Weekly Edition
Return to the Press page
| |||||||||||||
|
| |||||||||||||
Canonical to fund upstream Linux usability improvements (ars technica)
ars technica
reports on Canonical's efforts to improve the Linux desktop.
"Ubuntu founder Mark Shuttleworth announced Wednesday that his company, Canonical, will hire professional designers and interaction experts to improve the usability of the Linux desktop software ecosystem. They will work closely with upstream developers to bring a better experience to users of the open source operating system.
The charismatic frontman of the Ubuntu phenomenon made headlines for his keynote at OSCON earlier this year when he called for the open source software community to take on Apple and Microsoft and turn Linux into a platform that delivers superior usability and attractiveness."
(Log in to post comments)
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 11, 2008 20:59 UTC (Thu) by sbergman27 (subscriber, #10767) [Link]
I'm wondering what the next wave of criticism directed at Canonical will be. "Too little too late?" Maybe for a while. And then "Look! They're trying to seize control of Linux development! Canonical is the new Microsoft!".
Seriously, though. This looks like a good move for Canonical, Ubuntu, Linux, the *BSDs, and... well... pretty much everyone.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 11, 2008 21:03 UTC (Thu) by k8to (subscriber, #15413) [Link]
Usability folks working with free software developers is one of those things that's been hard to figure out, so far. If this is a success it should have good dividends. I hope it does.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 11, 2008 21:19 UTC (Thu) by drag (subscriber, #31333) [Link]
It should help.
First Sun and then much more recently Novell, had their own usability studies for Gnome desktop in the past. Both of these had very positive effects on the basic usability of the Linux desktop, even though some people tend to castigate the whole Gnome 'HIG' stuff that came out of the Sun studies. And somewhat ironically, it seems that a lot of the effort that Novell put into Gnome paid off in the form of Ubuntu garnering so much attention for creating a relatively slick environment.
Usabilities studies for free software have a similar effect to usability as giving a man dying of dehydration in the middle of the desert a canteen of water; It's desperately needed and should benefit everybody involved.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 11, 2008 21:17 UTC (Thu) by jspaleta (subscriber, #50639) [Link]
It's never too late.
I doubt that anyone would argue that upstream usability enhancements are a really good fit for how Canonical can contribute back to upstream. It would make buckets of sense for Canonical to pick up where Novell left off its drive to lead upstream usability efforts in Gnome.
I look forward to seeing new blood showing up in upstream development discussions, bringing Canonical's perspective to the development effort. When should we expect that exactly? It seems Mark didn't actually make any concrete statements as to when to expect the new team to be in place.
-jef
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 11, 2008 21:44 UTC (Thu) by SEJeff (subscriber, #51588) [Link]
Especially with some of the guys that work for canonical
- Neil Patel - Avant Window Navigator - MacSlow - Cairo Clock, crazy opengl apps - Ted Gould - inkscape
to name a few
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 0:20 UTC (Fri) by sbergman27 (subscriber, #10767) [Link]
"""
When should we expect that exactly? It seems Mark didn't actually make any concrete statements as to when to expect the new team to be in place. """
You sound skeptical, Jeff. Consider how much in the way of resources Canonical devotes during each release cycle to fixing all the poor usability that floats down from upstream every six months. Seems like I frequently hear members of some distro communities arguing that contributing work back upstream is in the best interest of distro maintainers? Do the people arguing that really believe it? If they do, they should have no reason for skepticism. Stemming the flow of unusability from upstream likely has some major "itch-scratching" motivations behind it.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 23:15 UTC (Fri) by jspaleta (subscriber, #50639) [Link]
I want to make sure that people have reasonably calibrated expectations.
I would hate for Canonical to get knocked around because Mark has jumped the gun..again...making announcement on an intention for Canonical to do something and getting people's hopes up before the wheels are actually in motion to get it done.
