LWN: Comments on "Nine Attitude Problems in Free and Open Source Software (Datamation)"

Nine Attitude Problems in Free and Open Source Software (Datamation)

vonbrand — Mon, 27 Oct 2008 01:18:27 +0000

Yet again, somebody who thinks that everybody must obviously do as they would like... the reality of the "community" is that it is inmensely diverse, and as it expands further it just gets more diverse. Sure, it includes (or doesn't, depending on how you define it/where you draw the frontier) people who don't worry about using propietary nVidia (or other) drivers or propietary programs, and also purists that wouldn't touch a piece of hardware with a ten foot pole if it is driven by closed-source firmware. Just define "community" as the group of people who go along with your particular world view, and everything is much simpler (and there is much less growth ;-)

Nine Attitude Problems in Free and Open Source Software (Datamation)

ovitters — Wed, 22 Oct 2008 22:30:18 +0000

Were the tiny improvements in Gnome 2.24 vs. 2.22 worth making a whole new release for? It's hard to build excitement when a new major release introduces a boatload of new bugs but otherwise doesn't change the user experience in any meaningful way.

Just because we don't report the hundreds of bugs that are fixed every release doesn't mean that such work hasn't been done. Further, GNOME uses a time based release method. There will be a release, even if no features have been added. What you're proposing is a feature based release method. We don't work that way, nor do we want to. The 6 months is further nice as it never gets too unstable. Try making software stable after it has been hacked upon for years and had no great user testing.

ou only need one rabid, foaming-at-the mouth Microsoft-hater to make the community look silly - and we have a small army of these guys.

I don't consider those part of the community (at least none where I participate in). Just because someone rants about Windows/Microsoft and praises Linux doesn't mean he either uses Linux or is part of the community. Of course, with a very wide brush you can consider everyone part of the same group, but (IMO) that is stating what people do in general based upon what you noticed in one country.

One Attitude Problem in Journalism (Me):

dgm — Wed, 22 Oct 2008 08:26:22 +0000

> Well the article was probably written with the assumption that you want software that will gain popularity, be usable, and be used by a much larger audiance then just your fellow developers in your project.

A matter of interpretation. I read the article as if you have the *obligation* to listen to any random person downloading the software. To put it simple: you don't. My point is that you may have popularity, usability *and* audience, by listening to the people that contribute (code or whatever), and safely ignoring the rest.

Agree, and disagree

mjthayer — Wed, 22 Oct 2008 07:26:28 +0000

That doesn't just apply to an application, it also applies to an operating system (or distribution in Linux-speak). Especially since "average users" may not know where one ends and the other starts: just the other day I was giving computer advice to a highly intelligent and - for his own needs - reasonably computer-literate person. He was not aware that MS Office was not an integral part of Windows, and assumed that he needed Vista on his laptop in order to be able to exchange documents with other people using Vista.

Talk is cheap

dkite — Wed, 22 Oct 2008 03:33:55 +0000

Thank you for stating the obvious.

Very few people contribute in any way to free software. Those who don't can
expect nothing.

In fact, they get much. An amazingly rich choice, of varying degrees of
finish and usefulness, which is in direct proportion to the number of users
who find it useful and are moved to contribute to make it better for
themselves.

It really is very simple. If you don't like it, fix it.

Derek

One Attitude Problem in Journalism (Me):

drag — Tue, 21 Oct 2008 15:36:10 +0000

> First, one has to realize that there's no "customer". As you're not paying for the software nobody has any obligation to listen to your opinions.

Well the article was probably written with the assumption that you want software that will gain popularity, be usable, and be used by a much larger audiance then just your fellow developers in your project.

In that case you can replace 'customer' with 'target audiance'.

If you have no desire for other people to actually use your software then it's obviously quite pointless to care what they have to say about it and there is no point in incorporating people who contribute documentation, user support, sysadmin support, and other non-programming improvements into your core project (and other points he is making in this article)

However if you do care about gaining a wider audiance or potentially displacing a existing proprietary software then your choices and focus are going to be quite a bit different.

