User: Password:
|
|
Subscribe / Log in / New account

Changes and complaints

This article brought to you by LWN subscribers

Subscribers to LWN.net made this article — and everything that surrounds it — possible. If you appreciate our content, please buy a subscription and make the next set of articles possible.

By Jake Edge
February 22, 2012

Free software is always progressing—at least by someone's definition of "progress". Over the last few years, or more, we have seen huge controversies erupt around changes that were being made to desktop environments, system software, and other parts of the free software stack. The controversies themselves are not that surprising, as people are often resistant to change (and some proposed changes are not necessarily good ones), but the tone and manner of the complaints is sometimes rather eye-opening. It often sounds like some believe that the developers of these projects set out to ruin people's lives with their changes. Somehow that doesn't seem likely, and starting a discussion from that standpoint seems rather unlikely to change anyone's mind.

It is unquestionably frustrating and maddening to upgrade a package or distribution and find that things no longer work they way they once did; that workflows or other habits are now "obsolete". But, presumably, the upgrade was done for a reason, typically to get new features and fixes. With upgrades come dangers, of course. Sometimes those dangers are from new bugs or incompatibilities, but other times the danger is that something that once "worked" no longer does.

This is a problem that is in no way restricted to the free software world. One could easily argue that the problem is far worse for proprietary systems, as upgrades are sometimes forced, which is pretty uncommon for free software. Distributions do reach their end of life, but there are plenty of alternatives to consider when that happens. Somehow, choices like Windows Mint or Mac OS X "Wheezy" don't seem to be in the cards. Nor does installing an alternative desktop environment, display server, init system, kernel, or audio subsystem. For the most part, you get what the proprietary vendors give you and you like it—or not.

It could also be argued that the free software approach (to the extent there is a single "approach") does not lead to desktop dominance. That has, of course, been argued ad nauseam here and elsewhere. But the fact of the matter is that hackers and projects are making their own decisions, freely, based on their interests and understanding of the problems they are solving. No one has set out to annoy anyone. That may be a side effect of the changes that are being made, but it certainly isn't the intent.

If you peer into the development mailing lists or talk to the proponents of these kinds of changes, they clearly see them as improvements. It's a little hard to imagine that folks would spend significant amounts of their time making the code worse. Opinions will differ on the changes, of course, and making one's opinion known is a time-honored tradition in free software (not to mention the internet and human society in general). But the vehement, sometimes demanding, tone that those complaints take is counter-productive. Part of what seems to be lacking is a certain of level of respect for the projects and the people behind them.

A much more effective way to make one's opinion known is through engagement. Unlike the proprietary world, we in the free software world can directly engage with the developers, describe the problems that we see, and try to change the direction of the project in a way that is more suitable for our needs. Most free software projects discuss planned changes well in advance of their implementation and give users lots of opportunities to try out early versions. But engaging the project is best done with well-reasoned, specific descriptions of problems, missing features, and so on—not endless streams of "Project XYZ sucks!" messages to mailing lists or comment threads.

Beyond just offering suggestions and/or complaints, though, we can also pitch in and fix the problems that we see. Given the large number of vocal opponents of various changes (and proposed changes) that we have seen, there should be a ready supply of developers and others to continue maintaining older code bases, or to work on getting features added to fill in gaps. Even when a project moves on from an earlier version (a la GNOME 3 or KDE 4), it's not like the old code disappears. We have seen some efforts (like the Trinity desktop environment for KDE 3 and MATE for GNOME 2) but, despite all the complaints, a big community has not sprung up around either.

Part of the problem is that the intersection between those who are vocally unhappy and those with the time and skills needed to help is probably fairly small. Free software thrives where people participate, but folks tend to want to work on things that scratch their own itch. If there aren't enough participants with the "right" itches, few of the complaints will actually get solved.

Another related problem is that there seems to be an increasing sense of entitlement from some within our communities. The huge base of free software that we use today has been given as, essentially, a gift, from tens of thousands of contributors worldwide. Just like "those who write the code get to choose the license", "those who work on the project get to set its direction". Users are, of course, an important piece of the puzzle, but frothing, name-calling, endless complaining does not rise to the level of a contribution.

It may well be that some or all of the problems that people see in various projects (GNOME 3, KDE 4, Wayland, systemd, the Journal, grub2, ...) are serious and will drive users away from the projects or the distributions that adopt them. But there's really only one way to find out. If users vote with their feet and move elsewhere, one suspects the projects will follow along behind.

None of this is meant to downplay the importance of users making their problems known, but there is a point at which the repetitive, often content-free, complaints don't serve any purpose—other than venting perhaps. These projects have most certainly heard most of the complaints by now, sometimes an enormous number of times. Have they seen constructive attempts to address those problems in even a small fraction of that volume? That, unfortunately, seems unlikely.

In the end, though, people eventually either get used to the changes or find some alternative that suits them better. Sometimes that alternative involves a fork of the existing code and it is not unheard of that the fork eventually rejoins or takes over from the original project. The EGCS/GCC split comes to mind, for example, and we may be seeing something like that play out again with LibreOffice and Apache OpenOffice. It's also worth remembering that GNOME 2 was originally met with a lot of unhappiness, perhaps even from some of the same folks that are now complaining that GNOME 3 "took away" things they were used to in GNOME 2. Over time, these things have a way of working out, but in the meantime, it's worth at least thinking about whether that venom-filled post is really going to be very effective.


(Log in to post comments)

Changes and complaints

Posted Feb 23, 2012 2:26 UTC (Thu) by horen (subscriber, #2514) [Link]

"We have seen some efforts (like the Trinity desktop environment for KDE 3 and MATE for GNOME 2)..."

Where distro developers have given us GNOME 3, I respectfully suggest looking at Cinnamon, instead.

Changes and complaints

Posted Feb 23, 2012 2:41 UTC (Thu) by naptastic (subscriber, #60139) [Link]

> Just like "those who write the code get to choose the license", "those who work on the project get to set its direction".

What for those of us who need certain features, or have certain preferences, but don't code? I know there are no guarantees, but I pay money for free software. Shouldn't that count for something? I'm getting a little miffed at being marginalized because I don't have enough years of coding experience to fix things myself.

And let's be honest, the changes we're talking about are not trivial. Even if you're good, you'd have to spend time familiarizing yourself with a codebase before making a change to a preference. I recently made a theme for Gnome 2 that's exactly like Ambiance but a little more compact. It took hours of searching through Gnome widget documentation and theme source files to find the changes I needed to make, and there are still problems with it. My point is, asking people who aren't already eating and breathing the project's code to improve it themselves is pretty callous. I'm not saying malicious; I'm saying insensitive.

Ardour is awesome. I file a bug or feature request, attach money to it, it gets done, and I fork over the cash. I just want Gnome 2 to snap windows to half the screen, and stop asking me to delete panel applets when it fails to load them. (Just wait 2 seconds and retry. Ask me again if it fails 3 times in a row.) If I could keep Gnome 2, and pay $75 for each of those changes, I'd do it in a heartbeat.

My point is that the people who maintain a project really do need to be responsive to their users. Not everyone who uses FOSS is a hacker anymore. The attitude of "if you don't like it, fix it" needs to go the way of the BKL.

Changes and complaints

Posted Feb 23, 2012 5:57 UTC (Thu) by kberg (guest, #4963) [Link]

I would like to second naptastic's comment about the effort necessary familiarize one's self with with a large complicated code base before even being able to make the smallest of changes safely. Many people, even if they have the skill, don't have the time or energy to do this.

I'd also like to state the importance of workflow. I've been refining my current workflow for the past 15+ years. I first started using Gnome when it was at version 1.2 so I've been through the 1.4 -> 2.0 transition. IIRC, the biggest controversy of that transition was not the change in how Gnome operated, it was that many beloved configuration options were hidden or removed. Over the course of 2.x series most of desirable options returned (at least all the ones I cared for). The problem people seem to have with the Gnome 2.x -> 3.x transition is that Gnome Shell forces one particular usage model on everyone regardless of what workflow is efficient for them. My particular work flow involves many windows open on several workspaces and those workspaces are organized in 2D grid (rather than a 1D linear arrangement). Certain programs I always place in certain specific workspaces so that I always know where they are (this makes use of geographic memory). Even if I switched to using a 1D linear arrangement, Gnome Shell's dynamic workspace system royally messes up my geographic memory.

My observations about Gnome development seems to be that the developers come across some sort of academic paper describing some aspect human nature that can be used to make computer use more efficient. The problem with a lot of academic papers is that they are often limited in scope and do not take into consideration many real world scenarios. Anybody remember Spatial mode in Nautilus? I recall that the basic premise was that it would improve usage efficiency by relying on the user's muscle memory to access files. I believe that the Gnome developers even referenced an academic paper that served as the basis for Spatial mode. The real world problem with that is that for muscle memory to be reliable, the set of files accessible would need to remain static and if you had enough files that required scrolling the window, then any gains from muscle memory would be lost since too may different operations would be need to access the desired file. Isn't it interesting that Spatial mode quietly disappeared from Nautilus in Gnome 2.30?

As I understand it, the premise of Gnome Shell is that people are more efficient working in a document centric fashion. This is true, but only for a certain set of people trying to get a certain set of tasks done. A document centric UI would be great for an author focusing on a book or article, or a photographer sorting and editing photos from a photo shoot. It does not work so well for someone who needs to use several different tools at the same time and rely on the desktop environment to act as the unifying glue between those programs. Imagine doing networking SW development where you are editing and building on your local machine, remotely logged into 3 different machines for testing, running debuggers, and watching logs. A more flexible desktop environment makes this much less painful.

I'd also like to point out an example of a tool that forces the user the work in a specific way may not be a path to success. Years ago I was evaluating version control tools to use. I'd come across an open source tool that looked promising called Aegis. It was not selected because it forced users to send every code commit through both a review process and a test process before the code would be put in the main line. Is such a process good? It's certainly good for code quality. Unfortunately, for a small team, the process overhead and delays to getting code merged made Aegis completely unusable for us.

I think that if you remove the vitriol from comments about Gnome 3, Wayland, Journal, etc., what you will find is that people are concerned about how they will adapt their workflow, or even if they will be able to adapt their workflow to some of the changes that are being made in infrastructure they require to get things done. People would prefer to make what they know more efficient rather than having to learn completely new interfaces and paradigms.

As for the comparison to the proprietary desktop environments from Microsoft and Apple, the workflow for them have not changed in many years. The workflow for Windows 7 is still the same as for Windows 95. The workflow on Mac OS X is still very close to what it was on the original Mac (and they changed both the underlying kernel and processor architecture twice in that time). What will be interesting is how different Windows 8 with its Metro interface will be compared to Windows 7 and how it will be accepted by the masses once it is widely introduced.

One thing that I think would help would be for developers of these projects that directly affect users to make an effort to find out what their current user base needs and wants and to work with users to figure out how to make the users' lives easier without making other users' lives harder. Essentially, to reiterate what has been written in LWN before, there needs to be more communication between users and developers.

Changes and complaints

Posted Feb 23, 2012 9:30 UTC (Thu) by dgm (subscriber, #49227) [Link]

> What will be interesting is how different Windows 8 with its Metro interface will be compared to Windows 7 and how it will be accepted by the masses once it is widely introduced.

An interesting observation is the way such changes get introduced by Microsoft. I have read that the Metro interface will be an option in Windows 8, and it will work alongside the traditional interface. There could be many reasons for that. It could be because Microsoft engineers are cautious and do not force disturbing changes onto their customers, or simply that they cannot afford to break their network of ISVs. Whatever the reason, it is the right way to do it. Disruptive changes should be introduced carefully, as an option.

Note that this is also the way the kernel introduces and deprecates code. First it is an experimental option, then mature or maybe on by default, then marked deprecate, and finally, after a long enough delay, removed. People is given time to adapt, and nobody complains.

I have come to the conclusion that most people are not against change, but abrupt change, specially when perceived as unnecessary.

Changes and complaints

Posted Feb 23, 2012 9:46 UTC (Thu) by Cato (subscriber, #7643) [Link]

I like the Ardour model - if more projects made it easy for money to be attached to bugs/features, always subject to a developer accepting the request of course, it could help fund projects, improve free software and make users happy... surely a win for everyone.

The Kickstarter model is interesting - while that's to fund large-ish projects for new films/products etc, the same model of pledging $X to a feature could be used to raise enough cash to fund a developer - or at least to give them some income to help out, even if not covering the full cost.

I'm not saying cash should be the only driver, and developers would still work on what they want to mostly, but if users want something and it's a reasonable feature/fix, why not allow users to contribute directly to that feature rather than a general donations pool?

Changes and complaints

Posted Feb 23, 2012 10:48 UTC (Thu) by neilbrown (subscriber, #359) [Link]

Yes.
We know we can "vote with our feet", and leave one project for another.
We know we can "vote with our effort" and contribute to a project to make it suit us better.
But some time we want to "vote with our wallet" and contribute cash. That isn't so easy.

If everyone who complained about Gnome3 handed over $10 or so it would probably be a fair sized kitty to keep Gnome 2 running well... or something.

Of course the "or something" is a difficult bit and making a system that is transparent and reliable would be a real challenge. But I think there is a real need here.

I think people are willing to pay, and other people are willing to do all sort of things if they were free from having to earn a living. But getting the middle-ware working is still an unsolved problem.

Paying for open source software

Posted Feb 23, 2012 12:56 UTC (Thu) by dan_a (subscriber, #5325) [Link]

Is there anything to stop you putting this as a job on vworker.com (or a similar rent-a-coder equivalent?)

For small changes to existing projects this should be something an individual user could fund. For bigger changes a group of users could get together to raise the funds. It would then be up to you to decide whether you wanted to try and get the changes committed to the main codebase or maintain them as your own private patch.

Changes and complaints

Posted Feb 23, 2012 21:06 UTC (Thu) by jmorris42 (guest, #2203) [Link]

> If everyone who complained about Gnome3 handed over $10 or so
> it would probably be a fair sized kitty to keep Gnome 2 running
> well... or something.

It is worse. They explicitly designed GNOME3 to not interoperate with GNOME2, you can't have both installed at the same time. And pretty soon every major distribution will be shipping 3 and not 2. So you would have to fund the creation of what would at minumum be a major fork of one or more distros. And commit to endless backporting as every major app defaults to building against gtk3 and the rest of GNOME3 plus all the *kit and *manager plumbing being wired ever deeper into everything.

That is the kind of attitude problem that really horks off us lowly users. We get the choice of launch a major fork, something we by definition are ill equiped to deal with, or accept whatever upstream decides to feed us today. Since the upstream on too many open projects don't care a whit about 'end users' we are in many ways worse off than a Microserf. At least Microsoft cares about really ticking off the userbase. Sure they care most about their own bottom line and second about the OEMs (their actual customers) and third about "developers, developers, developers!" but the users are at least on their list of concerns... just a bit down the list.

Changes and complaints

Posted Feb 23, 2012 21:30 UTC (Thu) by khim (subscriber, #9252) [Link]

So you would have to fund the creation of what would at minumum be a major fork of one or more distros.

The fork already exist.

And commit to endless backporting as every major app defaults to building against gtk3 and the rest of GNOME3 plus all the *kit and *manager plumbing being wired ever deeper into everything.

It's not a problem. Gtk2 and Gtk3 can coexist. It may be even possible to port MATE to Gtk3 eventually.

Sure they care most about their own bottom line and second about the OEMs (their actual customers) and third about "developers, developers, developers!" but the users are at least on their list of concerns... just a bit down the list.

Actually users affect bottom line quite directly thus they are pretty high on the list. Especially big spenders like MS Office users.

Changes and complaints

Posted Feb 23, 2012 22:56 UTC (Thu) by brettle (guest, #34988) [Link]

Consider using Elveos to crowdfund your favorite features. It's specifically for funding free software projects and is free software itself.

One thing that I think is holding the Elveos funding model back is the need to prepay for all the features you are sponsoring. That's why I sponsored a feature to allow contributors to contribute the same money to multiple features.

Changes and complaints

Posted Feb 23, 2012 2:48 UTC (Thu) by cmccabe (guest, #60281) [Link]

The problem with creating new user interfaces, like the GNOME 3, Unity, the Gimp, and to a lesser extent KDE 4.x developers did, is that new user interfaces aren't judged "fairly." Before you can get people to even try a new user interface, you often have to convince them that it will be worth their time.

The computing world is full of user interfaces that are sub-optimal. Look down at your keyboard. It's QWERTY, right? That is a sub-optimal layout. But nobody wants to take the time to re-train all of the world to use DVORAK, or a similar layout.

