Home > Gnome > Gnome 2.2 User Interface Discussion Gnome 2.2 User Interface Discussion Eugenia Loli 2003-02-01 Gnome 32 Comments Chrisime, member of the Gnome project, has put up a page discussing UI shortcomings on Gnome 2.2. Update: And here is another one, this time for Mandrake 9.1b3. About The Author Eugenia Loli Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker. Follow me on Twitter @EugeniaLoli 32 Comments 2003-02-01 11:35 pm I think Havocs reply is all that needs to be said: Most of the other points are just a matter of getting out Glade and making apps conform to the HIG; anyone is welcome to work on this. 2003-02-01 11:36 pm With that font, grammar and page design, she doesnt seem to be much of an expert on user interfaces. However, she does have some interesting and thought-provoking ideas. 2003-02-01 11:40 pm No wonder Linux is failing for the Desktop market. If people who design web pages with that font, size, and style–not to mention who cannot even spell “dokumentation”–are the ones who are pushing for “reform” of the user interface . . . then I hate to see the people who actually design the user interface. On the other hand, there are many good points, but these need to be refined and clearly set out by a professional. 2003-02-01 11:43 pm Both response’s to his thread seem to be ignoring the underling statement that Chrisime made, which was that these problems should be fixed permenently or else your stuck with the never ending cycle of a developer releasing a application that does not conform to the hig, and thereby forcing the ui developers to always play catch up. This imho is what makes gnome suck. 2003-02-01 11:49 pm I like to notice that I’m the author of the UI review. I only asked chrisime to put it up somewhere for proper discussion on the Mailinglists. Only to protect chrisime here. This review was shown and discussed in the german gnome channel to make sure that it doesn’t sound like a rant or troll. I also welcome the reply from Havoc and he is no doubt right. To add one last sentence to it. My main intention was to make people pay attention for the problems that I was seeing. greets, Ali Akcaagac I hope we can have a fair discussion about this. 2003-02-01 11:50 pm Unfortunately, there was a misunderstanding about who wrote the text. I just uploaded it to my home directory at http://www.gnome.org Originally, it is from oGalaxyo and I agreed with him to put it in my home DIR after having read the review. I hope that’s enough explanation for now 🙂 Greetings, Christian (aka chrisime) 2003-02-02 12:06 am Yes, that was one of the points. The problem is also to see the project as a whole not on a per app basis. From what I know (and of course I may be wrong here) is that the two examples of the capplets I brought up are already HIG’ified but looked into detail they still look unesthetical because of how the buttons are laid out, the origin from the frames etc. These things can all be sent to bugzilla and be solved (until someone break it a couple of days later) and it will never ever be really consistent. But for now these bugs are not important. Important is to have the people understand the issues around it and see the thing on it’s whole. I am aware that many people know about this already. Even new programs that got written, soon to be changed again because the Toolbar is being removed from GLADE and Bonobo to be replaced with the libegg one in GTK+ you shouldn’t play down this issue and underestimate the changes and work needed for the applications. There are also a bunch of CORE apps that belongs to the GNOME and 5th TOE release where the author don’t like the HIG at all: http://bugzilla.gnome.org/show_bug.cgi?id=93191 My intentions are not bad ones and I’m willing to help here but how can you help that’s the question. 2003-02-02 12:22 am (quote from the article) >> i have choosen the nautilus preferences dialog with the notebook widget to demonstrate the layout of the notebook widget inside the window. please look how much space this notebook widget has before it hits the border of the window decoration. I agree with everything in this article except the text in this quote (taken from the text under the nautilus pref window (screenshot)) The layout of the notebook is good imho, because it would look very odd when removing the redundant space (it would become far too small). I think it would be a better idea to decrease the width of those option menus instead and leaving some space between the text + option menus. Have a look at this modified shot from the article to see what I mean: http://home.quicknet.nl/qn/prive/rm.vaneijbergen/nautilus-notebook…. It also solves the inconsistent look you’ll get when just wacking those option menus in there, vertically between those labels. (but although it does look better, there’s still room for improvement) 2003-02-02 12:34 am Well I these are only examples I took to compare things. The Nautilus Preferences of course is a clean example. I needed some total different looking apps. Your Nautilus modification looks good. Please don’t nail me down with the examples choosen hehehe. 2003-02-02 12:44 am keep polishing Gnome because i love it!!! 2003-02-02 12:47 am be leet use enlightenment 2003-02-02 1:12 am Parse error: parse error, expecting `’)” in /home/osnews/web/settings.php on line 10 After this is fixed. 