Nine Attitude Problems in Free and Open Source Software (Datamation)

forthy — Tue, 21 Oct 2008 14:26:22 +0000

Note also that many people who rant about me-to-software don't even know the innovative. When you have OpenOffice.org (me-too), you have at the same time LaTeX+LyX (very original, innovative free software, with TeX being even older than the FSF, LyX being older than KDE, and even though still unmatched by Word and clones). Why do people copy an inferior approach like that of Word? Because the non-coding stupid users that have no vote (as the article tells you) want it that way!

Anyway, there's nothing wrong with clones. Commercial software does clones all the time, too. Word and Excel are in no way original software; not even Lotus 1-2-3 was. People just like things being done a particular way, and so clones come into existence. Just being different for the sake of being different doesn't cut the mustard.

Agree, and disagree

drag — Tue, 21 Oct 2008 13:53:51 +0000

> I would be wary of "Being content with imitating other operating systems". I agree that merely copying other system's features, blindly, is often bad.

Yeah.. You have to prioritize.

There are a minimum set of required features for any application. "Hard requirements", so to say.

OpenOffice.org is a excellent example of this:

A minimum required feature for any office suite is the ability to communicate with other people and share documents. The ability to save documents, send them to people, and have those people be able to read and edit those documents. As well as a edit documents sent from other people.

That's pretty obvious, right? A office suite that doesn't allow you to save documents or puts it into a format you can't extract from your local harddrive would be pretty worthless.

Well part of this minimum requirement is going to be Microsoft Office format compatibility. When people send you files they are, more then very likely, be files made with MS Office. And when you send documents to other people they are, more then very likely, be using MS Office.

So when comparing any Office Suite vs OO.org one of the top things you need to consider is it's ability for the user to interact with other people who are all using MS Office file formats.

If Koffice isn't able to read/write Microsoft documents to the same level of compatibility that OO.org can, then it loses by default. It doesn't matter if it's faster, easier to use, has more features, better documented, is better looking, or discovers entire new and revolutionary UI features.. it's going to be very much less useful for the average person simply because it fails to meet a minimal requirement.

-------------------------------------

So there is a priority that developers have to follow if they want to have a successful product (in the order of importance):

1. Minimal hard requirements --- Things that are absolutely necessary to get the job done. Without these set of requirements the software is just going to be mostly useless.

2. Features that users expect --- Things that users expect and need.. like Being relatively bug-free, having documentation, being usable usable, as well as features that are present in competing software that users are fond of.

3. Features that set your program apart --- The 'innovation' part of your program.. things that your program does better or newer that makes it more attractive.

Until you have priority 1 done then your software is worse then useless.. it's just a time sink for you and your project members. Until you have priority 2 done then don't expect anybody to actually want to use your software. After that then you can get working on number 3 and actually being competitive.

So basically this means that if your trying to compete with a existing, established, and successful software product then that does mean a whole lot of copying: Unless your able to provide the users with what they already have then they won't give you the time of day. Worrying about making that software cheaper/faster/better is hugely secondary, even though it's important.

Talk is cheap

dmaxwell — Tue, 21 Oct 2008 13:45:31 +0000

A "sermon" is basically talk and talk is cheap. For most developers, the hierarchy goes like this and for good reason:

1. Employer
2. Paying Customer
3. Fellow developer who either supplies a patch or detailed analyses that guide in actual code writing.
4. Skilled User (that one who is willing to run straces and the like)
5. Respectful nonskilled user who provides good detail and can follow instructions
6. Nonskilled deer in the headlights. This one is a quantity thing. You may want to heed a multitude of reports that "foo doesn't baz when I frob!"
.
.
.
.
.
29421 whingers

Developers often don't have a customer relationship with their users so they aren't vendors which means the developer is correct to have an expectation of respect or co-operation from users who contact him. The user got something for free that the developer didn't have to provide. So whinging and flamage won't go over well. Developers are only obligated to #1 and #2 and it is generally wise to try getting along well with #3 and #4. Our pulpit minister is completely off base as regards the other groups.

#1 and #2 not being served well isn't an issue with FOSS per se. Employers and customers are correct to expect what they paid for and to have answers and fixes for their issues.

One Attitude Problem in Journalism (Me):

dgm — Tue, 21 Oct 2008 09:52:13 +0000

> This is free people working on software. There are not going to do what they do not feel like doing. Trying to tell them how they should behave is completely pointless. Do not forget most of them are spending their spare time to do what they like; they are not going to do what they do not like.
> And this is open software. If you are convinced that something must be done, then do it yourself. Join whatever project you like and add your two cents.