I currently only see one job description in Ubuntu's employment page which delineate coordination with upstream projects, the Ubuntu GNOME Maintainer opening (which has been there since March of this year). The other postings don't make working with upstream a specific job duty, and I'm not seeing anything with regard to new hires for usability.
http://webapps.ubuntu.com/employment/
There's a lot of ways one can view that. Either Mark's got the team(s) mostly together already, and things are pretty much going to happen "real soon now", or the wheels haven't started turning yet to do the hiring which means everyone will need to wait for that process to complete. It's not clear which it is. He's mixed tenses in his blog post when describing the team so its not clear if the hiring process is an active one.. or if the hiring process is pretty much done and they are ready to rock and roll.
If hiring needs to be done to fill out the team, it could be a 6 months or more before there's any appreciable contribution to upstream from this effort. The fact that the Gnome Maintainer position is still listed as open probably is indicative that upstream Gnome won't be seeing a Canonical manpower injection until that position is filled.
If someone knows of specific Canonical job postings which have come and gone in the last few months which directly address usability or any of the upstream projects Mark has specifically mentioned in his blog post, that would be interesting points of reference. I could have easily missed these announcements. I've not subscribed to the rss feed so I have no information about previously posted and filled jobs. And the internet archive doesn't track that page. Anyone subscribed to the rss feed and seen an "X, OpenGL, Gtk, Qt, GNOME and KDE" engineering positions or usability come and go?
-jef
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 13, 2008 17:00 UTC (Sat) by sbergman27 (subscriber, #10767) [Link]
I'm sure that the BDFL of the most successful desktop Linux distribution of all time[1] appreciates your kind concerns.
[1] Note the use of the word "desktop" there. Sometimes people try to argue that Xandros is really the most successful, due to the EEE pc, or that some cell phone distro is destined to be the most successful.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 11, 2008 23:33 UTC (Thu) by dowdle (subscriber, #659) [Link]
They already had one guy doing that a while ago. I'm assuming he is still there?
I don't think anyone is going to be critical of Ubuntu for doing usability testing... although I'm not really sure how they are going to do all of the development needed in the various desktop environments and the apps... nor if they get all of that work done, how they are going to get it upstream. I do wish them luck with it.
You should recall though that Novell has done the same thing. They conducted a lot of usability testing, videotaped it, and have published the videos online for everyone. I'm not sure what, if anything, has come as a result of that. I would be nice to get a report from someone at Novell how they used the data. Maybe that information is already out there somewhere and I just haven't noticed?!
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 11, 2008 23:35 UTC (Thu) by dowdle (subscriber, #659) [Link]
Just to follow up on that last post... here's the URL for the testing Novell has done:
Wow. I hope this doesn't make it sound like I'm a Novell fanboi because I am NOT. :)
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 0:45 UTC (Fri) by qg6te2 (guest, #52587) [Link] One of the issues with many Human Interface studies/guidelines is the target audience. If the interface is dumbed down to the supposedly average Joe & Jane, you will be alienating a large minority who want to do more with their desktop, and/or who want to do things quicker. What's really needed is an interface that can cater for both audiences, with the more advanced features as an option. As one possible example, there should be an option for vertical-only maximisation of windows (for example, via double clicking the title bar). This helps immensely with multiple shell windows as well as file selectors - the user sees more, and it's a lot quicker to execute than manually resizing windows.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 2:12 UTC (Fri) by mattdm (subscriber, #18) [Link]
Yes! I can't believe how many "usability" studies I've seen which appear to focus on the first-ever interaction with the program. That's important, but it's only a fraction of the picture. Real usability is about day-to-day, not just the jump-off point.