Apple succeeded in getting people to try and like new user interfaces. Partly this is because they have good designers, but it's also because they aggressively market their products and are seen as a premium brand. The new Apple user interface _must_ be worth learning because it's the new Apple user interface. In the same way, the new Rolex watch must be worth the money because it's a Rolex.

User interfaces are often all about marketing and very little about actual functionality. Why did Apple promote a one-button mouse for so long? Because it was more efficient? Hardly. They promoted it because Steve Jobs though it was cool. (It's an Apple thing. You "wouldn't get it.")

The new user interfaces that are coming out of Cupertino and Redmond are often no more efficient or ergonomic than their predecessors. (The iPod wheel is a good example of a very unergonomic design, for example.) Despite this sad reality, open source software should still strive to copy the existing interfaces, rather than creating entirely new and different interfaces. Imagine how much more widely used Gimp would be if they had just tried to emulate Photoshop's UI, rather than creating a new one.

When we do create new user interfaces, there should be a clear and obvious efficiency benefit. And the creator should make people aware of those benefits as much as possible-- create awareness and build consensus rather than expecting people to immediately accept your new UI because you are so awesome.

Changes and complaints

Posted Feb 23, 2012 10:26 UTC (Thu) by Cato (subscriber, #7643) [Link]

Some good points. I believe the Apple mouse uses a single button because it avoids the user having to figure out which button to press on a 2 or 3 button mouse.

Only yesterday I was helping a highly intelligent professional with over 10 years experience on Windows who is still a bit confused on left vs. right button sometimes, and is often confused on when to click vs. double-click.

A single button with long-press (like Android's) might be a better design, as it's more discoverable.

Changes and complaints

Posted Feb 23, 2012 12:13 UTC (Thu) by jnareb (subscriber, #46500) [Link]

Some good points. I believe the Apple mouse uses a single button because it avoids the user having to figure out which button to press on a 2 or 3 button mouse. Only yesterday I was helping a highly intelligent professional with over 10 years experience on Windows who is still a bit confused on left vs. right button sometimes, and is often confused on when to click vs. double-click.
But because of lack of second button you have to use and know various combinations of mouse click with keyboard (Ctrl-Click equivalents).

Changes and complaints

Posted Feb 23, 2012 13:37 UTC (Thu) by raven667 (subscriber, #5198) [Link]

Not on MacOS you don't, you can optionally choose to learn the shortcuts but not knowing them doesn't break the UI.

Changes and complaints

Posted Feb 26, 2012 14:05 UTC (Sun) by faramir (subscriber, #2327) [Link]

How is a second or third button NOT discoverable? They are physically right there.

Changes and complaints

Posted Feb 27, 2012 17:46 UTC (Mon) by bronson (subscriber, #4806) [Link]

Inexperienced people see no reason to ever push them and are often too scared to even try.

Changes and complaints

Posted Feb 27, 2012 22:52 UTC (Mon) by HelloWorld (guest, #56129) [Link]

If people don't see a reason to push a mouse button in front of their nose, why would they try pushing a button for a longer time? I'm sorry, but that makes no sense at all.

Changes and complaints

Posted Feb 28, 2012 21:39 UTC (Tue) by bronson (subscriber, #4806) [Link]

Because they have no idea what would happen if they did push it and are therefore a little fearful of doing something wrong.

Have you never worked with inexperienced computer users?

Changes and complaints

Posted Feb 28, 2012 21:42 UTC (Tue) by HelloWorld (guest, #56129) [Link]

You didn't answer my question. If they, for whatever reason, don't try to press the right mouse button, then why would they try pressing the left button for a longer time? That just makes no sense.

Changes and complaints

Posted Feb 29, 2012 4:30 UTC (Wed) by bronson (subscriber, #4806) [Link]

I only showed why 2nd and 3rd buttons aren't very discoverable even though they're sitting right there on the mouse. I agree, long press is every bit as undiscoverable as double click or any other random gesture.

Changes and complaints

Posted Feb 28, 2012 21:43 UTC (Tue) by bronson (subscriber, #4806) [Link]

I'm not saying anything about long-press; that's probably just as confusing as double-click. Dunno.

I've seen people who think you double-click everything. EVERYTHING. It's amazing to watch.

Changes and complaints

Posted Feb 27, 2012 22:50 UTC (Mon) by HelloWorld (guest, #56129) [Link]

Speaking of Dvorak...
http://xkcd.com/561/

Changes and complaints

Posted Feb 23, 2012 6:00 UTC (Thu) by JoeBuck (subscriber, #2330) [Link]

My impression is that some of the people who are most vocally unhappy with the direction the desktop has taken recently are long-time contributors to the kernel and the core GCC toolchain, or to other non-GUI software. This is an unfortunate situation that will probably lead to fragmentation if the gulf is not healed.

I see problems on both sides: the kernel community has developed a culture in which abusive language is standard, and this kind of thing has affected other parts of the Linux community. It is common to insult the intelligence of programmers who, in the view of the accusers, have made bad design decisions (Linus does not post on the GCC development list to compliment people on the wonderful compiler, for example); the level of abuse tends to escalate until the "guilty party" agrees to make design changes.

But on the other side, GUI developers have been openly preaching that programmers' interests just don't matter to them. Programmers are power users, they are a small, insignificant fraction of the available audience. Opening a terminal is functionality that is best hidden; many terminals open at once is just not a relevant use case. Complaints that people with years of experience suddenly feel helpless don't seem to be taken seriously. If this state of affairs continues, the programmers that need a desktop will have to develop their own, and won't be eager to help those GUI programmers who build code that they don't care to use. We need each other.

So we need to do two things; we need to cut out the personal abuse and also start listening to each other. GUI developers need to listen to those with years of experience with Windows, Mac, and Gnome 2; at the same time, we old sticks in the mud need to be open to trying something new. The free software desktop has to be useful to software developers, because those software developers are contributing to making it better.

Changes and complaints

Posted Feb 23, 2012 7:30 UTC (Thu) by Los__D (guest, #15263) [Link]

There is only one way that "people with years of experience" suddenly feel helpless, where new users doesn't really have any problems: The are being stubborn and arrogant.

Their problem.

Changes and complaints

Posted Feb 23, 2012 8:25 UTC (Thu) by dlang (subscriber, #313) [Link]

and you have just proven that the abuse isn't all on the part of the long-term programmers

Changes and complaints

Posted Feb 23, 2012 10:55 UTC (Thu) by Los__D (guest, #15263) [Link]

I wrote my first program 30 years ago, so I'd consider myself a long-term programmer. I'm just not a stubborn conservative, afraid of changing workflows. That helps.

Whining about change doesn't (and shouldn't. The less conservatives are listened to, the quicker we get rid of historic garbage).

Changes and complaints

Posted Feb 23, 2012 11:47 UTC (Thu) by dgm (subscriber, #49227) [Link]

> I'm just not a stubborn conservative, afraid of changing workflows.

Congratulations, more power to you. You just need to admit that not everybody has to be like you, and that not being like you is not the same as being wrong.

Changes and complaints

Posted Feb 23, 2012 12:02 UTC (Thu) by Los__D (guest, #15263) [Link]

I have no absolutely problem admitting that. But them wanting to dictate what changes i.e. GNOME are allowed to make not to destroy their precious workflow (which they could change in a heartbeat, if they were just willing, instead of wasting their time kicking and screaming), I DO have a problem with.

Changes and complaints

Posted Feb 23, 2012 19:08 UTC (Thu) by lacos (subscriber, #70616) [Link]

"But them wanting to dictate what changes i.e. GNOME are allowed to make not to destroy their precious workflow (which they could change in a heartbeat, if they were just willing, instead of wasting their time kicking and screaming), I DO have a problem with"

You are *completely* misguided. Why do you think enterprise customers pay big bucks for stable enterprise distros? Because their workflows take time to polish and then become money-printing presses, figuratively. If you mess with their stability or workflow, you mess with their money-making capability. Go tell them "change your workflow in a heartbeat" and they will laugh you out of the room.

It is *exactly* the same with "conservative" programmers. Their efficiency (and project success and income) depends *critically* on their specific polished version of HCI.

"Easy" GUIs make easy things easy and complex workflows impossible.
Flexible GUIs make easy things somewhat harder and complex workflows possible.

Anyway I've given up on both GNOME and KDE years ago after an initial meet & greet, so I won't try to dictate anything on your precious mailing lists.

Changes and complaints

Posted Feb 23, 2012 19:16 UTC (Thu) by Los__D (guest, #15263) [Link]

You are actually comparing company-wide workflows, which must be figured out by commission, planned and tested, to a user?

You seem to be the misguided one.

Changes and complaints

Posted Feb 23, 2012 19:52 UTC (Thu) by lacos (subscriber, #70616) [Link]

"You are actually comparing company-wide workflows, which must be figured out by commission, planned and tested, to a user"

I most definitely am. You may have heard about productivity differences between programmers up to a factor of 10. Personal workflow is a significant part of that. But let's not turn it into a competition between programmers: the Zone(TM) is critical for getting things done, and if I'm interrupted by GUI concepts every minute for two weeks while I readjust, I won't produce anything for two weeks. I can't reason about a bug if I have to fight the GUI. The changed GUI would force actions that have become motor skills into my conscious, derailing my train of thought.

Additionally, one facet of a company's workflow is usually broken down into regulations for human workforce. Checklists, protocols, action plans.

Changes and complaints

Posted Feb 24, 2012 3:50 UTC (Fri) by ncm (subscriber, #165) [Link]

Not just "a factor of 10". More like three factors of ten, e.g. one programmer outproducing 500 colleagues on a single project is documented.

Nobody expects software that has not been tested thoroughly to work. A new kernel data structure that turns out to slow down sensible or common job mixes is rejected, no matter how clever. Somehow, though, a person reporting that a new UI design does not turn out, upon testing, to work better, or to work well, or even to work tolerably is a "whiner". When a clever new UI scheme slows down important work, we get accusations of Luddism.

What I have never seen is somebody explain how to achieve the same level of productivity with the new UI as with the old. Typically it's not possible. Instead, the task newly crippled is summarily declared to be marginal and uninteresting. Is there any usage pattern that can't be disparaged that way? I have not encountered any. Ultimately, anyone, taken alone, can be labeled marginal, and ignored.

When you find yourself looking for excuses to ignore people, that's arrogance.

Changes and complaints

Posted Feb 23, 2012 9:38 UTC (Thu) by jschrod (subscriber, #1646) [Link]

OK, you demonstrated that you're a part of the problem.

Joe presented specific criticism, not just the notion that "experienced people feel helpless". How about reacting to that instead? How about being a part of the solution?

Changes and complaints

Posted Feb 23, 2012 10:16 UTC (Thu) by dgm (subscriber, #49227) [Link]

> The are being stubborn and arrogant.

They are being conservative. And they don't want to change what works well, and is important to them.

We have seen user interaction paradigms, ideas and trends come and go. Some are good, many are not. The good ones stick and the rest fade away, but it takes a while, sometimes years. In the meantime, the obvious option is to wait and let things settle. But if you're not given the option because the change is forced onto you, ready or not... the only rational thing to do is complain. You're stuck between a rock and a hard place, and whatever movement you do is a losing movement, because it means changing environment and waste time adapting.

Is this being arrogant? Is this being stubborn? I don't think so.

I have been a KDE refugee, and now I'm a Gnome refugee too. Finally I have learned my lesson, and I'm moving to an environment that seems not interested in experiments, but in stability and keep things working, so people can do stuff. That's why I have switched to Xfce. Gnome will eventually undo the most glaring wrongs with the new design, as has KDE done. But I will not return to neither, because I don't want to be bitten by the same dog again. Twice is enough for me.

Changes and complaints

Posted Feb 23, 2012 19:36 UTC (Thu) by dlang (subscriber, #313) [Link]

heh, I'll point out it was two different dogs, not twice by the same dog :-)

as I see it GNOME goes through this sort of thing each time it does a major version bump (they had problems with the 1.x -> 2.x where they decided people didn't need options, then they gradually added options, then going from 2.x to 3.x they are again deciding that people don't need options). So while I have not been a heavy GNOME user (I used it for a while because it was the Ubuntu default), I'm not likely to give them much credit.

Changes and complaints

Posted Feb 23, 2012 16:25 UTC (Thu) by daglwn (guest, #65432) [Link]

Not true. Take the Wayland debacle. The Wayland developers seem to discount the importance of remote displays. This is not a "simple" user interface issue. It's core functionality that will just be gone. There is work I will not be able to do if our corporate distribution switches to Wayland.

So don't switch? That's fine for the individual user but in a corporate environment we don't have much choice if the vendor doesn't continue to support old versions forever.

Changes and complaints

Posted Feb 23, 2012 16:48 UTC (Thu) by JoeBuck (subscriber, #2330) [Link]

There isn't a "Wayland debacle": if the toolkits work on top of Wayland or on top of X, that holds us over until there's a good remote access solution for Wayland that beats what X does. Red Hat has a lot of corporate customers who will need remote access, so I'm confident that the resources to solve it will be available.

It's not as serious a concern as the desktop unhappiness.

Changes and complaints

Posted Feb 23, 2012 17:05 UTC (Thu) by daglwn (guest, #65432) [Link]

The X-as-a-client solution only works as long as the toolkits support X and the applications using the toolkits don't have to do anything special (build flags, etc.) to keep that X support.

I think it highly likely the toolkits will support X for a while but I don't know whether applications will have to link to them in a special way to get that support. Is it a mode bit in the toolkit library?

Changes and complaints

Posted Feb 23, 2012 20:59 UTC (Thu) by mathstuf (subscriber, #69389) [Link]

What do application developers do now when recompiling for Windows or Mac? IME, it's approximately nothing on the whole. I don't see why the decision between X and Wayland would be any different.

Changes and complaints

Posted Feb 27, 2012 22:45 UTC (Mon) by HelloWorld (guest, #56129) [Link]

If the toolkits drop X support (without offering a better replacement), then you should be complaining to them, and not to the Wayland developers for doing the right thing.

Changes and complaints

Posted Mar 1, 2012 3:25 UTC (Thu) by Zizzle (guest, #67739) [Link]

But that seems to the norm, at least in the GNOME/GTK+ world.

Hence why the mate project has to rename nearly every binary in the system.

The GNOME guys decided that GNOME2 was instantly deprecated as soon as you install GNOME3 - no parallel install possible.

I can see them doing the same thing once Wayland is usable.

Changes and complaints

Posted Feb 27, 2012 22:44 UTC (Mon) by HelloWorld (guest, #56129) [Link]

There is no "Wayland debacle". They simply made the technically sane decision to split up a large job (that of getting rid of X) into manageable pieces. Wayland only handles compositing and input, and these are completely orthogonal to network transparency.

Changes and complaints

Posted Feb 23, 2012 14:08 UTC (Thu) by marduk (subscriber, #3831) [Link]

This is probably the best comment I've read thus far. For the most part, I am in agreement. But ironically, I find that I am just the opposite. I think I was more stubborn and change-resistant when it comes to software in my earlier years. I started using Linux in the early 1990s and the internet before the WWW was popular. I remember when Netscape came out and it allowed web authors to change the background color of the web page and I considered that "cheating" and violating my user-defined preferences... back when "#BFBFBF" was the default background color. Nowadays if I see that color as the background color I barf. I didn't believe OpenGL/compositing belonged on the desktop, save for people who wanted to make their Linux boxes look like Macs.

Nowadays I'm mostly a "bring it on" kind of guy. I guess I just like to shake things up. Why not try something new, see what works and what doesn't. And while I'm still not on Facebook (although I used to be) and vim is still my editor/IDE of choice.

As for the kernel developer subculture... that's been discussed a lot before. I'm not one of those guys, and probably never will be. I do wish the back-and-forth could be more "academic" in style, but I'm on the outside looking in, so I can only say so much. I do think that sometimes when persons of that subculture enter in in dialog with peers of another subculture it can sometimes be counter-productive. I'm not sure anything can be done about that in the short term. My feeling is the problem will eventually fix itself.

Changes and complaints

Posted Feb 23, 2012 15:53 UTC (Thu) by k3ninho (subscriber, #50375) [Link]

There's a wider context you've missed: disempowered people who don't like what's going on in the culture around them (politically, culturally, etc.) and so raise a protest without necessarily effecting change. Web logs and forums allow the protest to be voiced.

It's made worse by the example set by the media. News journalism in print and on TV raises a racket, but isn't allowed to directly write the laws. So people who consume that media follow that model: raise a racket, but don't write the software.