2003-02-02 1:56 am Most of these problems are cosmetic, and although they play _some_ role, I think you’ll find that there are much bigger UI problems in Gnome than those listed here. I could sit down and handpick through every part of the Gnome UI and point out poor design left and right. I had every intention of doing this but through some freak incompatibility Gnome won’t work on this computer. Errrr. Within time I hope to complete my own analysis of the Gnome UI — although that will have to wait until I have a comp which does work with it. 2003-02-02 2:25 am There’s no need for any authors to say : “Oh oh oh, I was just expressing my opinion. I don’t want to start any flaming or cause any grief to anybody.” If you proposes or publish something, believe in it and be ready to take on accusation. 2003-02-02 2:56 am To be honest I just scrolled through quickly while taking a look at the included screenshots. All I was aware of is that the author is, rightly so, complaining about inconsistencies in Gnome’s UI. But funnily enough I found the text with all its lack of capitals and its colors highly unreadable. 2003-02-02 3:04 am Well I must admit that my english isn’t the best. After all it’s some years ago while I learned it at school. But I think it’s still good enough to express myself in the public. Regardless of this fact. I speak and know 4 different languages (german, turkish, greece and english). But explaining or phrasing sentences incorrectly with grammar mistakes doesn’t mean that I’m not right with the things I write. It makes the contents not less valuable or creditable. It doesn’t mean that by not being able to phrase sentences correctly that my capabilities and qualification to jude about these things is wrong. But since you scrolled through ‘quickly’ I must assume that you haven’t even tried to read and understand it. Greetings. 2003-02-02 3:28 am Well, I didn’t mind your english at all – the point was the Gnome UI. It is all to easy for the Linux camp to ignore their problems by picking a fault with a person who gives critique instead of saying “wow, thanks”. At the end of the day, the webpage’s purpose was to inform and it did that for me. Good job. 2003-02-02 6:13 am This whole mess started when GNOME decided to build upon the flaky X-Windows project; why can’t these people get it together enough to enforce consistency across all applications? ONE API. ONE way to do things. Anything else breeds GNOME as you see it today. It’s all good and well that there are fifteen window managers at any given time to choose from, it’s fantastic that choice is plentiful on a free OS, but there comes a time when the bazaar becomes rude anarchy. Play to your strengths, people, and please stop inventing new solutions to mitigate the side-effects of kludges on top of old workarounds – fix the root of it. Forget about hand-rolled widget sets and custom libraries for common functions. Standardize your tool kit. Stabilize API’s. I think we’ve seen enough anarchy in GNOME to last a lifetime.. now is the time for a finished product, not more of the same bloated crap GNOME’s been peddling lately. Oh, and maybe it’s time to buy that spell-checker from Sun. OpenOffice.org leaves much to be desired. 2003-02-02 6:19 am I think the focus is entirely in the wrong place, as I don’t see 1 pixel row of white space in the toolbars to be one of Linux’s most nagging problems in the way of usability. if you use the gimp and zoom this picture to 4:1 then look and compare the toolbars, they all look differently … As one other poster pointed out, I think it would take about 5 seconds to find other problems that are 100x more serious than this. 2003-02-02 7:14 am It might take 5 seconds to find other problems more serious than this, and it might take only five seconds to fix the problems as pointed out in the article – but the whole point is that IT”S NEVER GOING TO HAPPEN. This stuff won’t get fixed. GNOME isn’t exactly a professional piece of software, and the proof is on your screen.. watch for the same problems in version 3. 2003-02-02 10:15 am “It might take 5 seconds to find other problems more serious than this, and it might take only five seconds to fix the problems as pointed out in the article – but the whole point is that IT”S NEVER GOING TO HAPPEN. This stuff won’t get fixed. GNOME isn’t exactly a professional piece of software, and the proof is on your screen.. watch for the same problems in version 3.” And why not? Don’t you know how to program? Oh, and if you’re going to start on “Oh those OSS people. Always telling people they have to program this, and program that.” remember problems don’t get solved by talk, but by action. The thing with OSS is the “action” is no longer limited to a vendors whims, but is in your hands as well. Welcome to the new world. 