Usability experiences Posted Sep 12, 2008 6:20 UTC (Fri) by eru (subscriber, #2753) [Link] If the interface is dumbed down to the supposedly average Joe & Jane, you will be alienating a large minority who want to do more with their desktop, and/or who want to do things quicker.I don't think usability is just about making it easier for less-sophisticated users. Programs with poor interface typically make things bad for everyone, for example by making common operations require too many steps to accomplish, or destructive mistakes too easy to make. In the past summer I had the most enlightening experience of teaching an elderly newbie to computers to get online and use e-mail. Of course I set her up to use Linux, with a friendly distribution (Mandriva 2008.0 with some extra localization and KDE) I configured for her, but all the problems she encountered really had absolutely nothing to do with the OS or window system. We nowadays assume that when you have a GUI, everyone can easily learn to use it. Nonsense. We had to tackle basic concepts like use of mouse and menues, how to start programs and make choices. She had no idea about the difference between the hardware, OS, application and the internet. When the system inevitably showed some error or warning messages, and I tried to help by phone, it was difficult as she could not really tell me if the problem was a dialog box from the web browsers, something actually on a web page, or a message from lower layers. I can now well believe that most users like her immediately fall for the fake warnings from addware- and malware-pushing web sites or emails... (one good reason for using Linux in her case!). And she was not stupid at all, having had a long career in a quite demanding profession. It is just that she had no contact with the wacky world of computers. In this situation some of the configurability of the UI backfired. For example, in KDE you can put the task bar at any side of the screen you want by dragging, or make it hide by a click in certain place. Being uncertain with the mouse, she sometimes dragged or hid it by mistake, then wondered if she had broken the computer. Apparently using the mouse is much more difficult to learn than we who have learned it long ago remember. And it requires a degree of precision and muscle control that some older people may no longer have. I recommend that anyone who is working on user interfaces try replicating my experience: Try to find someone who has never used computers, and teach him/her the basics of daily PC and net use. It will make you see the whole field with new eyes.
Usability experiences of elderly users Posted Sep 14, 2008 6:37 UTC (Sun) by Cato (subscriber, #7643) [Link]
I've had a very similar experience with an elderly relative, who has in fact been using email since the mid 1980s but still struggles a bit with dragging things without intending to (motor control again). She's now on Ubuntu having used DOS/Windows for many years.
Being able to customize and then mostly lock down the desktop is very important - Windows makes it far too easy to drag the taskbar around the screen, and Thunderbird also needs more lockdown due to its drag and drop features.
I find GNOME is generally better for this type user than KDE, partly because it looks simpler and partly because KDE is going through a huge change with KDE4 and needs the rough edges knocked off (e.g. the ludicrous scroll menu on the K menu - is that still required?) I write this as a KDE3 user.
Another issue, though, is that some customizations are needed for accessibility, often to make the target for dragging/clicking sufficiently large - for example, GNOME makes it really hard to change the window border size so that someone with poor vision or motor control (many elderly people) can easily click and drag the border when resizing the window. There's discussion at https://bugs.launchpad.net/ubuntu/+source/metacity/+bug/1... - you actually have to customize the Metacity theme which looks easy but is quite hard if you don't want to hack the distro's own theme files, so I gave up. KDE and Windows are ahead here, it's a simple GUI option. Emerald may be a better WM for GNOME for this feature anyway.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 8:18 UTC (Fri) by Cato (subscriber, #7643) [Link]
Whatever simplification or usability optimisation is done for novice users, the advanced user can simply customize the desktop or specific applications, or replace them as needed - e.g. use KDE, Enlightenment, LXDE or specific WMs. I'm sure some of the older WMs do provide the sort of vertical-maximise feature you mention.
The main hassle for advanced users in my view is hardware support, which the sheer volume of novice users could really help with (by creating a big enough market that hardware vendors will document hardware or provide open source drivers).
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 16:09 UTC (Fri) by drag (subscriber, #31333) [Link]
Hehe.
It's amazing the number of people that seem to get upset that Gnome or whatever is loosing 'advanced' users because the 'advanced' users find it much to difficult to use gconf-editor or swap out the window manager without a GUI to help them do it.
Achieving high levels of usability is tragically difficult. I mean it REALLY REALLY is very very hard.