We can't give up noting the change in mind-set needed to engage with the free software culture. I've got to remember to ge involved and to play my part where contribution is expected and normal. Working out disagreement when you can say 'show me the code' is way better than a sea of critical blog postings, but that requires the soft skills to bring people inside the fold.

K3n.

Changes and complaints

Posted Feb 23, 2012 19:09 UTC (Thu) by lacos (subscriber, #70616) [Link]

"I see problems on both sides"

Thanks, this comment is the best I've seen in a long time.

Changes and complaints

Posted Feb 23, 2012 20:38 UTC (Thu) by drago01 (subscriber, #50715) [Link]

> But on the other side, GUI developers have been openly preaching that programmers' interests just don't matter to them.

This makes no sense at all because your basically saying that there *own* interests doesn't matter to them.

I can assure you that this is wrong.

Changes and complaints

Posted Feb 23, 2012 21:25 UTC (Thu) by jmorris42 (guest, #2203) [Link]

> This makes no sense at all because your basically saying that
> there *own* interests doesn't matter to them.

It would seem to be a problem but I have seen the pattern enough to believe it too. Too many of them seem to be on an almost religious quest to bring about 'The Year of Linux on the Desktop' and are convinced that if they could just make a dumbed down enough product that suddenly it would become the next 'big thing.' Yes we know it is daft, we know that, at a minimum, without a solution to the preload problem the 'Year' ain't ever happening; but they do seem to believe it to the point they will act on it.

So instead of building a desktop that would be ideal for themselves and the existing user base we see them building a desktop for an imagined new set of users, greater in numbers, large enough to sustain commercial development, able to contend with Apple for 2nd place or... whispered in the dark, someday Microsoft itself.

Changes and complaints

Posted Feb 23, 2012 21:44 UTC (Thu) by drago01 (subscriber, #50715) [Link]

> It would seem to be a problem but I have seen the pattern enough to believe it too.

Doesn't change the fact that this is wrong. I am way more productive using the GNOME3 interface as with an windows 95 style desktop. So are lots of other people (including the other GNOME developers).

> So instead of building a desktop that would be ideal for themselves [..]

Citation needed.

Changes and complaints

Posted Mar 3, 2012 2:20 UTC (Sat) by wmf (subscriber, #33791) [Link]

This makes me wonder: Is Unity developed using Unity? Do GNOME 3 developers use GNOME 3? If so, this suggests that it is possible to achieve some reasonable level of productivity. If they are cross-developing from XFCE or something, then there's a real problem.

The change is really tighter integration

Posted Feb 23, 2012 6:38 UTC (Thu) by grahame (subscriber, #5823) [Link]

The change is towards tighter integration between components of the operating system. In many ways this apes proprietary systems, and reduces the flexibility we have to roll our own solutions to strange problems. I think the value of that flexibility is being missed as the core, paid developers fix on the "desktop with a web browser" use case.

It's becoming more and more difficult to reject components of the stack that aren't fit for a particular purpose; systemd, pulse, dbus, PackageKit, abrt, etc. These tools all expose a complex interface that can't be easily emulated by another package, or just shimmed out.

It makes it easier for the people writing the desktop, but harder for people doing unusual things where they might want to reject one or more of those components.

The change is really tighter integration

Posted Feb 23, 2012 19:19 UTC (Thu) by lacos (subscriber, #70616) [Link]

Thank you. You got a grip on the problem at its core. Especially apt are the components you singled out; those are the ones that drive me crazy the most. Add in avahi as well.

Recent changes are turning the desktop into a single-purpose device. I don't mind a single-purpose device as long as it is my car or my toothbrush. But my desktop is my dang toolbox! I'm trying to work here!

The change is really tighter integration

Posted Feb 27, 2012 23:12 UTC (Mon) by HelloWorld (guest, #56129) [Link]

I don't see your point. systemd, dbus, pulseaudio, PackageKit and abrt are all entirely optional, and lots of people choose not to use them. Of course, you do lose functionality if you reject them, but that's normal, right? And it's not like everything breaks.

Also, it's all open source software, so anybody is free to make any modifications necessary for their use case. If people would hack as much as they bitch about other people's programs, we'd probably have had the year of the linux desktop 10 years ago.

Forced upgrades?

Posted Feb 23, 2012 9:14 UTC (Thu) by khim (subscriber, #9252) [Link]

Wow. Just wow.

One could easily argue that the problem is far worse for proprietary systems, as upgrades are sometimes forced, which is pretty uncommon for free software.

Is this some kind of joke or are you… gasp… serious?

Yes, in proprietary world people tend to push upgrades strongly, but this is not even close to what happens in Linux world. It's well-known fact that most organizations skipped both Windows ME and Windows Vista "upgrades". A lot of companies are still using Office XP for crying out loud! And yes it's still supported! Sure, they will be forced to abandon it soon, but hey, they had half-dozen years and three different versions of Office (2003, 2007, 2010) to pick upgrade path!

Is it even an option to skip one release for Ubuntu LTS user? No: Ubuntu LTS releases are spaced apart at two years intervals while support only lasts for three years (I'm talking about desktop: server is supported longer… and there are no riots on server front). And don't start about regular, non-LTS releases: they come out so often it just does not worth talking about.

Thus no, one could NOT easily argue that the problem is far worse for proprietary systems. Because they don't force upgrades on their users nearly as strong.

Sure, free software upgrades are free and proprietary upgrades cost money, but this argument only works if other costs are excluded - and these changes (when people need to spend time to relearn what they already knew) incur costs. No wonder there are riots!

The superstrongly enforced upgrades in the Linux desktop world are the very core of the issue under discussion! It's easy to don't like ribbon interface, but you have literally years to make yourself accustomed with it (MS Office 2003, 2007 and 2010 happily coexist in one system) - nothing like that exist in free software world. In most cases the option to choose GNOME2 or GNOME3, KDE3 or KDE4 is not even presented to end user! Sure, you can choose totally different distribution, but this looks like a little too step price for the ability to use latest laptop (and Sandy Bridge is not supported by Lucid Lynx, for example).

Unlike the proprietary world, we in the free software world can directly engage with the developers, describe the problems that we see, and try to change the direction of the project in a way that is more suitable for our needs.

And this, again, fail spectacularly. Compare situation with the product which enforces upgrades even more strongly: Google Chrome. It offers you an easy way to see what happens on the bleeding edge (for non-Windows users there are Chromium Builds). Compare with free software projects where you need a lot of efforts to even see latest version. And bug reports against distribution version (which is often the only version end user ever observes) are usually ignored as not relevant.

Distributions are biggest obstacle here: sure they solve a lot of other problems, but then basically build impenetrable wall between developers and users, which exaggerates the problem (again, created by distributions) of superforced upgrades. And distributors themselves usually lack the manpower to address problems raised by users.

Forced upgrades?

Posted Feb 23, 2012 10:34 UTC (Thu) by nim-nim (subscriber, #34454) [Link]

> Distributions are biggest obstacle here: sure they solve a lot of other
> problems, but then basically build impenetrable wall between developers
> and users,

Which is just a convenient excuse. Don't drink the Wallyesque kool-aid, sometimes the developer complaining (should I say boasting?) about the 'impenetrable wall' is even the same person as the packager that pushed the problem version in the distribution.

Not looking at distro reports is 100% the developer choice (and other software projects, which are not even managed by people paid by a distro company, have no such problems).

Drowning under distro problem reports is 100% due to buggy or incomplete software that needs fixing. If you move users from distros to somewhere else the software won't be any less buggy or incomplete and the reports any less numerous.

The same upstream bugs that go unfixed for years are duped by Fedora people, then Suse people, then Ubuntu people, then Debian people, following the deployment calendar of the problem software versions in various distributions, you can see the waves of dupes that occur when an version is deployed by a new distro, and it has nothing to do with any particular distro and everything to do with plain unappealing software bugs.

Solid projects are *happy* to see their software used in many different contexts since that helps identify difficult-to-trigger bugs (see this year's FOSDEM LibreOffice presentation, the many times Linus stated how happy he is to see the same kernel used on many archs and from embedded to big-iron, etc).

But it is easier to shoot the messenger than to fix problems.

Denial is powerful thing...

Posted Feb 23, 2012 11:50 UTC (Thu) by khim (subscriber, #9252) [Link]

Don't drink the Wallyesque kool-aid, sometimes the developer complaining (should I say boasting?) about the 'impenetrable wall' is even the same person as the packager that pushed the problem version in the distribution.

It does not make wall less impenetrable, sorry.

If you move users from distros to somewhere else the software won't be any less buggy or incomplete and the reports any less numerous.

Sure they will. People who consciously install and test nightly builds tend to produce much better, and more importantly, more timely bugreports.

The same upstream bugs that go unfixed for years are duped by Fedora people, then Suse people, then Ubuntu people, then Debian people, following the deployment calendar of the problem software versions in various distributions, you can see the waves of dupes that occur when an version is deployed by a new distro, and it has nothing to do with any particular distro and everything to do with plain unappealing software bugs.

Absolutely not. Most bugs are fixed pretty fast. What is not fixed are particular design decisions: should we merge this two dialogs or keep them separate, should we sort everything before showing on screen or should we rememeber old position, how this or that key combo should work. And by the time Fedora people, then Suse people, then Ubuntu people, then Debian people, following the deployment calendar of the problem software versions in various distributions start complaining it's always way too late: design decisions are approved and work is complete. You can try to fix them but then you'll get the feedback in a half-year or so when Fedora people, then Suse people, then Ubuntu people, then Debian people will start seeing new version. Which will trigger yet another wave of useless bugreports.

This is the same problem as with ISVs, really: nightly builds of the software are in the same position as out-of-the-tree ISVs software and since there are no easy way to deploy independent packages for ISVs there are no easy way to deploy in-development package for usually-in-tree software.

Solid projects are *happy* to see their software used in many different contexts since that helps identify difficult-to-trigger bugs (see this year's FOSDEM LibreOffice presentation, the many times Linus stated how happy he is to see the same kernel used on many archs and from embedded to big-iron, etc).

All projects are happy to see difficult-to-trigger bugs, no matter who finds them and where (they are especially happy to see reproducible difficult-to-trigger bugs, of course). What they are not happy to see is avalanche of pointless bugreports which question some design decisions long after the time of when these decisions were discussed, tested and implemented by a project. And distribution's main-in-the-middle role is important in breaking traditional alpha-beta-release cycle employed by developers of other platforms.

Denial is powerful thing...

Posted Feb 23, 2012 13:16 UTC (Thu) by nim-nim (subscriber, #34454) [Link]

> Sure they will. People who consciously install and test nightly builds
> tend to produce much better, and more importantly, more timely bugreports.

ROTFL. I run rawhide at home. Sometimes versions are pushed to rawhide hours before they get pushed upstream (because the rawhide packager is also the upstream main developer, and the rawhide process is faster than the upstream process).

That changes zip in the way problem reports are treated. Instead of ”your software is too old” you get:
– “I don't run a devel stack myself, please use old stable versions for everything but my own code and reproduce” (if you can't be bothered to run the early code of others, why do you insist users should run yours before you deign read their reports)
– “your software version is too new, we'll wait to see if there is still a problem later in stabler versions”
– “please retest or I close this report” months later (by someone who clearly never bothered investigating the first report)

Instead of “the decision has been taken a long time ago, it's too late to change it” you get “real users that do not run development distros will think otherwise” or “you'll see the awesomeness of our design once it has matured” responses.

And when the same problem hit more stable distros, and users do not find the design decisions awesome at all, and point out the very same problems than me months before, do they get a “we should have fixed that before, it was reported a long time ago” answer? No, they get exactly the excuses you just wrote.

Please climb down from your ivory tower.

(And I should also point out that some projects do take into account early bug reports, and they are also the same ones that take into account late bug reports, or reports against another distribution, and you never see them complaining loudly that distributions put an impenetrable wall between them and their users. They are awesome, even if they never make headlines discussing why they are entitled to ignore their user's feedback).

Denial is powerful thing...

Posted Feb 23, 2012 16:08 UTC (Thu) by khim (subscriber, #9252) [Link]

I run rawhide at home. Sometimes versions are pushed to rawhide hours before they get pushed upstream (because the rawhide packager is also the upstream main developer, and the rawhide process is faster than the upstream process).
That changes zip in the way problem reports are treated.

Of course. Why should it change anything? Now you have unstable, constantly moving system which includes god knows what.

– “I don't run a devel stack myself, please use old stable versions for everything but my own code and reproduce” (if you can't be bothered to run the early code of others, why do you insist users should run yours before you deign read their reports)
– “your software version is too new, we'll wait to see if there is still a problem later in stabler versions”

IOW: developers sensibly expect to see how their work behaves in isolation. This is how alpha/beta software was (and is!) always tested everywhere except in Linux world. And you are most definitely are not forced to participate in beta testing, but then, when the product is released, you are stuck with decision fixable at beta stage.

Sure, integration testing is important, but this is separate issue. It's not a replacement for plain old testing.

– “please retest or I close this report” months later (by someone who clearly never bothered investigating the first report)

This, again, is separate issue. This is the question of what is better: CADT model of handling bugs or bugs open and forgotten for years. I myself prefer bugs which are open for as long as it's needed to fix them (just recently bug which I've opened on GCC tracker was fixed... five years after I've filled it), but different people have different preferences.

Please climb down from your ivory tower.

Funny that users of other OSes (and there are more of them then users of all Linux Desktop goodies combined) don't think we work in ivory tower, only Linux people expect that we need to jump through 10 times more hoops to deliver software to 10 times less people.

One example: check this instruction. See the supported versions? Right: five year old and two year old compiler. Brand-new half-year old one is not even mentioned (today it's over year old and is supported... in experimental mode). The same with Mac (half-year old Leon is not even mentioned) and the same with Linux (GCC 4.6 is not supported yet. You may run into some build errors (and patches are welcome to fix them). Please see http://crbug.com/80071 before you proceed.): an aforementioned bug is actually fixed but I suspect recommendation to use older GCC will only be officially lifted after release of Precise Pangolin.

Distribution are supposed to help users with installation of software from the repo - and this process works reasonably well - but the fact that each release has it's own repo actively hurt ISVs (because there are no process for installing anything NOT in the repo) and the fact that some pieces developed by said ISVs end up included in the distribution does not change the equation much.

In the end it hurts users as well because many of them just want one or two bleeding-edge pieces - but this is basically impossible to organize in the distributions-driven world. Some software is backported in various PPAs, but this is half-hearted effort at best: there are no way to even deliver software to Ubuntu users using PPAs unless you'll build 3-4 different packages and other distributions require still more work.

As I've initially: I'm not saying that distributions are all bad, they certainly solve some real-world problems. But they also introduce some scalability problems with their "all or nothing" approach - and this hurts everyone: ISVs, developers of upstream packages and users.

In fact this is why GNOME3-like upgrades are met with such hostility: in the distribution-driven world you only have two choices:
1. to accept new interface right away when it's not yet refined enough, or
2. to reject the change - and be stuck with obsolete versions of all other programs.

In Windows world half-backed Windows ME and Windows Vista were just skipped and people went straight to Windows XP (after suitable hardware upgrade) and/or fWindows 7 - but they had access to all "latest and greatest" goodies in the meantime. In Linux we have huge flamefests instead.

You said it best yourself: it is easier to shoot the messenger than to fix problems.

Denial is powerful thing...

Posted Feb 23, 2012 19:11 UTC (Thu) by patrick_g (subscriber, #44470) [Link]

>>> In Linux we have huge flamefests instead

It's easy to access brand new software, even when you use an old distro. I installed Debian Squeeze on my laptop but I use Firefox 10.02 and LibreOffice 3.5.
It's a one clic install.

Denial is powerful thing...

Posted Feb 23, 2012 20:05 UTC (Thu) by nim-nim (subscriber, #34454) [Link]

> IOW: developers sensibly expect to see how their work behaves in isolation.
> This is how alpha/beta software was (and is!) always tested everywhere
> except in Linux world.

No, some developers want the luxury of dedicated tester drones, while rejecting for themselves the support constrains of traditional development houses. That's basically why the relationship sours now.

Everywhere except in the Linux world is where testers are paid to test instead of providing a free service.

Everywhere except in the Linux world is where desktop beta testing is opened to valued and respected customers – in the Linux world salespeople of paid-for distros will point customers to their community distro for beta testing, and then when they'll start reporting issues they'll get told by the developers of said distro “sorry, not good enough, go away”. In fact those developers seem to deliberately target a userbase that is nothing like the one that finances their employer, small wonder the whole thing seems to never go anywhere.