Right on the spot.

I once used to thought like the author of the article, too. It takes some time to realize that things are as they should be, for very logical reasons.

First, one has to realize that there's no "customer". As you're not paying for the software nobody has any obligation to listen to your opinions.

Second: there's no "community". We are not a corporation with a single image. We're not Linus' or Richard's merry boys. And of course we're not a bunch of Microsoft haters. Each developer and group is a separate entity, with his own background and objectives.

But the most important thing to realize is: contributors rule because it's the only way to make stuff happen. This is the key. Only projects ruled by the people that make them survive and evolve. Opinions and ideas are nice, but stuff that works is what makes the world tick.

One Attitude Problem in Journalism (Me):

eduperez — Tue, 21 Oct 2008 07:30:50 +0000

Trying to telle FOSS people what they should do!
This is free people working on software. There are not going to do what they do not feel like doing. Trying to tell them how they should behave is completely pointless. Do not forget most of them are spending their spare time to do what they like; they are not going to do what they do not like.
And this is open software. If you are convinced that something must be done, then do it yourself. Join whatever project you like and add your two cents.

Nine Attitude Problems in Free and Open Source Software (Datamation)

tetromino — Tue, 21 Oct 2008 02:00:50 +0000

Bollocks? No, the article is right on the money.

"Not knowing allies from enemies" - check. Just to take one example, look at the amount of badly reasoned Novell-bashing in the community.

"Talking software to outsiders" - check. The average man on the street doesn't care about the distinctions between GPL-2 and GPL-3, or the benefits of not including non-free firmware. What he cares about is the fact that if he is using free software, he is, for example,insured from his data becoming unreadable when the software he is using is discontinued. In our propaganda, we should be emphasizing the benefits to potential adopters, not concentrating on splitting legalistic hairs.

"Being content with imitating other operating systems" - check. How much of our software is truly innovative and trend-setting? Well, there is TeX, Apache, Perl/Python/Ruby, perhaps Firefox (although for the most part it just popularized Opera's innovations) ... that's about all that I could come up with, off the top of my head. The lack of innovation and the habit of merely reimplementing features others invent is not exactly building a strong case for the superiority of the free software process.

"Hostility to newcomers" - check, check, check, with thick red marker.

"Giving the developer a special position" - well, this is a bit strongly stated, the developer's position should be special because they are the ones actually creating the software. However, when developers ignore input from anyone who does not submit code, they end up creating software that's only usable by people who know how to code. That's OK for a kernel or a library, but it's definitely not OK for anything user- (or admin-) facing.

"Focusing on hating Microsoft" - check. You only need one rabid, foaming-at-the mouth Microsoft-hater to make the community look silly - and we have a small army of these guys.

"Taking commercial development as a model for growth" - in regards to releasing for the sake of release, I would have to agree. Were the tiny improvements in Gnome 2.24 vs. 2.22 worth making a whole new release for? It's hard to build excitement when a new major release introduces a boatload of new bugs but otherwise doesn't change the user experience in any meaningful way.

"Making market share the top priority" - check. If nobody uses our software, we merely do not win. But if through propaganda and clever marketing we introduce our software to a new user, who then discovers that it is fragile, buggy, unusable, and unfinished, then we actually lose.

The only point where I would disagree with the author is the last one, "Stopping short of a completely free operating system". I don't think that the entire community agrees on what constitutes a completely free operating system, and I don't think that everyone has subscribed to the original GNU vision of a 100% free stack. That said, if we do ever achieve a 100% Free system (from the BIOS up) that's actually usable and competitive, it would be a truly awesome thing.

Nine Attitude Problems in Free and Open Source Software (Datamation)

tbrownaw — Tue, 21 Oct 2008 01:51:40 +0000

1) Not knowing allies from enemies

At times, these pundits say things that nobody else will, as Jeremy Allison points out. But, more often, their motivation seems to be solely to make a name for themselves, regardless of the divisions they make in the community, and, by reading such pundits, we encourage the continuation of those divisions.