I've said it before that Ubuntu is mostly just a refreshed version of Debian. They use Debian packages and Debian software. They pull stuff almost straight from Debian's repositories and only do a relatively small amount of customization or their own packaging.
People misunderstand this to say that I think they are 'leaching' off of Debian or whatever.
The exact opposite is true! It's praising them for being smart enough to not re-do Debian's work and build something that is actually easy and pleasant to use out of the box!
Anybody who thinks that is 'leaching' or that they are 'dumbing down' the interface is simply ignorant.
You can easily spend 10% of your resources and 90% of the effort developing good documentation and good interfaces and easily end up with a program that is dramatically more useful then something somebody spent 100% of the time developing new features.
Seriously.
That's the problem with a lot of developers. They think of making things nice and easy to use as making interfaces for 'stupid people'. Then when they try to make something easy to use they have this 'stupid person' model in their heads and their interfaces end up just being shit.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 13, 2008 14:34 UTC (Sat) by qg6te2 (guest, #52587) [Link] the advanced user can simply customize the desktop or specific applications, or replace them as needed - e.g. use KDE, Enlightenment, LXDE or specific WMs. No, this is not what I meant. While "advanced users" are perfectly capable of replacing the window manager, I am also sure that for most of them this is an unnecessary Pain in the Butt (tm). The many flavours of "advanced users" should not get lumped into one extreme category. If so, we would all be using Gentoo and configuring the kernel + bazillion packages every day (no disrespect to Gentoo intended). People already familiar with machines want to get work done with a user interface that is conducive to fast work, while at the same time without feeling like they need to unduly modify everything. Furthermore, we are not all the average Jane & Joe, and I do not wish to get lumped with them -- while one cannot please everyone, satisfying two user interface targets instead of one would go a long way.
Ease of use for advanced users Posted Sep 14, 2008 6:25 UTC (Sun) by Cato (subscriber, #7643) [Link]
If you want more UI options within Ubuntu, you also have KDE and XFCE from official Ubuntu variants - I find both pretty customizable, having used both. If you want something else, there are unofficial variants using virtually all window managers (Fluxbox, LXDE, etc) and many WMs/DEs within the repositories. installing Xubuntu for example is just a matter of "aptitude install xubuntu-desktop", so it's hardly advanced.
It isn't realistic to demand huge customizability and great usability, there simply aren't enough advanced users *who agree on the same approach* to form a 'market' that can be addressed like this. If you disagree, it's pretty easy to create an Ubuntu variant - Ubuntulite (LXDE) started just like this, and runs in as little as 96 MB.
Linux and Ubuntu are already incredibly flexible, and I don't see this work making them any less flexible. Since Ubuntu is very close to Debian you retain a huge amount of control over your system, even if GNOME and some apps have usability-driven changes that you don't like.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 13:50 UTC (Fri) by niner (subscriber, #26151) [Link]
Well at least on kwin, vertical maximisation of windows is just a middle click on the
maximisation button. Horizontal maximisation is a right click on the same button (left does it both ways). It's actually one of the things, I'm missing most, when having to manage windows on a Windows machine (the other being alt+right click anywhere on the window to resize it)
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 17, 2008 5:51 UTC (Wed) by Mithrandir (subscriber, #3031) [Link] Well at least on kwin, vertical maximisation of windows is just a middle click on the maximisation button. Horizontal maximisation is a right click on the same button (left does it both ways). Works fine on a stock install of Ubuntu Intrepid. I guess that's down to compiz.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 18, 2008 16:08 UTC (Thu) by Lovechild (subscriber, #3592) [Link]
Why so hostile? Nobody has said anything and yet you are already in defense mode. In fact nobody as said anything but positive things so far about Canonicals decision to contribute and the area in which they will make these contributions.