Everywhere else in the Linux or free software world, people try to do their best for their current users, no questions asked (and those users usually contribute bits back in some other part of the free software ecosystem, it's a a big family). But some projects that grew out of this ecosystem have decided things were better “everywhere except in Linux world”, with all the empathy a spoiled teen has for the rest of his family.

Denial is powerful thing...

Posted Feb 23, 2012 20:46 UTC (Thu) by khim (subscriber, #9252) [Link]

Everywhere except in the Linux world is where desktop beta testing is opened to valued and respected customers – in the Linux world salespeople of paid-for distros will point customers to their community distro for beta testing, and then when they'll start reporting issues they'll get told by the developers of said distro “sorry, not good enough, go away”.

Have you ever participated in said beta testing programmes? If you try to push changes of the magnitude people demand from Linux Desktop developers then you'll just be laughed out and perhaps excluded from future testing. If Microsoft decided to go with new interface then this is what you'll get and no amount of complains can change that. Sure, you can offer some ideas about how to make it better, but the big decision itself? Unchangeable. This is what the parent article is all about.

Everywhere except in the Linux world is where testers are paid to test instead of providing a free service.

Testers are usually paid when product is sold for $$. Free products don't always have dedicated testers even everywhere except in the Linux world.

But some projects that grew out of this ecosystem have decided things were better “everywhere except in Linux world”, with all the empathy a spoiled teen has for the rest of his family.

This happens elsewhere, too. When there are not enough existing users (or when their numbers is dropping fast) they often are abandoned and new, incompatible, offer is presented (usually to the loud choir of complains). Some of the changes are successful (think Netscape to Firefox transition, or MacOS to MacOS X transition), most are not (PalmOS to webOS transition or Sinclair ZX to Sinclair QL). Of course only Linux Desktop developers think it's good idea to do such major changes without offering coexistence period, but as I've pointed out before this problem is blown out of proportion because of "all or nothing" approach to software practiced by Linux distributors.

Denial is powerful thing...

Posted Mar 2, 2012 17:34 UTC (Fri) by wmf (subscriber, #33791) [Link]

The issue is that desktop Linux is too marginal to live. Canonical can't even afford to develop Unity properly, so they definitely can't afford to develop a smooth migration path from GNOME 2. Same with GNOME 3. But they won't give up, so the situation will continue.

Forced upgrades?

Posted Feb 23, 2012 19:23 UTC (Thu) by dlang (subscriber, #313) [Link]

> The superstrongly enforced upgrades in the Linux desktop world are the very core of the issue under discussion! It's easy to don't like ribbon interface, but you have literally years to make yourself accustomed with it (MS Office 2003, 2007 and 2010 happily coexist in one system)

except for the fact that once anyone you deal with upgrades to MS Office X, they start sending you documents in the new default format for that version, and so you had better upgrade to it now if you want to be able to read and manipulate the documents that they send you.

Forced upgrades?

Posted Feb 23, 2012 21:08 UTC (Thu) by khim (subscriber, #9252) [Link]

except for the fact that once anyone you deal with upgrades to MS Office X, they start sending you documents in the new default format for that version, and so you had better upgrade to it now if you want to be able to read and manipulate the documents that they send you.

Nope, you don't need to do that.

Link to beta version is long gone, but released version of compatibility pack is still there.

It's not perfect and sure, new features will not be available to you but it covers 99.99% of important cases.

It's not uncommon for the large company to spend year to just develop an upgrade strategy! Upgrade itself may take two, three or five more years. And not everything is upgraded at once: company may decide to switch to MS Office 2007 and keep MS Visual Studio 2005. Or it may decide to keep MS Office 2003 and switch to MS Visual Studio 2008. Not something you can [easily] do in Linux world.

Changes and complaints

Posted Feb 23, 2012 9:57 UTC (Thu) by michaeljt (subscriber, #39183) [Link]

> A much more effective way to make one's opinion known is through engagement. Unlike the proprietary world, we in the free software world can directly engage with the developers, describe the problems that we see, and try to change the direction of the project in a way that is more suitable for our needs. Most free software projects discuss planned changes well in advance of their implementation and give users lots of opportunities to try out early versions.

Here I think is an important clue to the problem. Many people would like to be simple consumers of free software without having to follow the development of every part that they depend upon. Fair enough, you might say, pay for RedHat and you can do just that. It is made worse though by the fact that many of those people also want to stay at the bleeding edge in one or two places. They want to have the latest version of whatever they are actively following available while the rest just stays as it was or improves slowly and predictably. This goes against the whole philosophy of distributions, especially given the heavy use they make of dynamic linking, so that everything tends to be upgraded at roughly the same pace - either everything is very old (Red Hat and friends) or everything is very new (Fedora is a good example).

Changes and complaints

Posted Feb 23, 2012 14:46 UTC (Thu) by NAR (subscriber, #1313) [Link]

They want to have the latest version of whatever they are actively following available while the rest just stays as it was or improves slowly and predictably. This goes against the whole philosophy of distributions

Exactly. I might want the latest Firefox with the latest plugins, but I do not want new desktop toolbars or menus. I can do this on Windows, on a supported(!) OS I've got a user interface that didn't change in 15 years (there's a Win95 theme for XP and even for Vista) and the newest browsers. Would I be able to run the latest Firefox on Debian woody? I hardly think so. And as a user I absolutely don't care if Windows has "unmaintenable, unprofessional, messy" code to achieve backwards compatibility, I'm only interested in the result.

Changes and complaints

Posted Feb 23, 2012 16:26 UTC (Thu) by michaeljt (subscriber, #39183) [Link]

I've been wondering quite a bit about how this would be achievable on a modern Linux. Obviously statically linking would solve some of the problems, and I suspect that the larger disk and memory consumption would hurt more psychologically (those who noticed it at all) than in any noticeable way. Possibly with faster loading as a compensation prize. I'm not sure if the extra network bandwidth for updates would be painful (particularly in places with less of the stuff available), though one could imagine getting the static binaries unlinked from a distribution and linking them on arrival. Nonetheless, there is a lot of infrastructure that you really want to share, like graphical tool kit plug-ins and configuration, which might get rather painful if you were using different static versions of the tool kit libraries in different applications. (Evil voices inside me wonder about using the system copy of Python to access the installed tool kit, with script that can handle various versions of Gtk+ and Qt, and shipping a static fall-back in case that fails.)

Changes and complaints

Posted Feb 23, 2012 16:47 UTC (Thu) by khim (subscriber, #9252) [Link]

I've been wondering quite a bit about how this would be achievable on a modern Linux.

Modern Linux is fine. Modern GLibC is fine. Modern Xlib and few other pieces are fine. Even dpkg/rpm are fine!

They system breaks apart at the next level: sadly it was not a design goal for all these fancy repos to give you choice. They are designed with "everything must be of the latest version" approach.

Nonetheless, there is a lot of infrastructure that you really want to share, like graphical tool kit plug-ins and configuration, which might get rather painful if you were using different static versions of the tool kit libraries in different applications.

Right. That's why you should design you plugin systems to be upgradeable and side-by-side installable. This is not that hard, but it must be an explicit design goal. As long as answer to "how can I install some program" is "it should come from your distribution repo" we are stuck with all these problems.

P.S. The funny thing is that while developers are hurt by distributors they often depend on them, too: they provide new versions of things which are not compatible with older ones and not parallel-installable, simultaneously! This only works because distributor repos make it possible.

Changes and complaints

Posted Feb 23, 2012 16:57 UTC (Thu) by michaeljt (subscriber, #39183) [Link]

> Modern Linux is fine. Modern GLibC is fine. Modern Xlib and few other pieces are fine. Even dpkg/rpm are fine!

> They system breaks apart at the next level: sadly it was not a design goal for all these fancy repos to give you choice. They are designed with "everything must be of the latest version" approach.

Quite agree. I was including your "next level" in my vague "modern Linux".

> Right. That's why you should design you plugin systems to be upgradeable and side-by-side installable. This is not that hard, but it must be an explicit design goal. As long as answer to "how can I install some program" is "it should come from your distribution repo" we are stuck with all these problems.

Again, I agree, but my musings were/are about how to force what we have now into working with applications taken from other sources without things becoming too fragile. Not saying my evil thought above falls into that category!

Changes and complaints

Posted Feb 23, 2012 19:01 UTC (Thu) by tzafrir (subscriber, #11501) [Link]

It's not a design choice of Firefox to be very compatible. Linux distributions are about integration.

Want the latest Firefox with your Stable Debian system? http://mozilla.debian.net/ .

It works because there's a dedicated Debian packager maintaining this. But hey, the packages from Mozilla.com only work because of dedicated packagers, right?

The plug-ins of Firefox are really not designed to be installed as system packages (though from what I hear, the said packager is working on it).

Another question: a bug in libpng is discovered and fixed. What packages should be rebuilt?

Changes and complaints

Posted Feb 23, 2012 19:29 UTC (Thu) by lacos (subscriber, #70616) [Link]

"Another question: a bug in libpng is discovered and fixed. What packages should be rebuilt?"

The base system's provider would rebuild the shared lib and push the update. All dependent base system apps would benefit.

The vendor providing the extra package (statically linked with a specific version of libpng) should monitor all libs they link statically into the app. When there's a security advisory for libpng, they should backport the fix (or grab the new upstream release if appropriate), rebuild their app with the fixed lib, and push the product via a separate channel.

I have the impression this is how Firefox works on Windows. (Except they may not link statically, just maintain their private set of DLLs.)

Changes and complaints

Posted Feb 24, 2012 9:37 UTC (Fri) by michaeljt (subscriber, #39183) [Link]

> The vendor providing the extra package (statically linked with a specific version of libpng) should monitor all libs they link statically into the app. When there's a security advisory for libpng, they should backport the fix (or grab the new upstream release if appropriate), rebuild their app with the fixed lib, and push the product via a separate channel.

I don't even see why they need to do this manually. Linux distributions today handle security updates almost transparently to the system user, and this could be pushed up a level so that the statically linked "extra" package is automatically rebuilt and re-downloaded by its users. Granted, it ought to be tested first, but that doesn't really happen now in the dynamically linked situation we have.

Changes and complaints

Posted Feb 23, 2012 19:35 UTC (Thu) by khim (subscriber, #9252) [Link]

It works because there's a dedicated Debian packager maintaining this. But hey, the packages from Mozilla.com only work because of dedicated packagers, right?

Right. But there are a difference: you need one packager for Windows (90% of market), one packager for MacOS (7% of market) and one packager for iOS and one packager for Android (which are already significantly more popular then Linux) - yet you need dozen of packagers for Linux.

Another question: a bug in libpng is discovered and fixed. What packages should be rebuilt?

Ah, the infamous "security" strawman. Yes, all packages must be rebuilt in MacOS and Windows case but only one package in Linux case. Except it's not true: Firefox bundles it's own fork of libpng anyway thus it must be rebuilt, too.

The problem with this approach is that it just does not work. If users are not using your platform then it does not really matter how secure it is. Sure, if we are discussing some kind of military OS then it may make some sense… except it fails there, too: military starts from most popular OS, not from the most secure one.

Changes and complaints

Posted Feb 24, 2012 0:51 UTC (Fri) by jengelh (guest, #33263) [Link]

>one packager for Windows, yet you need dozen of packagers for Linux.

Judging from the fact that VMware, NVIDIA and ATI do not seem to give a damn about organized distro integration and instead produce self-extracting sh archives that spew untracked files into one's system, it seems more like there is just one person responsible for the Linux package, and the remaining 10 members of the team know how to edit the Windows MSI.

Meanwhile, it would not be all that hard to use [an own instance of] a potent cross-distro build system like Open Build Service to cater for a number of distros in two swoops.

Changes and complaints

Posted Feb 24, 2012 12:17 UTC (Fri) by khim (subscriber, #9252) [Link]

Judging from the fact that VMware, NVIDIA and ATI do not seem to give a damn about organized distro integration and instead produce self-extracting sh archives that spew untracked files into one's system, it seems more like there is just one person responsible for the Linux package, and the remaining 10 members of the team know how to edit the Windows MSI.

Well, that's certainly sensible choice: this about how much marketshares differ thus it's prudent allocation of resources.

Meanwhile, it would not be all that hard to use [an own instance of] a potent cross-distro build system like Open Build Service to cater for a number of distros in two swoops.

Yes, it's possible, but it does not solve the whole problem. It's not as hard to build your package (well, this is a problem, too, but a [relatively] minor one). It's the fact that you need to [test] it. Single self-extracting sh archive solves this problem to some degree, cross-distro build system does not. And even single self-extracting sh archive requires a lot of distro-specific code to handle [supposedly] trivial things (like the ability to run per-machine or per-session background process).

All these things are just a band-aids and don't solve the core problem.

Changes and complaints

Posted Feb 23, 2012 22:46 UTC (Thu) by misc (subscriber, #73730) [Link]

> They are designed with "everything must be of the latest version" approach.

No, that's designed of "everything must be of a specific version", which is the only sane way of designing it without having a combinatorial explosion.
( ie, how to test that you installation is good for kde 3 and 4, for gnome 2 and 3, with and without specific part, be it new and old xorg, with or without pulseaudio ).

Sure, someone could do a distribution with everything "old" except firefox. Still, some people would complain, because there is the new kde, etc. And if this was so easy, someone would have done it already.

The truth is that the only way to do it is to use something like gentoo ( DIY distribution ), and I see few people doing it. So either the drawbacks of gentoo are not enough to convince people to use it, even if binary one do not satisfy them, or the vocal minority is just that, a minority.

Changes and complaints

Posted Feb 24, 2012 0:11 UTC (Fri) by khim (subscriber, #9252) [Link]

No, that's designed of "everything must be of a specific version", which is the only sane way of designing it without having a combinatorial explosion.

Well, there are another way: actually build interchangeable components. Create some kind of "core OS" and make sure other subsystems don't share anything. This is how it's done in other OSes, after all.

ie, how to test that you installation is good for kde 3 and 4, for gnome 2 and 3, with and without specific part, be it new and old xorg, with or without pulseaudio

Well, OS exist to give users the ability to run the software of their choice. If it can not provide this capability even for small subset of the available software then something is most definitely rotten in the state of Denmark.

Sure, someone could do a distribution with everything "old" except firefox. Still, some people would complain, because there is the new kde, etc. And if this was so easy, someone would have done it already.

True. It's not easy. Does not make the problem any smaller. Either you need the ability to "pick and choose" or you need to actually make upgrades smooth and unnoticeable. No GNOME3 or KDE4 "revolutions".

The truth is that the only way to do it is to use something like gentoo ( DIY distribution ), and I see few people doing it.

Does not work either. Gentoo allows you to toggle different knobs. But different version? Nope. You can push some packages back and forth - but to very limited extent. Even mix of x86 and ~x86 is often problematic!

Changes and complaints

Posted Feb 23, 2012 19:32 UTC (Thu) by dlang (subscriber, #313) [Link]

hmm, I do just this very easily on Ubuntu, I install the PPA for firefox and it gets upgraded frequently to the latest and greatest, while the rest of my system remains the same.

I don't choose to use Unity or Gnome, and I find the KDE desktop to be consistent enough across the updates to not be an issue for me. I guess I don't use enough "Desktop Features", even the early 4.x versions weren't a real problem for me.

Changes and complaints

Posted Feb 23, 2012 19:55 UTC (Thu) by RobSeace (subscriber, #4435) [Link]

> Would I be able to run the latest Firefox on Debian woody? I hardly think so.

Not sure about Debian, but I personally run CentOS5 and download the latest Firefox right from Mozilla and run it with very little hassle... (I had to obtain a newer libstdc++ than the system one, and place it in the firefox install directory... Which, granted, an average user isn't going to want to do, but it CAN be done, at least...)

Changes and complaints

Posted Feb 23, 2012 11:34 UTC (Thu) by ms (subscriber, #41272) [Link]

It's a sad day indeed that a site like LWN is forced to write an article essentially reminding people to exercise their ability to empathise and see both sides of the coin. I fear this is, in many ways, really just a reflection of modern society as a whole and its apparently continuing journey to foster intolerance and outrage at the slightest opportunity.

Changes and complaints

Posted Feb 23, 2012 12:09 UTC (Thu) by dskoll (subscriber, #1630) [Link]

OTOH, some developers treat user concerns callously and with arrogance. There's plenty of blame on both sides. Trying to "engage" with developers can be a pretty rough experience. I've been burned too often to try to change people's minds. Instead, I just switch to what I consider to be saner and better software.

So far, thankfully, I haven't had to switch to a proprietary OS because there's always been a sane and free alternative.