Who is the "we" that reads them? (Regularly, as opposed to just whenever it's necessary in order to intellignetly discuss a link posted on Slashdot or LWN or some other place.)

Even worse is the division between free software and open source advocates. Admittedly, the philosophies are different: free software is about user freedom, while open source is about quality software. Yet, despite these ultimate differences, members of the two camps work on the same projects with the same licenses, and are seen by both themselves and outsiders as having far more in common than not.

The work gets done, so what's the problem?

So why dwell on the differences? The obvious truth is that free software and open source advocates will never find anyone with whom they have more in common than each other.

Arguments are the natural result of one group evangelizing ideas in the presence of a different group which thinks those ideas are misguided or harmful.

2) Talking software to outsiders

Recyclers don't pitch their cause by explaining where recycled goods are taken for reclamation, or explaining the process by which glass is melted down for reuse; instead, they talk about the benefits that recycling has on the average person's life.

In the same way, instead of talking about software or its licenses, the FOSS community needs to talk more about issues such as consumers' rights and privacy and free speech -- matters that extend far beyond the keyboard and terminal.

This would require that the practical benefits be promoted as the main point, rather than the license or abstract ideals being the main point.

3) Being content with imitating other operating systems

I haven't really seen this. Maybe I just haven't been looking hard enough?

4) Hostility to newcomers

This is the stereotype, but I think it has been improving.

5) Giving the developer a special position

If a non-developer makes a suggestion that would help the project, too often the response from a developer will be, "We look forward to your code" -- with the developing knowing very well that they are not talking to a coder.

Suggestions are cheap, or at least most of them are. Are developed, well-thought-out and supported (by data or just good logic) specific reccomendations given the same treatment, or is this just analagous to a spam filter? Of course that doesn't mean that a more polite and educational spam filter is a bad idea...

6) Focusing on hating Microsoft

However, some segments of the community seem less interested in computer freedom than in expressing loud and unwavering hostility to Microsoft.

And some are less interested in producing quality software than they are in agitating against practitioners of certain (fundamentally doomed anyway) business models.

But probably the greatest reason for rejecting the strong anti-Microsoft prejudice is that it can blind the community to other proprietary opponents. No one, for example, seems to be concerned about Apple's proprietary actions, even though in many ways it is emerging as FOSS' chief opponent.

Have you ever seen the comments on just about any Slashdot story mentioning Apple? I certainly don't see this lack of concern.

7) Taking commercial development as a model for growth

In some cases, borrowing ideas from commercial development may be useful. Yet it should never be forgotten that, while FOSS can work with commercial development, its goals are different. What happens, for instance, to the open source idea of only releasing software when it is ready when a project commits to regular releases? Sooner or later, problems in quality control will be inescapable.

This is a case of not giving the developers a special position (point 5). Developers are often perfectionists, but users (particularly corporate/institutional users) like predictability. Quality probably need not be inescapable, however -- that's what "planning ahead", "merge windows", "feature freezes" and similar things are for.

8) Making market share the top priority

However, as I have written before, more users mean nothing if they are gained by giving up FOSS ideals, or if those users do not support them. In the excitement of being recognized at last, the community needs to remember that the goal is not just to provide alternative software, but an alternative philosophy and relation to computing.

Whose ideals and whose goal? This goes back to point 1 and lumping the "open source" and "free software" people together as the "FOSS community". Probably the closest thing to a consitent ideal that could be given up would be to aquire users by marketing rather than superiority, but even that's questionable for the segment of the community that identifies primarily as being against proprietary models.

9) Stopping short of a completely free operating system

Now that the goal of a totally non-proprietary operating system is in sight, you might imagine that people would like to push on and finish the job. However, as reactions to the Free Software Foundations's recent relaunch of its high-priority list shows, a surprising number of people in the community feel no need to achieve the ultimate goal.

Which segment(s) of the community, and surprising to who?

Discussing the problems

My real point is that FOSS has grown so large so quickly that it has not had the time to question whether old attitudes were still useful or new approaches consistent with core values. Before it grows any bigger, the community needs to examine its attitudes and evaluate them. Otherwise, it risks, if not collapsing under the contradictions, then handicapping itself unnecessarily.