Now let's wait and see who is announced to be on this team, I for one, look forward to it.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 11, 2008 22:39 UTC (Thu) by andikleen2 (subscriber, #52506) [Link]
IIRC the last time this happened we ended up with the gnome file selector
box which everyone loves (not). Don't call me excited.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 0:09 UTC (Fri) by griffbrad (subscriber, #42805) [Link]
I think the initial designs for the new GTK file selector were done by Seth Nickel (http://www.gnome.org/~seth/blog/) who was a large part of the effort of defining the HIG for 2.x and making usability a core focus of the GNOME community. There are definitely issues with the file selector and many users, for instance, who were accustomed to using only the keyboard with the old file selector really loathed its initial implementations. However, I think it's generally agreed that the efforts of Seth and usability folks at Sun around the time of 2.x have been incredibly successful for GNOME as a whole.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 2:21 UTC (Fri) by yvesjmt (subscriber, #38201) [Link]
But isn't the GNOME file dialog remarkably similar to the one from Mac OSX?
A cynical could foresee that this new initiative will come up with a revolutionary "System Preferences" panel, inspired by... guess what? ;-)
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 9:59 UTC (Fri) by regala (subscriber, #15745) [Link]
because it comes from Apple ou MacOSX doesn't mean it's bad. We are discussing Desktop Usability, not Desktop Usability by carefully avoiding what is usable in other OSes.
honestly, why not just clone Acqua? Posted Sep 12, 2008 14:23 UTC (Fri) by yvesjmt (subscriber, #38201) [Link]
I didn't say I don't like the idea or that it didn't work. It's just funny.
An honest question: since we're not worried about avoiding other OS's ideas, can't GNOME or other desktop enviroment just be shipped by default with the same look and feel of Acqua? What's the issue, patents?
Application Design Posted Sep 12, 2008 18:27 UTC (Fri) by grantingram (subscriber, #18390) [Link] One reason is that Mac applications are designed to work with their own operating system conventions. For example the maximize button makes the window bigger rather than filling the whole screen. Mac applications typically use lots of smaller windows so you don't always want a single window to use all the screen. In other desktops (e.g. GNOME,Windows XP) the applications work much better when they use up all the screen. For example using OpenOffice in a number of small windows is slightly pointless as drag and drop doesn't work, the application doesn't communicate at all with the filemanager, etc etc. Shipping with an Aqua look (which several desktops have icon sets for) would be merely cosmetic without a lot of work behind it.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 13, 2008 23:58 UTC (Sat) by daniel (subscriber, #3181) [Link] There are definitely issues with the file selector and many users, for instance, who were accustomed to using only the keyboard with the old file selector really loathed its initial implementations.That is putting it mildly. I loath not only its initial implementation, but every revision I have ever run into, and the unbelievably painful keyboard interface is just the start of the damage. The GTK/Gnome file open is _so_ far from usable that I seriously wonder if whoever thought it beautiful at the time can ever be anything but a liability to a usability project. Sorry if that sounds grumpy, but using that dialog made me grumpy. To be honest, I do not know who designed it and I do not care. I just want the pain to go away. And of course it is not just the open dialog.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 7:11 UTC (Fri) by njs (subscriber, #40338) [Link]
> the gnome file selector box which everyone loves (not).
I like it; probably better than any other file selector box I'm familiar with. Not that I'm a connoisseur. I can type filenames, it has tab completion, ability to sort by last modified time, and quick access to usb volumes and remote hosts, and no clutter. What would you do to make it better?
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 10:08 UTC (Fri) by Zenith (subscriber, #24899) [Link]
+1 from me, I like it also.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 10:19 UTC (Fri) by tialaramex (subscriber, #21167) [Link]
The main issue I've seen with this dialog is that it's too slow.
This is compounded by applications (ab)using it to pick things from directories like /usr/bin/ that are by definition going to be huge. In most cases the application really wants a list of "applications" (e.g. "Text editor" or "Firefox Web browser" not "/usr/bin/firefox") and then a place where advanced users can manually enter an entire command line.