Changes and complaints

Posted Feb 23, 2012 12:22 UTC (Thu) by ms (subscriber, #41272) [Link]

I agree - as another comment here suggests, software is seldom the problem: the problem is normally personalities, and my professional experience bears that out enormously. My previous comment doesn't absolve any of the many sides from blame: I'm sure we can all benefit from stepping back from the brink whenever possible.

In many ways I consider myself very fortunate: I long ago switched to using ratpoison - a window manager that has seen almost no development for the last 5 years - and thus my workflow and the way I use my computers has been very stable indeed. I have never looked back from all but abandoning the use of the mouse - especially from a health point of view. In the gnome2 days, I used to run a lot of gnome daemons in the background so that I didn't have to type !?mount whenever I plugged in eg my camera, but that's currently broken since gnome3 and I've not bothered to try and fix it up. But then I'm so far off the expected use case that I don't have any expectation it should work: I'll take responsibility for that and I'm happy to revert back to the command line and do things manually when stuff breaks. And I run Debian Sid on my desktop and have done for the last decade. I kid you not.

Changes and complaints

Posted Feb 23, 2012 20:54 UTC (Thu) by drago01 (subscriber, #50715) [Link]

> OTOH, some developers treat user concerns callously and with arrogance.

No really if you want someone else to change his mind you have to convince him that your proposed solution is better. Just demanding, name-calling and ranting is not the way to do that.

Or he can simply disagree with you which is fine as well. Demanding that others implement software the way you want it is arrogant. Having a different opinion is not.

Changes and complaints

Posted Feb 25, 2012 6:25 UTC (Sat) by speedster1 (guest, #8143) [Link]

>> OTOH, some developers treat user concerns callously and with arrogance.

> No really if you want someone else to change his mind you have to convince him that your proposed solution is better. Just demanding, name-calling and ranting is not the way to do that.

Are you saying Free Software developers *never* treat users callously and with arrogance? I certainly have seen this happen. It is tempting for people who are talented and skillful to be dismissive of others with less talent or skill (in their own area of interest). Reminds me of some conversations that used to go on in my research group at grad school...

Of course, there are people working at proprietary companies who treat their users badly as well, but the same arrogant person has more incentive to treat users respectfully when working at his day job than when coding Free Software on his own time -- it's easier to get fired from a company than a Free Software project, and the former is more likely to be one's main source of income. It's hard to know how to motivate such a person to be respectful when doing volunteer work; in some cases peer pressure might work, depending on leadership of the core project team (doesn't help of course when the arrogant, grumpy devs *are* the core team)

And on the other hand, a user can also be prone to arrogance, thinking the devs should hang on his every word... in which case it's time to "go long" on fire retardent.

Changes and complaints

Posted Feb 23, 2012 11:47 UTC (Thu) by error27 (subscriber, #8346) [Link]

"It often sounds like some believe that the developers of these projects set out to ruin people's lives with their changes. Somehow that doesn't seem likely"

I disagree. It's just like when you tazor someone. It feels good. In the kernel I've seen people who were amused about breaking out of tree drivers. Those guys are not customers so it's funny.

For gnome3, old time gnome users are not the target audience, so its funny when you break alt tab for them. Ha ha. Nothing they can do about it. Although, in fact you can do something about it, which is to release a forked version like the Mint people are doing.

This is hardly the first time when personality issues have been a problem. Look at the history of cdrecord, reiserfs, glibc and emacs. So often it's not about technology.

Changes and complaints

Posted Feb 23, 2012 16:06 UTC (Thu) by tetromino (subscriber, #33846) [Link]

> For gnome3, old time gnome users are not the target audience, so its funny when you break alt tab for them.

IMHO, that statement is mind-bogglingly wrong. The new Alt-Tab behavior is quite possibly the single most useful UI change that Gnome 3 has introduced. I find it incredibly convenient to be able to quickly switch between windows for a specific application with Alt-` without needing to wade through everything else I have open. I find it incredibly convenient to be able to switch from firefox to gnome-terminal with just one press of Alt-Tab without need to wade through all the other firefox windows I have open. And I find it incredibly convenient for the desktop environment to arrange the Alt-Tab and Alt-` lists based on access order so that the documents I switch to frequently can be switched to quickly.

TL;DR: Gnome 3's Alt-Tab behavior is objectively better than what Gnome 2 had. One just needs to spend a bit of time getting used to it.

Changes and complaints

Posted Feb 23, 2012 17:19 UTC (Thu) by error27 (subscriber, #8346) [Link]

It's not "objectively better" until someone does a study.

Anyway, I don't think it's worth arguing about because the Mint people are taking care of it and I can use something else in the mean time.

Changes and complaints

Posted Feb 23, 2012 17:36 UTC (Thu) by dashesy (guest, #74652) [Link]

Maybe it is better, but wasn't it better to use a new combination for a new paradigm (even just to reduce the amount of complaints)? The Windows key is otherwise useless, it could be put to a good use-case, it is closer to the Tab, and most new keyboards support it (assuming that you have a modern graphics card you probably have a new keyboard too).

Changes and complaints

Posted Feb 24, 2012 16:41 UTC (Fri) by jeremiah (subscriber, #1221) [Link]

Bad assumption FYI. I know I'm not the only one out here using an IBM Model-M. Are there are most definitively no "windows-keys" on it. Please feel free to ignore this comment, although true, it's meant mostly in jest.

Changes and complaints

Posted Feb 23, 2012 17:53 UTC (Thu) by cortana (subscriber, #24596) [Link]

gnome-shell's behaviour breaks as soon as you want to switch between different open windows of different programs. For instance, terminal 1 (where I've just finished editing a config file for a server on a remote host) to chromium (to read some documentation for the client I'm about to use) to terminal 2 (where I'm about to try the client against the updated server config). This was easy and convenient in GNOME 2, but in GNOME 3 it's a pain: where I used to be able to do the window switch in one operation, I now have to use *both* alt-tab and alt-` to get to the Window I want. And every time it happens it breaks my train of thought and I get distracted by the user interface. Every time this happens, the UI has failed. I have given up for now and am sticking with GNOME 3 in fallback mode for the foreseeable future.

Changes and complaints

Posted Feb 23, 2012 19:06 UTC (Thu) by tzafrir (subscriber, #11501) [Link]

What would it take to patch the terminal app to spawn each window as a different "application"? (Terminal-1, Terminal-2, etc.).

Changes and complaints

Posted Feb 23, 2012 20:27 UTC (Thu) by cortana (subscriber, #24596) [Link]

gnome-shell decides whether windows are part of the same application by looking at their WM_CLASS. More details can be found here: http://live.gnome.org/GnomeShell/ApplicationBased

In practice, you'd need gnome-terminal to be launched with a different --class argument every time it is invoked. There's also a bug whereby the --class argument is ignored when gnome-terminal is already running (the process that's launched just tells the existing process to create a new window, and it doesn't pass the class through).

But doing this makes gnome-shell unable to know which icon to use for an application, and I think (though I didn't try it for a while) it will just display all the applications as "gnome-terminal" rather than their actual title.

It's a start though. I certainly think there is some scope for programs to declare themselves as "I don't want all my windows to be raised at once and alt-tabbed into as one", perhaps by setting a window property or something. This seems more productive than my grumpy original comment! :)

Changes and complaints

Posted Feb 28, 2012 11:08 UTC (Tue) by daenzer (✭ supporter ✭, #7050) [Link]

> In practice, you'd need gnome-terminal to be launched with a different --class argument every time it is invoked.
> There's also a bug whereby the --class argument is ignored when gnome-terminal is already running (the process
> that's launched just tells the existing process to create a new window, and it doesn't pass the class through).

--disable-factory works around that problem.

> But doing this makes gnome-shell unable to know which icon to use for an application,

I just ran 'gnome-terminal --disable-factory --class foobar', and the dock on the left of the Activities screen had a second terminal icon with a 'foobar' caption. I also tried 'gtk3-demo --class foobar', and again the icon and caption were correct.

This is with version 3.2.y of the relevant components.

Changes and complaints

Posted Feb 28, 2012 12:04 UTC (Tue) by cortana (subscriber, #24596) [Link]

Yeah, I think gnome-shell falls back to using the window's icon. Unfortunately this was (last time I did it, anyway) of a much lower resolution than the icon referred to in the .desktop file.

Try making two terminals with teh 'foobar' class and alt-tab between them. Again, last time I tried, it became unusable--each individual 'foobar' window was treated as belonging to a separate "application".

I'd try all this again, but I'm sick of gnome-shell crashing my VirtualBox VM. Maybe when llvmpipe makes it to end-users it will become more usable and I'll give it another go.

Changes and complaints

Posted Feb 23, 2012 16:07 UTC (Thu) by ebassi (subscriber, #54855) [Link]

I disagree. It's just like when you tazor someone. It feels good.

I'd like to think that not everyone feels good when tasering people, like apparently you do. I know I wouldn't.

For gnome3, old time gnome users are not the target audience

wrong. gnome developers are long time gnome users. a lot of the people that are working on gnome are the same as the people that started the project in the first place.

Changes and complaints

Posted Feb 23, 2012 22:19 UTC (Thu) by jonasj (guest, #44344) [Link]

It's just like when you tazor someone. It feels good.

That was the most uncomfortable comment I have ever seen on LWN. The next time you want to talk about how violence makes you feel good, I really hope you take it to your blog or something instead.

Changes and complaints

Posted Feb 24, 2012 7:16 UTC (Fri) by error27 (subscriber, #8346) [Link]

Sorry, I didn't mean to advocate violence.

It's just that when I look around at the world and see all the cruelty, I don't see people as meaning well. I see people who enjoy it. I see the pepper spray cop telling his buddies, "Ooo... Let me handle this. :)" The Stanford prison experiment showed that meanness is the norm, not the exception, and when I look into my own heart I know this is true.

I apologize for my comment.

Changes and complaints

Posted Mar 1, 2012 18:24 UTC (Thu) by slashdot (guest, #22014) [Link]

Violence/cruelty can indeed be enjoyable, but helping other people and doing a good job can be enjoyable too.

Of course, one expects that the latter sentiments prevail between developers and their users.

Changes and complaints

Posted Feb 29, 2012 14:41 UTC (Wed) by error27 (subscriber, #8346) [Link]

Actually I fixed all my complaints with gnome3 using the extensions. It was simple. I am an idiot and a dick. Sorry.

The value of backward compatibility

Posted Feb 23, 2012 14:22 UTC (Thu) by bkw1a (subscriber, #4101) [Link]

I think developers often underestimate the value of backward (and forward) compatibility and overestimate the value of new features. Users sometimes find that they're locked in to a particular version of a desktop toolkit, for example, because upgrading the toolkit would break existing applications. And without the upgrade, there are newer applications that can't be installed. Compare this with the long-term stability of things like perl, latex and the basic X libraries, which change slowly and with a lot of thought about whether existing applications will be broken.

New features are great, but if I can't take advantage of them without upgrading the entire operating system, they're useless. This may not be such a big deal for individual users, but it's a huge deal for people who support many computers.

The value of backward compatibility

Posted Feb 23, 2012 22:10 UTC (Thu) by drago01 (subscriber, #50715) [Link]

> Users sometimes find that they're locked in to a particular version of a desktop toolkit, for example, because upgrading the toolkit would break existing applications. And without the upgrade, there are newer applications that can't be installed.

Specific examples? For at least gtk (and qt iirc) this is not true. When major changes happen (gtk1->gtk2 or gtk2->gtk3) both toolkits don't conflict with each other so that you can simply install both.

The value of backward compatibility

Posted Feb 23, 2012 22:26 UTC (Thu) by bkw1a (subscriber, #4101) [Link]

It's been a while since I've tried this, so I'll give it another go. One thing I'd like to get running under Centos 6 is "okular", which requires KDE 4.6 (CentOS 6.2 has KDE 4.3.4). When I have a chance, I'll see if I can compile the necessary version of KDE and install it.

The value of backward compatibility

Posted Feb 23, 2012 22:40 UTC (Thu) by dlang (subscriber, #313) [Link]

requires a specific version of KDE or of QT?

The value of backward compatibility

Posted Feb 24, 2012 13:57 UTC (Fri) by bkw1a (subscriber, #4101) [Link]

>> requires a specific version of KDE or of QT?

When configuring okular with cmake, I see the complaint:

ERROR: the installed kdelibs version 4.3.4 is too old, at least version
4.6.0 is required

The value of backward compatibility

Posted Feb 24, 2012 19:31 UTC (Fri) by ABCD (subscriber, #53650) [Link]

Okular is itself part of the kdegraphics component of the KDE Software Compilation (or whatever upstream changed the name to this week). As such, it does tend to depend on the version of kdelibs that was released at the same time. Okular itself is available in Centos 6 as part of the "kdegraphics" package.

The value of backward compatibility

Posted Feb 24, 2012 19:41 UTC (Fri) by dlang (subscriber, #313) [Link]

do you have to upgrade all of KDE to put in a new version of the kdelibs? or do they do a good job of making them backwards compatible so that you can put in a new version of kdelibs without having to upgrade all of KDE?

The value of backward compatibility

Posted Feb 24, 2012 22:38 UTC (Fri) by Jonno (subscriber, #49613) [Link]

You can upgrade only kdelibs if you want to, but if you want to upgrade anything else, you must also upgrade kdelibs. E.g. kde modules are backwards compatible, but not forward compatible.

The value of backward compatibility

Posted Feb 24, 2012 22:51 UTC (Fri) by zlynx (subscriber, #2285) [Link]

Having run Gentoo in the past with both Gnome 2 and KDE 3 I know for a fact that toolkit upgrades break applications.

No, not all the time. But it appears that testing of applications is not well done in the toolkit world. They seem to assume all software will be built against the current toolkit version.

It made it very risky to put the newly released toolkits into my Gentoo override files. The Gentoo distro knew about the problems and made it rather difficult (big lists of packages to put into override files) to get the new Gnome until all the bugs were found and worked through.

It has been rather a long time, but I do recall broken behavior with, I think, Inkscape and the Gnome 2.22 to 2.24 upgrade.

Complaints and changes

Posted Feb 23, 2012 17:52 UTC (Thu) by man_ls (guest, #15091) [Link]

It is not often that I get to disagree so fundamentally with an LWN.net editorial. Not just with details, the idea or even with the tone, but with the intended audience. I think that an op ed is indeed needed, but directed to developers of desktop environments and similar software; one that requests them to listen to their audiences and not make them angry and frustrated, specifically about user interface changes. And asking these developers to correct course if such a situation might arise eventually.

GNOME 3 and Ubuntu Shell have modified their UIs, intentionally, so that some users perceive them as broken: apparently they have been adapted to touch interfaces and small screens, leaving some power users (those with a mouse or a big screen) in the cold. I wonder why these developers have not committed to keep a compatibility mode going forward, as long as it is needed. KDE 4 was released far too soon: LWN commenters say that 4.5 was already usable again, meaning that there were 4 useless point releases. KDE developers are adamant that developer needs had to be put before user needs, and they don't seem likely to correct course soon. As to Wayland, based on recent LWN.net articles 432 comments are something to take into account, and I wonder why its developers have not put a prominent message on the front page stating: We will strive to make Wayland network transparent as soon as possible. In short, we need Free software to say out loud: People, we care about you.

I think that the present article asking for respect and contributions is unwarranted. As developers we must indeed be respectful and focus on technical work. As consumers, however, our most important contribution is to make ourselves heard. And in the examples given it is not easy to contribute as a developer, for different reasons. In GNOME it is not possible to code backwards support into the desktop environment, because it is not where the core developers want it to go; in Ubuntu Shell there is a company which funds development; and in KDE we cannot prevent the release of immature code. There is also a common reason: they are large, complex codebases not easy to delve into, so forks are often not feasible (though their mere existence is revealing). We cannot fool ourselves: we are end users (i.e. consumers) of desktop environments, because we need them to be standard and there is not much we can do to set their directions. Except complain, or complain bitterly.

I have not entered these flames, but that is because I don't care a bit about these projects: since before the current deluge of desktop environments I am a happy XFCE user, and its developers have not disappointed us changing our environment (or saying they would want to do that, for whatever reason). If I had cared about GNOME, KDE, Ubuntu, or even about network transparency in Wayland, I would have tried to let their developers know that I am not happy about it and that the changes suck. If this is interpreted as a flame, then let be it. By now developers of these projects are aware that part of their user base is not happy. For every user that complains there are 9 that leave silently, and for every user that complains bitterly there are probably 99 that have already left. Let us face it: our marketshare is not increasing, and desktop linux is probably decreasing below 1% (cannibalized by Mac OS X users and Windows 7 defectors). Jake says:

If users vote with their feet and move elsewhere, one suspects the projects will follow along behind.
But people who complain do not want to leave the project! They use and probably love the previous versions; it is no wonder they get emotional when they are forced out.