This assumes that the "FOSS community" is a single cohesive organization. It isn't, it's a loose collection of ill-defined groups moving in generally the same direction. It doesn't (can't) determine overall attitudes in that kind of top-down philosophical manner, partly because there isn't a single set of core values. Instead change takes a more evolutionary nature, and the contradictions are a vital part of this.

Nine Attitude Problems in Free and Open Source Software (Datamation)

himi — Mon, 20 Oct 2008 23:13:05 +0000

Azureus is an also-ran? Maybe if you don't happen to use it yourself, or you prefer another client, but it's hardly what you might call a failure, and it's an excellent bittorrent client.

I think this reveals an attitude problem in the general user population: different is worse. Most of your examples of "poor knockoffs", "bad clones" and "also-rans" can only really be considered in a negative light if you compare them with some arbitrary competitor (often the Microsoft one, for some reason). Considering them on their own merits they're generally excellent pieces of software, often with significant advantages compared to their dominant competitors.

himi

Nine Attitude Problems in Free and Open Source Software (Datamation)

nix — Mon, 20 Oct 2008 22:41:55 +0000

I was covering the unpaid ones, as those are generally the ones who get
hit with ridiculous Do This instructions from random people who aren't
paying them. The ones who get paid are generally given Do This
instructions from people who *are* paying them. This must be dealt with in
an entirely different manner :)

Nine Attitude Problems in Free and Open Source Software (Datamation)

JoeBuck — Mon, 20 Oct 2008 22:32:17 +0000

I didn't interpret the article as calling for random members of the general public to have votes. Rather, he was talking about non-developer key contributors. These might include documentation writers, QA people, application experts (e.g. graphic artists for a graphics application), etc., who in a particular case might contribute as much to the success of a project as the developers do. For example, sometimes I haven't noticed a design problem in my software until I attempted to document it; then I realized that the features I was having a hard time explaining were just broken as designed.
<p>
Not that I'm all that keen on votes, as a general rule. If a project makes a decision based on a 51-49 vote about an important issue, something's usually wrong, and it is usually better to work harder to achieve a solution that almost everyone can live with.

Nine Attitude Problems in Free and Open Source Software (Datamation)

wertigon — Mon, 20 Oct 2008 22:24:48 +0000

Who said it's a "free time" developer? I get paid to crank out GPL code you insensitive clod! >_<

Nine Attitude Problems in Free and Open Source Software (Datamation)

chromatic — Mon, 20 Oct 2008 21:09:20 +0000

Just look at your desktop environment and note how many applications are clones and how many are truly innovative.

Why should I care how "innovative" a piece of software is, if it meets my criteria better than any other piece of software? (One criterion is "is the software Free"?)

Nine Attitude Problems in Free and Open Source Software (Datamation)

robilad — Mon, 20 Oct 2008 20:35:32 +0000

Our work is never over! ;)

Nine Attitude Problems in Free and Open Source Software (Datamation)

sdalley — Mon, 20 Oct 2008 20:32:04 +0000

Yes, your post portrays the attitude problem perfectly.

Nine Attitude Problems in Free and Open Source Software (Datamation)

nettings — Mon, 20 Oct 2008 19:50:26 +0000

the article was actually quite insightful imho, although the editor's introductory line "sermon for the day" seems to suggest otherwise. i must confess i read it only because i expected some fun with somebody making a fool of himself. in that respect, i was disappointed :-D

Nine Attitude Problems in Free and Open Source Software (Datamation)

nix — Mon, 20 Oct 2008 19:06:38 +0000

Wow, look at the half-baked reasoning.

So you'd like free-time developers (now a minority) to do whatever some
user says, even if they disagree with it or think it unimportant? I can
verify that when a developer *does* agree with something and think it
important, that thing will get done, even in the absence of code.

And I see no 'fixation on Microsoft as the enemy' among the actual
*developer* community. There's wariness because they have *actually
threatened* us with things like patents, and there's dislike of their
systems, but fixation? 'Enemy'? You've been reading Groklaw too long.
That's not how the hackers think. It's just another system, a not very
nice and distinctly non-Unixlike system that we might have to ensure
portability to but which thank goodness we don't have to develop on.

Agree, and disagree

dwheeler — Mon, 20 Oct 2008 18:39:07 +0000

It's an interesting post.