2003-02-02 10:33 am Well, if it’s true you can find huge usability problems in Gnome in 5 seconds, you should file them in bugzilla so they are taken care of for the next HIG revision. P.S.: remember we refer to Gnome usability problems, not Linux’s. They are out of the scope of Gnome 2003-02-02 10:47 am “but the whole point is that IT”S NEVER GOING TO HAPPEN. This stuff won’t get fixed.” WRONG! I just sent in a patch to fix the 1 extra pixel Glade toolbar problem! :p 2003-02-02 11:23 am Face it! Common developers are not (smart enough|too lazy|etc.) to follow any HIG. Enforce the HIG into the libraries. Then when someone uses a gnome_dialog all margins and standard buttons are right from the start. No need to handcode this. It’s a waste of developer time. Of course not all of the HIG can be done in the libs, but many things (like margins for example) can. This is one field where I think KDE is more convinient from a developers point of view. 2003-02-02 12:27 pm There are some errors in the ads here on osnews… No comment Microsoft OLE DB Provider for ODBC Drivers error ‘80040e14’ [Microsoft][ODBC SQL Server Driver][SQL Server]Could not allocate space for object ‘ExternalArticles’ in database ‘Prototype’ because the ‘PRIMARY’ filegroup is full. /system/includes/adserverutils.asp, line 59 2003-02-02 3:24 pm No, this is not so easy. I hear this ‘file it to bugzilla’ all the time. As long as the developers are not open for these things as long nothing will happens. There are 3 situations in GNOME: – Apps that got written already using all kind of Methods such as using Glade, Bonobo, Gnome-App. These apps are ready already, work so far but their GUI looks unestehtical. Solving all this stuff will take huge amount of time and manpower. There is no need for 10 individual people that sent in 20 different looking patches again. HIG is the HIG and people should follow it as is regardless of the fact that some likes some stuff more and the others less. There is no need to have patches from 10 individuals where one follows the HIG exactly, the other nealry exactly and the last one more HIG conform. What Gnome needs are 10-20 people that know what they are doing, that know how an interface has to look like and who play together completely. Fixing all that stuff will take ages, not to mention that there are a lot of applications under the hood of a GNOME release that either belongs in GNOME or in 5TH TOE, where the author doesn’t like the HIG at all. The reply “I don’t HIG and I therefore don’t care” is a blocker and these apps will probably never get a unified look over the rest of GNOME. – Then there are apps that got started to be written already by third party members and when we count them all together we have around 150 internal GNOME modules (apps and libraries) and another 200 GNOME applications written by third party members. Many of these apps people used all the sorts and different types of programs to create them with even GLADE. Sending in a patch NOW for GLADE may fix the program now but it won’t fix the programs that where using already badly generated *.glade files and even worse some applications have the *.c generated code embedded in them. This is all hard to fix and it may be harder in several situations to get the people interested for the HIG. – Then there are applications that are planned by people and soon people start to program them. But what should they user ? There are no programming manuals nor are there any guidelines telling them what to avoid and what to use and what soon will happen. I mean most of this knowledge is knowledge that only depend to the 10 developers that regulary hacking in GNOME and GTK+ but the outstanding people don’t know it. There is no guidelines and nothing that suggests them what to do and how to do it and what they should care for. developer.gnome.org is horribly outdated and this yet calls again for correct manuals so people know. Summarized up. Personally I don’t want to talk about the contents of the HIG since this is not my job and I’m not related or responsible for it in any ways. I think that the 3-4 people working on it are making an almost good and necessary job. I’m not a friend of all stuff written inside it no doubt but it has been decided and on the long run makes sense. But please remove these ‘the size can be around 10 pixels or between 5-10 pixels’ that I (if I recall correctly) found inside it. It has to be ‘5 pixels no pixel more or less’ or ‘it has to be 10 pixels no pixel more or less’. And there is no need for having 10-20 patches from individuals, as said above, where the one follows the HIG perfectly, the other almost HIG’ified and the last one ‘more HIG conform’. It would be better to get a team of professional GUI reviewers together that consists of 10-20 people that got exact guidelines what to do. And some stuff is really hardly to change because it requirs codechange on the long run. Some UI glitches there and here always happens that’s no doubt but right now in GNOME it’s more or less a real urgent. The GIMP is another example it’s not a GNOME app but it belongs to 5th TOE. My intention was more or less to inform the people. Getting the people to see the problem as a whole. Hopefully you understand the situation now. As more apps there is as more “if not impossible” to fix them all. 2003-02-02 3:26 pm That is a good point. I remember when KDE 3.0 RPM’s came out for MDK 8.2 the first time wizard would come up every time the user logged in. All you had to do to dissable it was add one line to a config, something like “first login =false” and the package bundled with mandrake was not fixed till something like the 3rd or 4th 9.0 beta which used KDE 3.0.3 2003-02-02 6:52 pm I would like to inform that I have ‘massively’ reviewed the Text and updated it some mins ago. I like to let you know this in case you are interested. 