Usability studies are something that you do when you do not have user feedback; it's not a substitute. Once you have user opinions there is no need to do fancy studies, you just have to listen. Users may not know what they want, but they sure know what they don't want. This is the first and foremost lesson that all software projects need to learn, especially in Free software where there is little lock-in.

Complaints and changes

Posted Feb 23, 2012 19:00 UTC (Thu) by khim (subscriber, #9252) [Link]

Let us face it: our marketshare is not increasing, and desktop linux is probably decreasing below 1% (cannibalized by Mac OS X users and Windows 7 defectors).

Actually your link shows Linux's marketshare at 0.62% six years ago and at 0.82% today. I don't think it's precise enough to do any large conclusions.

For every user that complains there are 9 that leave silently, and for every user that complains bitterly there are probably 99 that have already left.

Perhaps, but what does it change? As alankila said: Linux currently is for the 1%, not the 99% - you can afford to lose some of that 1% if you gain nearly any fraction of the 99%: number of users would increase dramatically.

Usability studies are something that you do when you do not have user feedback; it's not a substitute. Once you have user opinions there is no need to do fancy studies, you just have to listen. Users may not know what they want, but they sure know what they don't want.

Right, but irrelevant. They know what change they will complain about. They don't know what changes they will eventually accept. And this is vital metric. This is why Mozilla plans to track users: they are losing to Chrome (again, according to the link you've shown) - and they don't even know why.

As to Wayland, based on recent LWN.net articles 432 comments are something to take into account, and I wonder why its developers have not put a prominent message on the front page stating: We will strive to make Wayland network transparent as soon as possible.

Perhaps because good people are not supposed to lie?

In short, we need Free software to say out loud: People, we care about you.

s/we care/we DON'T care/

I think the people are doing what they are supposed to do but that they themselves are still in denial as to about why they need to do that.

Situation is very simple:

  1. Linux desktop goes nowhere.
  2. Small cosmetic changes don't cut it.
  3. We need to do something drastic if we want to attract new users.
  4. Old users may complain (or complain bitterly).
  5. This is irrelevant because if we'll not attract new user soon (as in: the next 3-5 years) then it's the end.

It's as simple as that. Some people may try to sugarcoat the truth, some are still lying even to themselves, but that's the sad truth.

As consumers, however, our most important contribution is to make ourselves heard.

As consumers you (well, we because I'm a Linux user, too) are irrelevant because there are just not enough of us. Yet Bug #1 is as relevant today as it was half-dozen years ago. The fact that someone is finally taking it seriously is encouraging - even if I'm not all that sure they have chosen the right path to fix it.

Complaints and changes

Posted Feb 23, 2012 20:07 UTC (Thu) by kh (subscriber, #19413) [Link]

I had in the last ~ five years gotten some non-technical users to switch to Ubuntu. It was easy to understand and work with, and they were happy with the reliability and security of the system. They bragged about the ease of installing printers and finding useful free-software programs and web tools. They came to understand and enjoy the desktop switcher. They got some of their friends and family to switch. Linux did not have all the games and a few other programs (e.g. accounting software), but they kept around a Windows partition for those few times they needed it. I kept hearing stories that they found they booted to Windows less and less. They would upgrade the OS themselves. They even came to like Google Docs & OpenOffice/LibreOffice better than the Microsoft product - they were not made to use the ribbon interface that forced them to work how Microsoft's UI experts told them they should. They felt treated with respect.

They have all now switched to Windows 7 as their primary desktop because of the change to Unity. They don't particularly like Windows 7, but feel it is an improvement over Vista, and much less trouble than Unity because Windows 7 gives them more freedom to work as they wish. They can still use Firefox and LibreOffice with Windows.

So when you say that your motivation is to give up on the 1% developer community in search of the 99% of other users, I can completely understand your motivation, but I think you have blinded yourself to the actual results of your actions.

Complaints and changes

Posted Feb 23, 2012 21:22 UTC (Thu) by jonasj (guest, #44344) [Link]

Why are they not just still running Ubuntu's latest LTS with GNOME 2? Did the introduction of Unity in a newer release cause the old, still-supported-for-more-than-one-year-from-today release to be ripped out from their harddrives?

Complaints and changes

Posted Feb 23, 2012 21:34 UTC (Thu) by jmspeex (subscriber, #51639) [Link]

I would assume that the answer is a combination of:
1) Once you upgraded (the newer version has to be better, right), you can't easily downgrade
2) Why re-install, when you've just been burned and you know it's a dead-end because in one year, LTS will stop being supported and you won't be able to use gnome 2 at that point.

Personally I switched to XFCE, but I can see how people who just came in from Windows would go back.

Complaints and changes

Posted Feb 23, 2012 22:12 UTC (Thu) by jonasj (guest, #44344) [Link]

The poster I was replying to specifically talked about Ubuntu, where:

1) Once they upgraded from v10.10 to 11.04 (first release with Unity), they still had GNOME 2 (11.04 stuck with GNOME 2), and could just pick it from the login screen.

2) But the next LTS will still have the gnome-panel package in the archive! Yes, it might be officially called "fallback mode" now, but it is, for all intents and purposes, the exact same experience as GNOME 2 -- or is there something I'm missing here?

Complaints and changes

Posted Mar 1, 2012 3:45 UTC (Thu) by Zizzle (guest, #67739) [Link]

> Yes, it might be officially called "fallback mode" now, but it is, for all intents and purposes, the exact same experience as GNOME 2 -- or is there something I'm missing here?

You have clearly not used GNOME3 fallback mode.

All the panel applets are missing.

The GNOME developers have made it clear that they will stop maintaining it ASAP.

Complaints and changes

Posted Mar 4, 2012 12:02 UTC (Sun) by jonasj (guest, #44344) [Link]

I have used it, don't make claims about things you don't know. (I personally don't need any of the missing applets.)

They have also made it clear that anyone is welcome to continue maintaining it on gnome.org infrastructure.

Complaints and changes

Posted Feb 23, 2012 23:36 UTC (Thu) by dgm (subscriber, #49227) [Link]

Because, linux being what it is, you _have_ to upgrade to get access to new drivers and new applications. People buy new machines, and want to use improved software.

Complaints and changes

Posted Feb 24, 2012 2:22 UTC (Fri) by dskoll (subscriber, #1630) [Link]

I make everyone at work run Linux. One of our sales guys put Ubuntu on all of his home computers (he had got used to Linux and started to like it.)

When the Unity interface came out, it drove him mad and he defected (to Debian, thankfully... not to Windows.)

Our sales guys run Debian Squeeze and GNOME 2. I will move them to XFCE rather than undergoing the pain of transition to GNOME3.

Complaints and changes

Posted Feb 24, 2012 7:40 UTC (Fri) by jonasj (guest, #44344) [Link]

I would like to ask you a question about the "pain of transition to GNOME 3" that you mention. I assume in this context by GNOME 3 you refer to gnome-shell, and not the newer infrastructure (GTK3, etc.).

See, like your users, I'm running GNOME 2 on Squeeze at the moment, but when time comes to upgrade to a newer distro release, I expect I will run the same interface on it as I do now: that is, gnome-panel. The thing that is difficult for me to understand about all the people complaining about GNOME 3 taking away their old and loved interface, is that they are still, for the time being, shipping gnome-panel, with practically identical functionality and behaviour as in GNOME 2. I feel like either there must be some major problem with version 3.x of gnome-panel that I'm just oblivious to, or everyone's complaints just don't make sense. It's driving me crazy trying to grasp why people complain about the GNOME interface change. Can someone who is angry with GNOME 3 tell me, what do you have against the so-called "fallback mode", a.k.a. *the same interface you used to use, just with a different version number*, which is still being shipped in GNOME 3.x releases? Is it really just small things like holding down alt to change panel configuration? Or is there something major I'm just missing?

Complaints and changes

Posted Feb 24, 2012 12:12 UTC (Fri) by dskoll (subscriber, #1630) [Link]

I'm moving my users from GNOME to XFCE for one simple reason: XFCE doesn't introduce radical changes. For my non-technical users, stability of workflow is far more important than whiz-bang new GUI paradigms.

My users could almost certainly cope with GNOME 3 after a period of adjustment, but why should I lose productivity if I don't have to?

Complaints and changes

Posted Feb 28, 2012 0:42 UTC (Tue) by HelloWorld (guest, #56129) [Link]

Oh, so you're actually claiming that moving from Gnome 2 to Gnome 3 (potentially with fallback mode) is a more disruptive change than switching to Xfce?

Complaints and changes

Posted Feb 28, 2012 1:04 UTC (Tue) by neilbrown (subscriber, #359) [Link]

I found this to be true.

Complaints and changes

Posted Feb 28, 2012 1:15 UTC (Tue) by HelloWorld (guest, #56129) [Link]

Considering that fallback mode is pretty much the same as Gnome 2, the only possible conclusion is that you're confused.

Complaints and changes

Posted Feb 28, 2012 2:46 UTC (Tue) by neilbrown (subscriber, #359) [Link]

It is certainly possible that I am confused. It is also possible that I have been swayed by all the negative publicity. But when I upgraded to openSUSE 12.1 and got Gnome 3, some stuff that I valued just stopped working, so I had to make an active decision. "Switch to Xfce" seemed easy, and it turned out that it was.

Complaints and changes

Posted Feb 28, 2012 9:54 UTC (Tue) by paulj (subscriber, #341) [Link]

Fallback mode is GNOME2 only for a limited period of time - soon as software rendering is in place for gnome-shell (which might be the case already, shipping in Fedora 17?). So the GNOME2 -> (GNOME3 | XFCE) disruptive change will soon be unavoidable.

Complaints and changes

Posted Mar 6, 2012 8:30 UTC (Tue) by jonasj (guest, #44344) [Link]

GNOME developers have stated publicly that anyone who wants to are welcome to continue maintaining it on gnome.org infrastructure. So just because GNOME 3 won't depend on it any longer once software rendering is in place, it will still be available.

Complaints and changes

Posted Mar 1, 2012 7:58 UTC (Thu) by cmm (guest, #81305) [Link]

> Considering that fallback mode is pretty much the same as Gnome 2

False, it's not even close.

Complaints and changes

Posted Feb 23, 2012 22:04 UTC (Thu) by daglwn (guest, #65432) [Link]

>> I wonder why its developers have not put a prominent message on the front
>> page stating: We will strive to make Wayland network transparent as soon
>> as possible.

> Perhaps because good people are not supposed to lie?

>> In short, we need Free software to say out loud: People, we care about
>> you.

> s/we care/we DON'T care/

And with one post khim demonstrates how completely misguided this editorial is.

Complaints and changes

Posted Feb 23, 2012 22:23 UTC (Thu) by dlang (subscriber, #313) [Link]

> Situation is very simple:
>
> Linux desktop goes nowhere.
> Small cosmetic changes don't cut it.
> We need to do something drastic if we want to attract new users.
> Old users may complain (or complain bitterly).
> This is irrelevant because if we'll not attract new user soon (as in: the next 3-5 years) then it's the end.

so stop pretending that what you are creating is an upgrade of existing packages.

Go off and create new ones, use a new name, don't claim (by implication) that what you are doing is just a new version of the existing software.

This is something that Canonical did mostly right. They created a new name (Unity) for their new desktop, and left the option in place to use the old desktop (GNOME), along with the option to use other desktops (KDE, etc). Unfortunantly, their decision to do this coincided with the GNOME madness which is a classic case of how not to do this (make the new thing have the same name as the old one and be incompatible with it)

The KDE 3->4 transition was ugly, but for a very different reason (premature transition), and while there are a few cases of "there's no maintainer for that so it's not getting reimplemented", for the most part, their response to complaints has been either "you can disable that new feature if you don't like it" (frequently with a "here's how"), or "we're planning to do that but haven't got to it yet". This seems to show a very different attitude.

Complaints and changes

Posted Feb 24, 2012 9:55 UTC (Fri) by man_ls (guest, #15091) [Link]

A brand new package instead of an upgrade would be an improvement for two of the cases described above: GNOME and KDE. I understand that the temptation to experiment with your current userbase must be overwhelming. My answer would be: fine, but as long as you are ready to roll it back if your users don't like it.

Ubuntu Unity (not Ubuntu Shell, sorry!) and Wayland are however new packages, and the situation is still not good. Wayland is a good example: even the idea of change in a pre-release package can scare people off. Something which should in principle be seen as a welcome renovation does instead create a backlash; and given the precedents we cannot blame those would-be users.

Complaints and changes

Posted Feb 24, 2012 0:16 UTC (Fri) by dgm (subscriber, #49227) [Link]

> Perhaps, but what does it change? As alankila said: Linux currently is for the 1%, not the 99% - you can afford to lose some of that 1% if you gain nearly any fraction of the 99%: number of users would increase dramatically.

A false dichotomy. You don't have to upset your current users in order to gain new ones. in fact, it's doing the opposite that Microsoft, Apple and now Android have grown. Network effects are your best friend. Or would have been, should I say.

In two months it's going to be a year from the release of Gnome 3. What fraction of that 99% will have Gnome 3 won in that time?

Complaints and changes

Posted Mar 1, 2012 3:51 UTC (Thu) by Zizzle (guest, #67739) [Link]

> In two months it's going to be a year from the release of Gnome 3. What fraction of that 99% will have Gnome 3 won in that time?

Excellent point.

Is Fedora the only distro that defaults to it?

So GNOME went from having nearly all the distros use it by default (GNOME2) to one?

I suspect that GNOME3 not only has no significant gain of the 99%, but has lost most of the 1%.

Complaints and changes

Posted Feb 24, 2012 2:24 UTC (Fri) by dskoll (subscriber, #1630) [Link]

s/we care/we DON'T care/

Thank you for clarifying that. Now we see why the editorial is completely wrong.

When developers are that arrogant, they don't deserve users. I suppose that will, in the end, reduce the volume of complaints.

Complaints and changes

Posted Feb 24, 2012 2:36 UTC (Fri) by viro (subscriber, #7872) [Link]

> When developers are that arrogant, they don't deserve users.

Non sequitur - it's not developers, it's khim. If that demagogue has ever developed anything other than verbal diarrhea and, perhaps, political connections, I'd be very surprised.

Complaints and changes

Posted Mar 1, 2012 3:52 UTC (Thu) by Zizzle (guest, #67739) [Link]

Yeah and ad hominem is so much better than non sequitur.

Complaints and changes

Posted Feb 23, 2012 19:48 UTC (Thu) by mpr22 (subscriber, #60784) [Link]

I wonder why its developers have not put a prominent message on the front page stating: We will strive to make Wayland network transparent as soon as possible.

First, is there really anyone left who thinks "as soon as possible" means anything other than exactly what the speaker wants it to mean?

Second, would anyone believe them if they did say that?

Complaints and changes

Posted Feb 23, 2012 22:33 UTC (Thu) by jmorris42 (guest, #2203) [Link]

> Second, would anyone believe them if they did say that?

I certainly wouldn't. I suspect they wouldn't merge a patch set that implemented network transparency if it were delivered to them written, debugged and ready to roll. They reject the idea.

Because everything I have read about Wayland says it comes from the same "UNIX Hater's Handbook" school as exemplified by Lennart Pottering and rest of this 'alien tech' that has been showing up in our distributions the last couple of years. They reject every assumption the [U|LI]NIX/GNU/X stack was built up from. Everything is a file? Bah! Windows rules the world and it is built on Everything is an API so there. Network transparency? Useless! That may have been ok in the Elder World but in a world of tablets and web apps? Small interchangable parts? No place in the modern world for that! Knowable systems? For who? We are building for people who would have 12:00 blinking on their VCR... if they still made em.

I really wish they would give up trying to remake our world and go try to make a better ReactOS since Windows seems more to their liking.

In the end we just might want to ponder the ramifications of giving commit access to to the key cultural artificts of the UNIX world to people who hate everything we believe in. Way I see it is we UNIX folk are in a bad spot, we got just popular enough a whole bunch (by our standard, a pittance of total Windows users though) of Windows refugees came into our camp and have basically taken over, leaving us strangers in what we thought was our own land. Just listen to em talk, they don't want to learn the UNIX Way and they don't want to attrack more of their tribe to come to us. They want to remake the current landscape into one more attractive to their old tribe and loudly and derisively dismiss us 'old greybeards' who don't like their plans.