I agree that disrespecting contributors - who happen to not contribute code - is a BAD idea. Communities should respect EVERYONE who contributes useful material, whether it's code or not. User interface design and testing are painful, but the people who do it should be respected not abused. Thankfully, there are many welcoming communities, as well as some abusive ones.

I would be wary of "Being content with imitating other operating systems". I agree that merely copying other system's features, blindly, is often bad. But users are normally not tabula rasa (blank slate); if they're used to doing things in a particular way, it's often better to present things to them the same way. In user interfaces, you better have a REALLY good reason for doing something different; otherwise, do it the way the user is probably used to. The user will be grateful!

Nine Attitude Problems in Free and Open Source Software (Datamation)

einstein — Mon, 20 Oct 2008 17:59:56 +0000

> Just look at your desktop environment and note how many applications are clones and how many are truly innovative.

Well, let's see... windoze users jaws drop when they see the compiz cube on my desktop at work. Even on a low spec machine, linux is doing things that neither microsoft nor apple has been able to provide on high spec machines.

As for azureus or open office, they may be offering similar functionality as existing products, but both are pretty useful IMHO.

Nine Attitude Problems in Free and Open Source Software (Datamation)

pheldens — Mon, 20 Oct 2008 17:19:20 +0000

I think attitude is what makes FLOSS fun.
I hate corporate grease balling, especially because they screw you over in the end just as hard.

Nine Attitude Problems in Free and Open Source Software (Datamation)

raven667 — Mon, 20 Oct 2008 17:02:04 +0000

I disagree with your summary, I think that this article is right on target. Just because one is a
user or fan of something does not mean that they should be blind to its faults. Take a look at
some of the references at the Linux Haters Blog to see examples several of these points.

It is not a secret or debatable that most projects are run by the developers and that non-
developer types are second class citizens in the decision making process for most FOSS projects.
This breeds software that only a developer could love because they are the only ones whose
opinions are considered in the process. "Show me the code" or "Send in a patch" is reminiscent
of the old practice of bogus competency tests for voting registration in the poor and minority
areas of the USA in the last century, it is a justification for keeping out the unwanted opinions of
the customers and users of FOSS software projects.

One only need to point to popular sites such as Slashdot or even the comments in this fine
publication for plenty of examples of the vocal anti-Microsoft camp. Some popular information
outlets such as Groklaw have very little else to recommend them except for rampant anti-
Microsoft paranoia. This fixation on Microsoft as the enemy is ultimately detrimental as it
reflects poorly on the community as a whole, making it seem trivial and shrill, it is polarizing and
leads to rejection of FOSS for many people who are not "true believers of the Great Evil".

There is much in the FOSS desktop software ecosystem which is just a poor knock-off of existing
proprietary software where the FOSS software is an "alternative" rather than the trendsetter or
leader of its particular category. Just look at your desktop environment and note how many
applications are clones and how many are truly innovative. There are a few good ones like
Firefox, Pidgin/Adium but there are plenty of bad clones like Evolution, OpenOffice and also-rans
like Azureus/Vuze. In non-desktop software there are plenty of infrastructure components
where the FOSS implementation is the leading implementation such as for web servers, java app
servers, large server operating systems in general but in desktop software there is very little.
They are mostly following behind Apple who is leading or following behind Microsoft who is
several steps behind.

Nine Attitude Problems in Free and Open Source Software (Datamation)

wertigon — Mon, 20 Oct 2008 16:58:52 +0000

Actually, he has quite a few good points in there.

Like, why should we *not* prioritize a free operating system? Personally, I feel the FSF is very much needed, and distros such as gNewSense makes a wonderful job of pointing out the remaining bugs for a 100% free alternative Linux Distro - But at the same time, a regular user won't care why something is broken, he or she just wants to fix it and move on with whatever they are doing.

However, the goals doesn't have to be mutually exclusive, and I'm still optimistic for the future. As long as the Internet keeps on being as free as it is, Free Software will get better, faster, harder and stronger. :)

Nine Attitude Problems in Free and Open Source Software (Datamation)

stumbles — Mon, 20 Oct 2008 15:43:07 +0000

Bollocks. I thought about posting more, but that sums it up.