One of the first and saddest things you learn in UI design is that app authors would rather use the wrong UI element and save ten minutes than use the right UI element. All those inappropriate OK/ Cancel dialogs you still see in Windows apps? That would be because causing a modal OK/ Cancel dialog in VC++ is two lines of code, whereas building a usable equivalent with verb labels and not blocking the UI takes way more work. It's a "no brainer".
Ideally you try to make the Right Thing easier, but sometimes developers are just ducking out on necessary work. Thus there is no easier way to do the Right Thing, so you need to make the Wrong Thing harder.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 19:42 UTC (Fri) by njs (subscriber, #40338) [Link]
It used to freeze up when you typed "/usr/bin" because it was grinding away loading tab completions, but that was fixed a while ago, and just typing the filename is quick. If you want to actually browse through the contents of /usr/bin then it will take a while to finish loading, but that's not really its fault, and you don't want to do that anyway.
And all the apps that I've run into recently have widgetry like "Open with [list of registered apps for this file type] / [Other]", where "Other" is what opens the file selector. (I'm sure there are exceptions, but high-profile apps seem to get this right.) It is silly that one can't type a command line, but in practice I haven't run into much of an issue; it's not like your average GUI app takes particularly interesting command line options.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 22:37 UTC (Fri) by zotz (guest, #26117) [Link]
[(e.g. "Text editor" or "Firefox Web browser" not "/usr/bin/firefox") and then a place where advanced users can manually enter an entire command line.]
Yup,
this one bugs me from time to time. Give me a place to paste in the results of which please. If it is there, I either haven't found it, or forget where it is when from time to time I need to do this.
drew
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 11:03 UTC (Fri) by epa (subscriber, #39769) [Link] A better filepicker is an improvement, but not as good as getting rid of the filepicker altogether. In my view they're a prime example of user interface cruft. There is already a file browser available (hopefully a good one, that gives you a good 'spatial' feel of 'this file goes here, and that one goes there') and it should be used to select files for opening and closing too. ROX gets this right - look for 'drag and drop saving'. It is a shame GNOME and KDE are still perpetuating the idea that each app needs to have its own little box for choosing files, which is a legacy of single-tasking days.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 11:24 UTC (Fri) by michaeljt (subscriber, #39183) [Link]
ROX looks very nice, and I've always liked the NeXt/MacOS idea of appfolders. How do they manage though, with things like file type associations if they use zero install-type installation? And things like the "You have inserted a DVD, what would you like to do with it?" dialogue box? Or in general, things where one application has to be aware of other installed applications?
Drag and Drop Posted Sep 12, 2008 13:57 UTC (Fri) by grantingram (subscriber, #18390) [Link] The problem with Drag and Drop is that it is somewhat poorly implemented, though I do remember the sheer beauty of a consistent visual interface with the old Acorn Risc OS with great fondness. For example I've just tried dragging e-mails, firefox tabs and pdfs between applications and the filemanager. The end result was I had to do a Ctrl-Alt-Backspace to get the X server unstuck... This sort of thing encourages me to keep using the file selectors. Which I agree is quite annoying at times when you are trying to save lots of documents to the same place from different programs.
Drag and Drop Posted Sep 12, 2008 14:15 UTC (Fri) by michaeljt (subscriber, #39183) [Link]
Actually, another problem with it (usability!) is that it is not accessible to people who can't use a mouse, so you need to have (at least) a keyboard fallback anyway.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Dec 17, 2008 8:43 UTC (Wed) by mfedyk (guest, #55303) [Link]
I hope it is nothing like Spacial Gnome.
If there is anything that makes Ubuntu popular it is that small tweaks have been made like not defaulting to Spacial Gnome.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 21:34 UTC (Fri) by endecotp (guest, #36428) [Link]
> What would you do to make it better?
> I can type filenames,
I had no idea that either of those features existed (and I'm not stupid, or a novice); they are totally hidden. There is nowhere visible that my typing will "go"; there is, to use the usability terminology, no affordance to suggest that typing a filename is an available option.
I would also suggest that as the filesystem is a tree, the dialog should look at least somewhat like a tree. And it always annoys me that the root of the filesystem is called "FileSystem", not "/".
I encourage everyone to read "The Humane Interface" by Jef Raskin and especially "The Psychology of Everyday Things" by Don Norman. In fact, I think they should be compulsory reading for anyone doing design.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 13, 2008 2:27 UTC (Sat) by drag (subscriber, #31333) [Link]
> I had no idea that either of those features existed (and I'm not stupid, or a novice); they are totally hidden. There is nowhere visible that my typing will "go"; there is, to use the usability terminology, no affordance to suggest that typing a filename is an available option.
Here is probably a shocker then: It works the same way in Microsoft Windows. (Just not as well)
One big advantage is that with Microsoft Windows it only does matching on the first letter. With Gnome you get a little box that appears and you can narrow down matches; which helps a lot when your dealing with 200 shares on a corporate network and they all begin with the first three letters..
I haven't tried it in OS X, but I bet it works the same way there also.
-----------------------
Another one is that Drag-n-Drop mostly works in Gnome.. as long as you stick with Gnome, or at least Gnome-aware stuff.
Ever tried to drag-n-drop a icon from Nautilus to Gnome-terminal? It's pretty clever. Even works for URLs from Epiphany.
Another shocker.
This works in OO.org 3 beta, at least. But try dragging and dropping a mp3 file from nautilus into OO.org. :)
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 13, 2008 8:29 UTC (Sat) by endecotp (guest, #36428) [Link]
> Here is probably a shocker then: It works the same way in Microsoft Windows
OK, it all makes sense then. People who are familiar with Windows will be able to use it. I don't fall into that category.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 15, 2008 1:36 UTC (Mon) by drag (subscriber, #31333) [Link]
It works better then in Windows.
Theres always lots of stuff that isn't obvious about any decently complex user interface that I've seen. It's one of the ways you can strike a balance between the needs of advanced users while meeting the needs of non-experts.
You give users every option under the sun and it'll just confuse and bewilder them.
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 13, 2008 6:05 UTC (Sat) by njs (subscriber, #40338) [Link]
Oh, right, one thing that's a little confusing is that there are actually two ways you can type filenames and make things happen -- you can start typing to pop up a "search within this list-box" thing (which all GTK+ listboxes seem to support). Not very discoverable, but very useful in general and no cost for non-expert users.
That's not what I was talking about, though. Click the little "pencil-and-paper" icon at the upper-left; this gives you a text entry box with drop-down autocomplete and all that. If you click it once, then forever after it appears in all GTK+ open dialogs (unless you toggle it again). AFAICT, anyway. I'd forgotten you have to click, because I did so a long time ago :-).
A button does not seem "totally hidden" to me; it's certainly much more discoverable than the "expert user configuration" that a lot of people whine about wanting.
I have no idea what you mean by "looking like a tree", and while I'd personally prefer to see "/" as the label on the whole-filesystem option too, I can't imagine how Jef Raskin or Don Norman would!
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 13, 2008 8:41 UTC (Sat) by endecotp (guest, #36428) [Link]
> Click the little "pencil-and-paper" icon at the upper-left
Ah! I had honestly never seen that either. I had noticed the chain of per-directory buttons along the top of the dialog (where "/" aka "File System" seems to be represented by a little white swirl with a green line above it !?) and had mentally bundled that icon in with them, I suppose.
> I have no idea what you mean by "looking like a tree"
Well, just choosing the first hit in Google image search:
The point about "/" is that it's a 30-year established name for the root, and it's too late to change it (even if we wanted to). "File System" just makes me think "but I have (at least) two filesystems, which one does it mean?".
Canonical to fund upstream Linux usability improvements (ars technica) Posted Sep 12, 2008 6:51 UTC (Fri) by petegn (guest, #847) [Link]
so that means it is going to become yet another M$ Corp windBloWs clone
|
|
||||||||||||