Complaints and changes

Posted Feb 23, 2012 23:12 UTC (Thu) by daglwn (guest, #65432) [Link]

To be fair, change in necessary and often good. I don't particularly have a problem with most of what Lennart has done, though binary logs are concerning.

And there is a fair amount of cognitive dissonance in the "traditional" UNIX world. "Everything is a file," except when it isn't (sockets). "Everything is a file," yet Linus bashes a language (C++) that directly supports modeling this kind of inheritance concept efficiently. We've seen disastrous consequences when filesystems don't get updated to some new VFS layer feature.

None of the philosophy from any side can be applied ideologically everywhere. I like the idea of Wayland. The X codebase is a mess and it really does have a lot of unused stuff. But like you, I have no confidence that there will ever be any acceptable network transparency support. It's a critical feature without which I can never adopt Wayland.

Complaints and changes

Posted Feb 24, 2012 22:59 UTC (Fri) by zlynx (subscriber, #2285) [Link]

I like one Wayland idea someone had. He pointed out that with the latest Intel chips you can do a realtime video encode of each composited window and ship that over the network using less bandwidth than most of the current VNC solutions.

Complaints and changes

Posted Mar 1, 2012 13:50 UTC (Thu) by nix (subscriber, #2304) [Link]

Yeah... but video encoding is lossy and intended for animated real-world scenes. Have you ever seen what a window containing lots of text (or even line drawings) looks like after a video encoding pass? Ugly and unreadable, that's what.

(This is not an indictment of video codecs: they are largely very good at what they do. But what they are designed to do is encode animated real-world scenes, not flat window images.)

Complaints and changes

Posted Mar 1, 2012 22:13 UTC (Thu) by zlynx (subscriber, #2285) [Link]

If you did not move the text around, I believe that after several frames it would be clear and readable as each frame increased the resolution.

I have watched shows that had quite a bit of text in them. Some sports shows have a bunch of statistics and on HDTV they are pretty crisp.

Complaints and changes

Posted Feb 23, 2012 23:19 UTC (Thu) by dlang (subscriber, #313) [Link]

while I am unhappy about many of the same changes you are, I don't think that problem is that these people are writing the software, it's that people are so locked into their disto that they are not willing to avoid these new things.

systemd is not in every distro, sound was already an API (ALSA) so that's not a change of type, etc

Ubuntu became huge fast because they provided a better experience for people who just wanted to use their desktop machines (including many developers), if they aren't providing what you want (and this includes with the kubuntu/xubuntu/etc options), switch to a different distro

If Fedora is going in the wrong direction in it's attempt to be the bleeding edge of everything, go to a different distro that's going a different direction.

For the most part, the introduction of these new pieces does not prevent other approaches from working.

In places where changes to a new thing _will_ prevent old things from working

for example, if people start pushing a new GUI toolkit that uses Wayland and does not use X, or only uses it like the GNOME failback mode, then it's time to speak up loudly, not because people are trying something new, but because they are doing so in a way that makes it impossible to just ignore that project, you also have to then ignore every other project that makes use of that one project

Or with journald where they appear to be inserting themselves between the application and syslog in a way that will prevent syslog from working (either the trusted properties can be forged by systemd, in which case they can be forged by the application as well, or they can't be forged, in which case systemd cannot just forward the messages to syslog in a way that makes it possible for syslog to not know or care that systemd exists). This is mitigated a little bit by the fact that (at least for now) systemd is still an optional component and people can opt to use something different (like Ubuntu uses upstart)

Complaints and changes

Posted Feb 23, 2012 23:37 UTC (Thu) by khim (subscriber, #9252) [Link]

Because everything I have read about Wayland says it comes from the same "UNIX Hater's Handbook" school as exemplified by Lennart Pottering and rest of this 'alien tech' that has been showing up in our distributions the last couple of years. They reject every assumption the [U|LI]NIX/GNU/X stack was built up from.

Reject? Hardly. They question them. The world changed in the fourty years - yet a lot of stuff in Desktop Linux is the same not because it makes any sense but because "it was always done this way".

Everything is a file? Bah! Windows rules the world and it is built on Everything is an API so there.

Well, ioctl(2) is not a new invention thus it's pretty clear that "Everything is a file" was never a reality, more like some kind of impossible ideal. And this RFC is over twenty years old so you can not say that move to APIs is all that new.

Network transparency? Useless!

Not exactly. They observed the very simple fact: network transparency is "boon for a few, but PITA for many". Such feature makes sense but it should not affect the life of majority.

Small interchangable parts? No place in the modern world for that!

Indeed. What exactly small interchangable parts approach brings to the table? The ability to tinker? You have source code for that. The ability to save few megs of space? Is it still important in a world where memory cost is about $10/GB for RAM and less then $1/GB for flash? Security? This is kind of valid… but it's not clear what system is more secure: the one which includes bazillion tiny pieces with huge number of interconnects or the one where there are less pieces but less joints, too.

I really wish they would give up trying to remake our world and go try to make a better ReactOS since Windows seems more to their liking.

I don't see where you've gotten the ridiculous idea that they actually prefer Windows. They are new generation that does not worship "UNIX way", true, but why do you think anyone who don't worship "UNIX way" should immediately worship Windows? They borrow worthy ideas from MacOS and Windows, true, but to say that systemd is just clone of launchd... nope, not even close.

In the end we just might want to ponder the ramifications of giving commit access to to the key cultural artificts of the UNIX world to people who hate everything we believe in.

Sure. But it's the only viable alternative. New generation always looks critically in it's predecessors thus this revisionism was, in fact, inevitable. Sure, some good properties of "old ways" will be lost… and then reinstalled back. This is how it happens in other industries, too.

Way I see it is we UNIX folk are in a bad spot, we got just popular enough a whole bunch (by our standard, a pittance of total Windows users though) of Windows refugees came into our camp and have basically taken over, leaving us strangers in what we thought was our own land.

Again: where have you gotten the ridiculous idea that this revolution has anything to do with Windows? Sure, Windows, MacOS and other, less popular, OSes serve as source of inspiration (and Windows is the most popular OS) but AFAICS this is mostly "new generation" natural scepsis WRT "old, tried and true" ways of doing things.

Complaints and changes

Posted Feb 23, 2012 23:49 UTC (Thu) by dlang (subscriber, #313) [Link]

> Indeed. What exactly small interchangable parts approach brings to the table?

what Interchangeable parts approach brings to the table is the ability for you to write a new tool that you think will do the job better and try it out without having to replace everything.

It's the very thing that makes these new things possible.

They want to take advantage of the fact that the parts are interchangeable to change some of them, but then deny that capability to the people who come later, because they know what's best for people, both now and for the future.

If they would accept the fact that their solution is not the "One and only true Solution", which means not only accepting that it's not right for everyone now, but that it may not be right for everyone in the future either, they would make it possible to replace their tool rather than make it try and do everything.

Complaints and changes

Posted Feb 24, 2012 0:24 UTC (Fri) by khim (subscriber, #9252) [Link]

what Interchangeable parts approach brings to the table is the ability for you to write a new tool that you think will do the job better and try it out without having to replace everything.

If you want to experiment there are always source code as I've said. Systemd, for example, is actually implemented as bunch of daemons, not as a single monolithic tool. But. These daemons are designed to be used together. If you want to replace something then it's your responsibility to keep everything working smoothly.

Somehow the fact that Linux kernel is developed as a single large monolithic kernel (and not as a HIRD of Unix-replacing daemons) have not stopped all the progress, why should similar development in userspace lead to a disaster?

If they would accept the fact that their solution is not the "One and only true Solution", which means not only accepting that it's not right for everyone now, but that it may not be right for everyone in the future either, they would make it possible to replace their tool rather than make it try and do everything.

Well, sure. But their tools are replaceable. You only need to convince distributors to replace them, that's all.

Complaints and changes

Posted Feb 24, 2012 2:29 UTC (Fri) by dskoll (subscriber, #1630) [Link]

What exactly small interchangable parts approach brings to the table?

A usable system.

I like to compose email in a real editor, not a Gtk or Qt text widget. Kmail has support for an external editor. Thunderbird has an add-on for that. Claws-mail supports it. Evolution does not, except possibly for some weird embedded vim hack (I use emacs.)

I asked the evolution developers to consider adding support for an external editor, and received khim's sed command:

s/care about users/DON'T care about users/

Meh. Why bother?

Complaints and changes

Posted Feb 24, 2012 7:41 UTC (Fri) by drago01 (subscriber, #50715) [Link]

> I asked the evolution developers to consider adding support for an external editor, and received khim's sed command:

> s/care about users/DON'T care about users/

So your definition of "care about users" is "you command => developers implement" ?

Sorry that not how it works.

Complaints and changes

Posted Feb 24, 2012 12:06 UTC (Fri) by dskoll (subscriber, #1630) [Link]

You completely misunderstand my point or (more likely) deliberately twist what I say. Are you a GNOME developer, by any chance? Because this is the pattern of behavior I've observed in my brief interaction with them. Any suggestion, no matter how sensible, is treated with hostility right from the outset.

So as I said: Why bother? It's easier to take the path of least resistance and de-gnomify my home and workplace.

Complaints and changes

Posted Feb 24, 2012 12:26 UTC (Fri) by drago01 (subscriber, #50715) [Link]

> You completely misunderstand my point or (more likely) deliberately twist what I say.

No I did not (intentionally) twist anything. You just said that you asked the evolution developers to implement a feature (allow using external editors) and they have rejected that. That very fact does not imply any hostility. If anything I can conclude that they disagreed with your reasoning.

> Any suggestion, no matter how sensible, is treated with hostility right from the outset.

You seem to interpret it that way but that is not necessarily true; perceiving any disagreements or different opinions as hostility towards you just ends in emotionally heated debates with no useful outcome.

Complaints and changes

Posted Feb 24, 2012 16:52 UTC (Fri) by dskoll (subscriber, #1630) [Link]

You just said that you asked the evolution developers to implement a feature (allow using external editors) and they have rejected that.

It was the way they responded. They rejected it with hostility and disdain, which was not necessary.

You seem to interpret it that way but that is not necessarily true; perceiving any disagreements or different opinions as hostility towards you just ends in emotionally heated debates with no useful outcome.

I have dealt with many developers over the years and can assure you that I know the difference between hostility and simple disagreement.

Complaints and changes

Posted Feb 24, 2012 23:03 UTC (Fri) by zlynx (subscriber, #2285) [Link]

Heh. I bet you got hostility and disdain because they get *so many* complaints about Evolution's editor. I hate the thing too.

Complaints and changes

Posted Feb 28, 2012 0:56 UTC (Tue) by HelloWorld (guest, #56129) [Link]

> I certainly wouldn't. I suspect they wouldn't merge a patch set that implemented network transparency if it were delivered to them written, debugged and ready to roll.
They probably would, because network transparency at that level *doesn't make any sense*. Network transparency needs to be done at the rendering level, but Wayland has nothing to do with rendering as it only handles compositing.

> They reject the idea.
This is simply nonsense. Whenever the Wayland developers were asked about network transparency, they said the same things:
- it's orthogonal to what Wayland does
- it can be done on top of Wayland
Apparently, people like you just refuse (or are unable to?) to understand this and prefer to uninformedly bitch around instead.

Complaints and changes

Posted Mar 1, 2012 13:32 UTC (Thu) by farnz (subscriber, #17727) [Link]

And note that there's a second way to look at Wayland, which makes it a net gain for people for whom remoting is important.

Wayland is equivalent to just the bits of X11 that you get if you use DRI2 and XI2 for rendering and input (respectively), down to the network transparency capability. One way to depend on DRI2 is to require OpenGL 2.0 or later to render - OpenGL 2.0 is simply not available in an indirect context.

At least if a project decides that it requires Wayland and won't run on pure X11, you know that it won't work in a network transparent fashion. It's entirely possible, however (I've done it, quite deliberately, for in-house applications where network transparency is a non-issue) to write X11/OpenGL applications that are not network transparent.

Would the Wayland critics be happier if we simply continued down the (already established) path of extending X11 in ways that are not network transparent? It would still be X11, but if you use the new functionality (and toolkits probably will), you simply don't work over a network. At least with Wayland, you know that it's not X11, and your X11 expectations will not be met.

Complaints and changes

Posted Feb 24, 2012 10:12 UTC (Fri) by man_ls (guest, #15091) [Link]

First, is there really anyone left who thinks "as soon as possible" means anything other than exactly what the speaker wants it to mean?
But that is exactly my point! You don't have to compromise your developer schedule, you don't have to bend over backwards for your users, you don't have to change your architecture to accept every silly request that comes your way. Just [sitcom voice] show you care! [/sitcom]. Be reasonable about the features you want to implement, but also try to be flexible if your users are requesting it.

Being Free software is no excuse to disregard your users; treat them as if you were a business, be professional about it, and users will come. It will be hard to get good referrals for new users if your existing userbase is leaving in disgust.

[offtopic] Sigh, I am a starving hacker right now, but I miss my email notifications. And the ability to filter a few trolls that populate these threads. I will have to shorten my subscription time by going professional again! [/offtopic]

Complaints and changes

Posted Feb 28, 2012 0:39 UTC (Tue) by HelloWorld (guest, #56129) [Link]

> As to Wayland, based on recent LWN.net articles 432 comments are something to take into account, and I wonder why its developers have not put a prominent message on the front page stating: We will strive to make Wayland network transparent as soon as possible. In short, we need Free software to say out loud: People, we care about you.
They didn't put that on the Wayland front page because it would be nonsense. What needs to be done over the network is rendering, but Wayland only handles compositing. Thus, the Wayland people did the only sensible thing when they wrote this in their FAQ:
> Is Wayland network transparent / does it support remote rendering?
> No, that is outside the scope of Wayland. [...] This doesn't mean that remote rendering won't be possible with Wayland, it just means that you will have to put a remote rendering server on top of Wayland.

> Usability studies are something that you do when you do not have user feedback; it's not a substitute. Once you have user opinions there is no need to do fancy studies, you just have to listen.
I'm sorry, but this is just complete nonsense. The truth is that when you listen to "user opinions", you're listening to opinions from a very specific group of people. When you announce a change, who's going to say anything about it? Those who don't use your software? No, because they'll likely judge your project by your current offering and not by something that was proposed somewhere on some mailing list and can't be tried out because it doesn't exist yet. Those who use your software and like the proposed change? No, they're likely to just sit there and wait for good things to happen. But experience shows that people do complain when they feel something is being taken away from them, and this is what I've seen happen over and over again, here and elsewhere.
Engineering psychology is a science. If you don't approach it in a scientific way (and "listen to your users" is anything but), nothing good is going to come out of it. When you have a 1% market share, then listening to your users is a stupid thing to do, you should listen to the 99% who don't use your product and find out why.

It's a matter of trust

Posted Feb 23, 2012 21:27 UTC (Thu) by jmspeex (subscriber, #51639) [Link]

The way I see this, it's not that much a matter of how happy you are with the changes, but a matter of trust in the developers. I don't know about the code itself, but from a user's perspective, gnome 3 is a complete rewrite compared to gnome 2. I would say even KDE 3 is closer to gnome 2. So in practice, despite sharing the same name, this is *exactly* the same as if gnome developers had just decided to say "we've canned the project and we've decided to start something else". That they want to start a new project is fine, but abandoning all users of the current project is not so much. The same pretty much happened between gnome 1 and gnome 2. At the time, I switched to KDE and went back to gnome 2 later. Doing this again for gnome 3 basically now tell me "don't ever use this again because you *will* get burned". Even if gnome 3 ends up evolving in a way that I like (or say I eventually adapt to it), I now *know* that I'll get burned when they decide to again drop support and start with something new that will be called gnome 4 and will not look even remotely like gnome 3. I just cannot trust this project anymore. Of course, at the same time these are mostly volunteers and I paid them nothing, so they're free to develop what they can. I just don't want to have anything to do with this anymore.

Now, when it comes to comparing with proprietary software, I would say it's the complete opposite. Sure there's no guarantee that the current version will keep being supported, but in practice, when new versions come out, the general problem is almost never that there's too many changes. Most of the times, there's actually too *few* changes and they're just forcing upgrades to make more money. I would actually think that any proprietary software company (other than the huge ones) that tried doing the same thing as gnome did between version 2 and 3 would actually lose so many of its customers that it would likely go out of business unless it kept supporting the old version for at least several years. I mean in general, I can't think of many (if any) complete rewrites that really benefited users in the long term. CADT model all over again.

Yeah, right

Posted Feb 23, 2012 22:53 UTC (Thu) by southey (subscriber, #9466) [Link]

I have to say this is perhaps one of the worst unbalanced articles that I have read on LWN over the many years that I have read it! Please don't repeat this again because you are far better than this!