2003-02-02 9:48 pm It’s a brave person that stands up against the might of the OSS development community and dares to criticise. IMHO: those who think these are minor points beware, gnome has suffered a lot of bad press over the years, and like most linux stuff has been desperately un-userfriendly and inconsistent. The Gimp being a great example: a superbly powerful application bogged down in an impentrable UI that doesn’t conform to any UI guideline anywhere, making the already steep learning curve of such a complex app nearly vertical. Gnome, and indeed all of the OSS need to address these UI issues. Consistence and usability and GUIs are not dirty words they are they key to acceptance for the billions of windoze users and newbie computer users out there. 2003-02-02 11:36 pm neatean, <pat>back</pat>. I totally concur. Too often those with some amount of UI design knowledge who critique free software UI’s hear “don’t criticize that which you’re getting for free”. I often say that Bill Gates doesn’t have to lift a finger to crush linux on the desktop because there are so many people in the free software community doing his job for him. As for gimp as an example of bad UI: One of gimp’s myriad UI problems, pencil-thin buttons that stretch so far across their parent widgets that you can’t tell they are buttons (and due to their one-way largeness don’t often any benefts from Fitts’ Law) is a problem that would only require 10-15 lines of code max per instance to solve. There is really no excuse for this design, yet it continues to this day. Why? Kernel bugs seem to get patched in hours, yet this simple, easy-to-fix problem still perplexes free software developers after all these years. Maybe in a decade or two we’ll have a fix. To those who think the points are minor, remember that effect of a bunch of little annoyances and problems are cumulative, and put together can add up to a really nasty user experience. And I’ll add that the reactions to Ali’s critique on the GNOME list shows the double standard of Free software: nitpicking over technical stuff like how many extra cycles some kernel modification saves is lauded, nitpicking over the design of the end-user experience is severely chastised. A developer community that doesn’t value the end-user experience has lost the battle for the desktop before it ever started fighting it (not that this will stop them from whining about how governments and schools are slow to adopt their half-assed software). 2003-02-03 2:55 am “And I’ll add that the reactions to Ali’s critique on the GNOME list shows the double standard of Free software: nitpicking over technical stuff like how many extra cycles some kernel modification saves is lauded, nitpicking over the design of the end-user experience is severely chastised. A developer community that doesn’t value the end-user experience has lost the battle for the desktop before it ever started fighting it (not that this will stop them from whining about how governments and schools are slow to adopt their half-assed software).” There’s no double standard. Good and bad ideas show up on the list. Leaving aside personal opinions of “good-bad”. Good ideas (ideas that survive the critique process) get in. Bad ideas get left behind like they should. I assume everyone is familiar with the peer-review process, and the present level not only of scientific progress we enjoy, but the offshoots? The thing I think a LOT of people forget (especially those who are only familiar with the “old” way) is that this equation holds. Developer=User,User=Developer. The new way requires that a UI designer wear more than just one hat, and can carry the Developer hat in hand. Waiting for someone else to do something for us is the “old” way (the proprietary way). I’ll leave the history lesson for others. A new paradigm always takes awhile for people to get use to. 2003-02-03 4:32 am I happen to be one of those developers writing an app and oGALAXYo pointed this out to me in an irc channel a couple of days ago, i’ve now stopped coding the app because I don’t know what libs to use to get a proper toolbar or a toolbar that conforms to the HIG. Basically it doesn’t make any sense for me to follow the HIG if the tools don’t even conform to HIG (ie: GLADE) secondly I want to write my program and really don’t want to have to bother reading the HIG if I don’t know what libs the proper HIG stuff is in. Sure I could spend time sitting there writing GUI code but if there is a better way I’d like to use it. I’ve already set the GUI up using Glade and it was fast and easy BUT it doesn’t conform to HIG and the toolbars look ugly compared to the ones in libegg; etc etc. You get the idea, the solution to this problem is to have ONE lib that holds all of the proper HIG stuff and PROPER documentation. The changing of the libs and putting diff toolbars in diff libs is annoying and I don’t follow core or cvs enough to know whats going on and whats where. There needs to be a team put into place that cleans the mess up otherwise this whole usability push is senseless and all it does is make programmers who want to write a quick app, annoyed. I’m willing to code the gui but now there is no documentation on what to use which leaves me confused.