Most free software projects discuss planned changes well in advance of their implementation and give users lots of opportunities to try out early versions.
Sure there is often vague reports of bugs fixed in say nightly builds but that after the fact. So where does one exactly find [m]ost free software project that discuss these changes, let alone with their community?
This especially means more than a one liner that say a limited set of things that may get in. On LWN.net, the only clue we have are the wonderful reports from various summits but actually very few of these actually list planned changes. The Weekly Edition does sometimes cover what people have already proposed for the Linux kernel but there is never a guarantee that feature or patch will make in. Yet the Linux kernel is very good at ensuring that things just work so an upgrade is an upgrade.

Second, did you actually read what you wrote?
Sure user get lots of opportunities to try out early versions but nowhere do you indicate that users can actually get things changed in that process. So it relies on some sucker to get hit first. So, yes, users have a right to complain loudly!

Frankly, everyone has to remember that a project is a huge community where everyone has a major role in the the success of the community.

Yeah, right

Posted Feb 24, 2012 8:55 UTC (Fri) by drago01 (subscriber, #50715) [Link]

> I have to say this is perhaps one of the worst unbalanced articles that I have read on LWN over the many years that I have read it! Please don't repeat this again because you are far better than this!

I tend to disagree. It was about time for such an article is published, given the recent flamewars that come up for every bigger change.

> So where does one exactly find [m]ost free software project that discuss these changes, let alone with their community?

Depends on the project but mailing lists, IRC channels, wiki pages ...

> Sure user get lots of opportunities to try out early versions but nowhere do you indicate that users can actually get things changed in that process.

Sure they can file bugs and / or provide _constructive_ feedback on mailing lists. That means no emotionally loaded post full of insults but a simple description of what you are doing why you think something is broken along with suggestions on how it can be solved if you have any.

> So, yes, users have a right to complain loudly!

Sure they can but that is a waste of everyone's (including the users) time. Getting involved and / or provide constructive feedback / bug reports is way more useful.

Yeah, right

Posted Feb 24, 2012 12:10 UTC (Fri) by dskoll (subscriber, #1630) [Link]

Getting involved and / or provide constructive feedback / bug reports is way more useful.

Not in my experience. In my interactions with GNOME developers, my suggestions and bug-reports have been greeted with anything ranging from hostility to derision.

So how is that supposed to be "useful"?

In some cases, the politics and personality of a software project are such that it's useless to offer feedback or bug reports; it's better just to stop using the software and move on.

Yeah, right

Posted Feb 24, 2012 12:19 UTC (Fri) by drago01 (subscriber, #50715) [Link]

> Not in my experience. In my interactions with GNOME developers, my suggestions and bug-reports have been greeted with anything ranging from hostility to derision.

Can you point me to specific examples? ... at least for gnome-shell (i.e the project where I am involved in) I can't recall a bug report that has been handled that way.

Changes and complaints

Posted Feb 24, 2012 8:49 UTC (Fri) by gnu_andrew (subscriber, #49515) [Link]

I think a lot of the problem with big changes like GNOME 2 to 3 or KDE 3 to 4 is that users just don't see why things have to change so dramatically, to the point where features that were there before and working are no longer there because they haven't been migrated to the new version. Most users are generally appreciative when they can understand the work being done. When they see vast amounts of time being invested in a complete redesign which changes the way they work, rather than their own bugs being fixed, it's understandable they get annoyed. Part of that frustration is that they don't have a choice buto switch in reality; even those with the skills (which will decrease as a percentage over time) can't replace a dedicated team overnight. This need for complete overhaul reminds me of the proprietary approach where new versions need to attract sales. We don't need to do this. Why not concentrate on instead making the existing releases as bug-free and usable as possible without starting from scratch again?

Changes and complaints

Posted Feb 24, 2012 8:58 UTC (Fri) by dlang (subscriber, #313) [Link]

to be fair, the KDE change was only partly a cosmetic/usage redesign, there was a significant change to the underlying library being used (moving to a new version of QT)

that didn't require that they do all of the changes that they did, but they did have to re-write a vary large portion of the code for the new library version, making the other changes while they were already substantially re-writing things makes a lot of sense.

Changes and complaints

Posted Feb 25, 2012 2:10 UTC (Sat) by gmaxwell (guest, #30048) [Link]

"there seems to be an increasing sense of entitlement from some within our communities"

I don't believe this is really true and I'm disappointed to see this rather unnuanced view promoted on the page of LWN.

The people I've seen frustrated by unwelcome changes in Gnome are perfectly happy to let the gnome developers do whatever they want— they aren't suffering the from an illusion of believing it's all being written for them.

The problem is that we are effectively forced by the surrounding ecosystem to use Gnome's latest and greatest, or suffer extreme interoperability and maintenance costs from using a less common unsupported configuration. They've gone very deep on OS integration without heed to the harm it causes people who want to opt out of their decisions.

I've contributed greatly of my time to the common free software ecosystem and I don't think it's right to call it entitlement when I complain someone's latest contributions make that common ecosystem more costly or use.

The real targets of the complaint shouldn't be Gnome however, it should be the distributors— but the commercial distributors at least have the good practice of hiring the people that work on their software, so it seems there is often no neutral party to appeal to. (And even in the community distributions, the commercial funded desktop developers have a significant upper hand in setting the agenda).

It's a complicated and interdependent situation— and I agree that just ranting won't fix it. But dismissing the complaints of people who feel harmed by some of these changes is also not productive.

Changes and complaints

Posted Feb 25, 2012 14:25 UTC (Sat) by bloopletech (guest, #71203) [Link]

It is a shame that the article author went to such effort to rite a reasoned and eloquent piece, as the point has nevertheless been completely lost on the commentators.

The fact that some people here disagree so much with a post by a well respected editor implies that perhaps those people are wrong.

In a democracy, you have a *moral right* to speak your opinion, and in the large, to have it followed by your representatives. This is because the representatives have accepted the responsibility of acting on behalf of others, and because the representatives are receiving *your* money or resources.

In a meritocracy, you may or may not have a *moral right* to speak your opinion and to have it followed; it depends on how much you are contributing. The more you contribute, the more right you have to have your opinion heeded.

** FOSS is a meritocracy, not a democracy. **

The question is then: what have you contributed?

If a large portion of the people using a project disagree with the most active contributors, perhaps the active contributors should listen; however whether we have a "large portion" here is in question. Also, the opinion of the "large portion" has to be weighed against the cost of following the opinion. (The people doing the weighing should be the active contributors).

A lot of people are apparently suffering Unwarranted Entitlement Syndrome. UES is a terrible illness where people are convinced that FOSS projects must do what they say, even though they have contributed nothing, or very little, to the project.

Really, we shouldn't insult sufferers of UES; they are clearly operating under a delusion. But we shouldn't facilitate their delusions by treating them as if they are realistic.

Changes and complaints

Posted Feb 25, 2012 14:28 UTC (Sat) by bloopletech (guest, #71203) [Link]

Golly, an attack on people's understanding and I got a word wrong in the first sentence! That should be write, not rite.

Changes and complaints

Posted Feb 25, 2012 17:02 UTC (Sat) by man_ls (guest, #15091) [Link]

You seem to forget that end-users are contributing something: they are taking time out of their lives to comment, complain and flame. Feedback is about the most valuable contribution a mere consumer can give a project: it serves to save on focus groups, usability studies and the like. Any commercial enterprise can understand that.

Changes and complaints

Posted Mar 1, 2012 3:59 UTC (Thu) by Zizzle (guest, #67739) [Link]

Haha.

Jon Corbet (chief editor here) was seen saying this about GNOME3 the other day on Google+:

"It is really nice to concede (in an after-publication addition) that "If you want to be able to view more than one window at once you will still be able to do so," but it makes it fairly clear that, whoever those users are that they have in mind, they work very differently than I do. I'm discouraged."

Changes and complaints

Posted Mar 1, 2012 3:12 UTC (Thu) by Zizzle (guest, #67739) [Link]

Yeah, bugger the users, what do they know?

How dare they invest in learning to use our software and expect to keep using it the same way!

I think we should make GNOME3 so that it can not be installed in parallel with GNOME2 just to show them!

Full screen windows for everyone! No-one uses the networking features of X anyway!

Changes and complaints

Posted Mar 1, 2012 10:56 UTC (Thu) by ovitters (subscriber, #27950) [Link]

Dude, troll somewhere else.

Changes and complaints

Posted Mar 2, 2012 5:10 UTC (Fri) by Zizzle (guest, #67739) [Link]

Wow you really showed my with your well reasoned argument.

Do you dispute any of these?

a) GNOME2 can NOT be installed side-by-side with GNOME3? (i.e. gnome decided to instantly dump the previous version despite no feature parity)

b) The workflow in GNOME3 is the same as GNOME2? (i.e. no respect shown for current users habbits/expectations)

c) GNOME devs are talking about full screen windows as the default?

https://afaikblog.wordpress.com/2012/02/10/a-new-approach...

d) The next display protocol (Wayland) has no plans for a remote protocol?

e) Given (d) and (b) remote GNOME looks like it will die a quick death?

Seriously, like others have said, I would not have cared about GNOME3 if I could still get GNOME2 in a distro with a recent browser. Who is to blame more, the distros or GNOME? I think the fact that GNOME2 cannot be installed side by side with GNOME3 puts more of the blame on GNOME.

Changes and complaints

Posted Mar 1, 2012 13:05 UTC (Thu) by slashdot (guest, #22014) [Link]

The problem is that changes must NEVER remove existing features and workflows!

If you think you can make a better UI by redesigning it from scratch, awesome, but you MUST do one of these:
1. Make the new UI a strict superset of the old, so that it can be used like the old if desired (and it's possible to only enable some of the changes)
2. Keep maintaining the old one forever, or at least until it's clear that it's obsolete (note that, instead, the new one might turn out to suck)
3. Make it clear that your desktop is just a research toy, and strongly discourage distributions from making it the default

Simply, the desktops must treat the UI just like the Linux kernel treats the user<->kernel API, which is never broken, except for obscure parts that are known to have no users.

Breaking existing features and workflows is NEVER ACCEPTABLE!

Anyone who disagrees with that should not be given any power over design decisions, and he should not be given the power to commit code without review.

Leadership is needed

Posted Mar 3, 2012 22:31 UTC (Sat) by blujay (guest, #39961) [Link]

Finally, someone else who gets it. I agree with you completely.

It seems to me that the problem is that the inmates are running the asylum. Hey, "crazy" people can come up with some good ideas--but they shouldn't also be the ones deciding when to replace the old with the new.

What is lacking in the FOSS [GUI] community is strong leadership. It's too much CADT right now.

Of course, volunteers can't be ordered about. What's needed is for the volunteers to submit voluntarily to some wise leadership and oversight. It reminds me of Debian: the DDs are in charge of their packages--but the ftpmasters are in charge of what makes it into the archive and gets released. If an upstream or a packager makes a really bad decision, that's fine--but the ftpmasters can prevent it from being released in Debian.

In the same way, GNOME and KDE, et al, need some cooler, wiser, visionary heads to oversee what gets released. Some devs want to redesign the GUI, or rewrite some core functionality from scratch? Fine--but those don't get released until they're stable and have feature-parity with existing software. Let them dogfood (the verb) until it's as good as they can make it, and let seriously interested testers use a development PPA to test it--but the rest of the world shouldn't have to put up with the garbage until it's not garbage anymore.

Changes and complaints

Posted Mar 6, 2012 22:47 UTC (Tue) by CycoJ (guest, #70454) [Link]

The big problem I have with a lot of the people complaining is the dishonesty I see in the complains.
The primary example are the complains about wayland. It has been repeated and repeated again that wayland devs supposedly refuse to implement network transparency. This has been refuted multiple times (see e.g. some comments above), but this argument is still being brought up. So the only the only conclusion I can come to is that people making this argument are being deliberately dishonest to advance their point.
Similarly every time critics are being told that they can continue using X11, the argument is made that toolkits might possibly remove X11 support at some point in the future. That argument is also somewhat dishonest, the likeliness this is going to happen any time soon is miniscule. All the toolkit support more OSs than Linux, so they would have to suddenly decide to not support any of the BSDs anymore from one moment to the next.

Then there is the ignorance and arrogance about other peoples workflows, it's all about that your workflow has been disrupted and you don't want to learn anything new, but don't care about that it might have improved the workflow for everyone else. I actually find that especially unity has much improved on the workflow of Gnome or KDE. It is much easier to control your programs without a mouse for example. I find it interesting that the people who call themselves "power users" are so obsessed about clicking icons, search based interfaces like unity are much much quicker.

Changes and complaints

Posted Mar 6, 2012 23:14 UTC (Tue) by dlang (subscriber, #313) [Link]

be careful in claiming dishonesty.

many of the people complaining about wayland see dishonesty in the statements 'refute' their worries about network transparency for example (I am not going to re-hash this topic here). There are enough people making the claim that "Linux is all that matters" that worries about toolkits becoming Linux/Wayland only is a valid concern (you may disagree on the probability, but it's bordering on dishonest to say that there is no reason to be concerned about it)

Complaining about breaking existing workflows isn't incompatible with other people experiencing better workflows. Creating new options that some people like better is a good thing. Regressions for existing users is however, a very very bad thing. It takes probably 100 new 'good' things to counter a single regression.

That doesn't mean that regressions can always be avoided, or that it's right to avoid them at all costs, but far too many projects are so focused on the new stuff that they give short shift to the concerns about regressions.

you can't change a GUI layout at all without some cost to existing users (muscle memory in using the old GUI if nothing else), so if you are going to change it, you need to think hard about it and make sure that the improvements really are worth the cost.

Changes and complaints

Posted Mar 7, 2012 1:49 UTC (Wed) by CycoJ (guest, #70454) [Link]

Well it's still dishonest to say that Wayland developers have said there will not be any network transparency, or saying network transparency is not possible with Wayland. If you believe the refuting statements are dishonest than you should say "the wayland developers are lying when they say there will be network transparency, or that network transparency is possible".
Also please tell me which toolkit developers have said linux is all that matters?

I also disagree with the premise that it takes 100 new "good" things to counter a single regression. Why is the workflow that the existing user has gotten used to 100times more important than the workflow of new users who previously did not use the tool, because it didn't work for them. What about users (or developers) who actually find that they can work better in a different way, so they are not allowed to change something to their new way, because the new way might upset some other existing users? Take it even further what if half the users decide they can work better differently, 60%, 80%...?

Changes and complaints

Posted Mar 7, 2012 6:47 UTC (Wed) by dlang (subscriber, #313) [Link]

> also disagree with the premise that it takes 100 new "good" things to counter a single regression.

the workflow of your current user is more important than the workflow of the new user simply because your existing user _IS_ an existing user.

This also doesn't mean that you can't offer a new workflow, you just need to keep the existing workflow working

People like new features, but if the software/service/device can't be counted on working, it doesn't matter how good it is when it is working, it can't be counted on so people dump it for one that does.

Developers like to talk about neat new features, but users talk about the time that they lost everything due to a crash 10 years ago. Once you develop a bad reputation it's really, really hard to clean it up

pick any hard drive brand and think about how many people you know that will tell you never to buy that brand, based on the fact that they had a bad experience with it and lost data (or heard that someone else did)

100:1 may not be the exact right ratio, That was a number I grabbed out of thin air, but if you think it's 1:1 (breaking one existing person is Ok if it fixes another person), you are dreaming.

go back and look at the industry talk around linux years ago before Linus instituted the 'no regressions' policy. Developers were willing to break existing users if they thought that they were improving things for more users. The public, users and experts didn't see it that way. They saw it as "Linux just can't be counted on because what works today may not work after your next upgrade". Even years later, you still run into people with that mindset.

The "Enterprise" kernels where they backport massivly but don't change the version number but managers are happy to upgrade blindly is one example of this crazyness, because "everyone knows" that upgrading versions is risky, but upgrading a patch release isn't"

Never mind that this has been proven to be false time after time (how many thing break with each Microsoft service pack?), the risk is in the amount of change and the amount of testing that the changes get.

This is also why time-based releases are gaining popularity, each release is smaller, and so the total testing of the codebase is higher for each release. This doesn't just mean the QA testing of the project, the more code remains from the prior release, the more normal users are using it, and in the process testing it.

I could probably go on for a while, but this is probably a good subject for our esteemed editor to write about, and he can put it much better than I can, and pull in far more examples and references.


Copyright © 2012, Eklektix, Inc.
This article may be redistributed under the terms of the Creative Commons CC BY-SA 4.0 license
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds