Submitted by “Theme development is a tedious and difficult task, and for the GTK devs to be so careless in breaking their API at every turn disrespects the many hours people put into making themes for it. […] I was given to believe that this breakage stems from a Microsoft-like climate of preventing users from customizing their systems, and deliberately breaking the work of others so that your ‘brand’ is the best. Anytime I hear the word ‘brand’ being used in Linux, I know something valuable is being poisoned.” I find the tone of this one a bit too harsh and overly negative at times, but his point still stands.
That doesn’t sound very Microsoft like, while not quite as customizable as Linux, Windows has always had a fair share of options. No, Gnome has always worshipped Apple, and are following the user-hostile approach of their idol, but without the talent to make hard-coded defaults any good.
Gnome has always worshipped Apple and Mac OS, but they’ve always completely, totally and fundamentally misunderstood the world that Mac OS operates in. It operates in a world of straight-jacket hardware and options such as theming because it just isn’t intended to be Windows, isn’t intended to run on a wide variety of hardware and doesn’t have the extremely wide range of software written for it. Even then, amusingly, Mac OS is far more configurable than Gnome is.
Gnome is fundamentally at odds with those two things – yes it does need to run on a variety of platforms for its userbase (it’s supposed to be a ‘Unix’ desktop for fuck’s sake) and it desperately needs applications and developers to move forwards. Put those two things together and you have why Gnome is dying in its own vomit.
What’s more funny is how the Apple worship and general ‘usability’ direction came about. It didn’t come about because Gnome’s developers genuinely believed that it was the right direction to go. It was because the functionality that they needed to write to compete with KDE, Windows, and Mac OS for that matter, was simply too difficult to do. Gnome is still the same hashed together, poorly thought out technical mess loosely based around Gimp’s toolkit that it was in the mid-nineties when it was thrown together as a response to KDE.
They then dressed up cutting features and configuration you would find in all those other desktops as being for reasons of ‘usability’ to hide these uncomfortable facts. In fact, I don’t think Gnome’s developers have even consciously thought about this. It just seems to be an endemic mental disease in the project. People like Jeff Waugh over the years have exemplified all this, and the whole notion of a Gnome ‘brand’ – whatever that is.
My thoughts exactly.
“We^aEURTMve fought long and hard to give GNOME 3 a consistent visual appearance, to make it synonymous with a single user experience and to ensure that that experience is of a consistently high quality. A general purpose extensions and themes distribution system seems to threaten much of that.”
that is pure apple.
It seems rather odd to be complaining so strongly about gnome, when Qt exists, is much more stable and well supported, is easier to code for, has better documentation, and much better cross-platform support.
I’ve been using only Ubuntu for a while now, so I know not if fedora etc have made their qt themes look and behave nicely alongside their gtk themes, but in unity and in KDE, Qt looks beautiful.
It would surely be easier to improve Qt themeing as canonical have, than to keep complaining about gtk, because we all know that gnome do things their own way, and don’t really care very much about anyone else.
Edited 2012-11-08 21:30 UTC
Unfortunately, switching to Qt theming is a lot more difficult than it sounds.
There used to be a GTK+ theme which delegated the work to Qt but, as far as I can tell, the last commit to it seems to have been back in 2009. (Around the same time Qt gained an officially supported, bundled theme which did the reverse, theming Qt using GTK+ themes)
As for switching the apps themselves over, I can tell you, as a programmer, that’s not a very viable option. GTK+ and Qt aren’t just widget toolkits. They both provide their own approaches to a LOT of low-level functionality that neither C nor C++ offer as core features. (The event loop, signals, network-transparent file I/O, reference-counted object system, etc.)
Finally, just finding replacements isn’t a viable option since any given user who’s grown comfortable with an application will probably see all the competitors as significantly inferior… if for no other reason, due to their design philosophies not lining up with the user’s comfort zone.
I used to be a happy KDE 3.5 user but I always ran a mix of Qt and GTK+ applications. Since KDE 4 came out and I switched to LXDE, I’ve just moved more and more to the GTK+ side of things.
For example:
– Konqueror 4 is apparently bitrotting except for minor fixes by KHTML devs.
– Without KParts, Dolphin is inferior to PCManFM.
– Nothing else comes close to Pidgin or Kopete.
– Kopete is under-maintained, buggy, and lacks IRC.
– No other Qt or GTK+ app is as responsive as Geeqie (formerly GQView).
– etc. etc. etc.
…not to mention how Qt 4’s movement threshold for clicking and dragging is so small (compared to Qt 3 and GTK+) that, in Konqueror or Dolphin, I make accidental copies more often than I open files. (I have a Logitech G3 mouse and a big desktop)
…or how, being used to LXDE with lightweight GTK+ apps, I can actually notice and get annoyed by how the closest competing Qt4 apps tend to be a few hundred milliseconds more sluggish on my AthlonII X2 270 with 16GiB of RAM.
Edited 2012-11-08 22:39 UTC
I was mostly referring to the development of new applications, to be honest.
gtk seems much more developer-hostile, but most Linux applications seem to use it.
I agree that many Qt applications are bloated, but that is not something inherent to the toolkit, rather being a tendency of certain types of developers.
Looking at the (new, and not yet very stable) razor-qt desktop environment, fast Qt apps and environments are definitely possible.
I am yet to find a file manager and terminal that I like more than dolphin and konsole. It’s just a shame that kde itself is not as nice as unity.
KDE telepathy is moving towards being a good chat client, but for now, the best option is still pidgin, unfortunately.
I can’t say I’ve experienced the clicking threshold problems you describe, but considering I’m mostly a keyboard user, I am probably not able to judge.
Edited 2012-11-08 22:56 UTC
Maybe I’m just not using it right, but, last time I tried PyQt, I found PyGTK to generally be more comfortable. (Not to mention more favorably licensed and with a more sane reaction to receiving Ctrl+C)
Also, I’ve started playing around with Vala and it seems to be harder to shoot yourself in the foot with it than C++.
(If you’re not familiar with it, it’s a language that tries to layer a more Java/C#-style syntax on top of GObject and compiles to pure C. It’s also been designed to be more friendly to being called by pure C code than C++.)
Ever since Yakuake’s KDE4 port made it too heavy, I’ve grown quite attached to my urxvt setup. (I use the bundled “kuake” plugin with GNU screen for tabs)
As for Dolphin, I find it a GNOME-reminiscent step backward from Konqueror 3 in various respects. As for PCManFM, the only things I’ve seen Dolphin do better are the filter bar and the split/unsplit toggle button… and the former is currently the new feature in development for PCManFM.
I actually think Plasma+Kwin is nicer than Unity+Compiz… but heavier than LXPanel+Openbox and I can’t find where that Plasma equivalent to the (non-negotiable) LXPanel “Directory Menu” widget went since I last used it in KDE 4.4 or 4.5.
The official python API is PySide, not PyQt. That is the one actually developed by digia.
Regarding vala, it seems like a really cool idea, but is very immature and has a fair distance to go before it reaches widely useable potential.
I actually wanted to play around with it, until I realised it was basically just for gnome. I’ve only had bad experiences with gnome, so…
Dolphin does less than konqueror and kommander, but it does have a much more clean and pleasing UI, whilst retaining features like pane splitting, kparts and integrated terminal. Also good is that they have all of the panels and toolbars still fully customizable and removable.
Nautilus has been okay, but pcmanfm/thunar just don’t look that good, and the reason we’re using file managers instead of just a terminal is for the GUI, right?
It could well be that I have no problems because I’m running a Phenom II 965 quad core, and intending to upgrade to haswell next year, but even on my athlon, the themeing of KDE was something that kept me there, as opposed to the lighter desktops, despite the minor performance regressions. I know you clearly favour performance, but I like my working environment to feel enjoyable, which for me means a consistent and modern appearance.
I can’t help you with that directory panel, unless it is folder view, or the classic start menu, which are still bundled with default KDE as plasmoids.
It’s been a while. Last time I actively tried writing Qt applications in Python, PySide didn’t exist and, last I considered trying again, PySide didn’t have a Windows release.
I’ll have to give it another try at some point… assuming it’s compatible with something like Py2Exe.
What made you conclude it’s just for GNOME?
I’ve had no problem playing around with GNOME-free GTK+ programming in Vala and, while it means you can’t do object-oriented stuff, it does have an experimental mode which kills the Glib dependency so you can compile to more pure C.
That depends on your definition of clean and pleasing.
1. I find the variable icon sizing and blur-inducing animations for showing and hiding disk usage bars in the places pane to be irritating despite preferring KDE’s support for application-specific place bookmarks.
2. I never use the icon size slider in the statusbar and much prefer how PCManFM puts a free space counter there.
3. I have no clue what you mean by Dolphin supporting KParts. It still seems to be firmly designed to only use the DolphinPart KPart because the author doesn’t want people to realize that he’s reinvented Konqueror.
(He’s of the belief that a file manager should not also be a browser and, to protect that viewpoint, he refuses to implement “preview using KParts” because Dolphin already uses KIOSlaves and all it takes for a primitive browser is KIOSlaves and KHTMLPart)
I don’t understand. How do you define “look that good”?
http://i.imgur.com/a0AW9.png (comparison)
To me, they look pretty much the same aside from KDE applications using some icons not present in the Elementary theme I’ve grown fond of.
I do agree that Thunar is undesirable… but that’s because the Xfce devs are adamant that Thunar will never have tabs.
I think we might have different definitions of “modern”. I find Oxygen to be unpleasant and consider quite a few aspects of “modern” theming to be counter-productive glitz.
My (currently Lubuntu-based) LXDE desktop may not follow current trends, but it feels pretty darn modern and consistent to me.
No, it’s sort of a hybrid of the two. It’s like the classic start menu except, instead of a tree of categories and application launchers, the menu allows quick navigation of the filesystem.
(The LXDE one just shows folders for quickly opening a PCManFM instance while the KDE one shows everything for quick launching)
I have one for my home directory and one for my media library.
Edited 2012-11-09 00:19 UTC
You can use PySide with pretty much all of the exe freezing options out there now, and it has better support for Qt and all of its platforms than PyQt ATM.
My bad, I was referring to gObject as being essentially a gnome project, because their developers have such a level of control over it that they basically own it.
I leave it with bookmarks, but I must agree that those usage bars are annoying.
You can turn off the zoom slider and turn on a usage text item, this is KDE after all.
I mistook KIOSlaves for kparts. I was referring to handy things like easy disc conversion, good FTP support and their ilk.
I should have noted that I meant their appearance in kde and in unity. This includes the feel of the use, which seems very dated in most GTK2 apps, although it has been some time since I tried pcmanfm.
I agree that oxygen needs a refresh, but I have been using unity, which looks just as good in Qt with unity 2D as it does in compiz.
The general feel of Qt is sleeker though, as an example razor-qt vs lxde. In my opinion, razor-qt feels slicker despite being buggier.
That perception is hard to pin down, but I think it is due to the squared panels with gradients that pop as opposed to smooth and subdued panels.
It would be the same reason unity and kde have nicer looking window decorations than openbox.
I understand what you mean, and have definitely seen it before, but unfortunately, I am totally unable to name it off of the top of my head.
Also, as an aside, there is this for looking at Qt apps. It is only relatively young, but shows some promise.
http://www.qtdesktop.org/applist
Good to know. I’ll have to give it a try and see if I can duplicate the kind of responsiveness Geeqie gets out of GdkPixBuf in Qt’s equivalent classes.
(I have a project planned that needs a highly-responsive image tagging GUI and Geeqie is my gold standard for responsiveness while an image hasn’t finished loading.)
Fair enough. I’ve actually been meaning to track down a widget-agnostic event loop and signals system so I can move all the GTK+ dependencies in my creations into UI plugins for easy replacement with Qt.
I’ll admit I didn’t go looking as thoroughly as I should have.
Was that added since in KDE 4.4 and 4.5 or am I just too tired to be competent?
Agreed. From KDE 3.5 to present, I’ve always kept Konqueror around as my FTP/SFTP/FISH/etc. client regardless of what I do with other things.
Lately, though, I haven’t been using any kind of GUI client. I just use rsync. (Pull it here, edit, push… and I get a free extra backup out of it.)
I’ll admit PCManFM is a tiny bit dated in not yet supporting a breadcrumb trail address bar or showing the number of children in the size column for folders.
That aside, I’m not really sure what you mean. Maybe I just value performance too much to get comfortable enough to notice any progress that has been made.
You are aware, I hope, that Unity 2D is being end-of-lifed? Last I heard, Canonical was planning to replace it with regular Unity using software-rendered compositing in either 13.04 or 13.10.
From the screenshots, I can see what you’d mean by looking slicker for the panels and popup notifications.
However, in some ways, my LXDE desktop is actually slicker. Most of what I see would become almost identical once I squashed those panels down to my desired height and turned off compositing and fake translucency to squeeze out more CPU cycles.
(I’m using panel graphics from a previous Lubuntu release that make LXDE’s panels look a LOT like the dark Razor ones and I’ve customized my tray icons so they all look like monochrome carved depressions in the panel unless they’re trying to get my attention)
I actually don’t like those seamless window decorations. Lubuntu does offer some but I turn them off because I like the active region for the WM to be clearly distinct from the active region for the client.
(Not to mention how the default window borders don’t offer enough contrast between active and inactive windows)
I’ll take a look at it but this isn’t a case of “He doesn’t know where to look”. This is a case analogous to “He’s running Vim because, with fifty-million other editors having a Vi-like mode, none of them cloned it well enough to satisfy.”
Making a satisfactory knock-off of a program is very hard and, for most of my GTK+ apps, nobody has even come close to mimicking all the little features that, to other people, probably seem like pointless wastes of effort.
Don’t quote me on this, but I believe it has been there since 4.3 or so. I may well be mistaken, however.
Indeed. often, the GUI just interferes.
I think it is mostly 3D shiny gradients, squared raised elements etc that annoy me. That being said, I like android’s holo themeing, and think that the iPhone looks cheap and dated in comparison, due to things like the 3D bubbles in messaging, etc. Very 2000s style kitsch.
All too aware. They actually suggest LXDE as a replacement for those who can not run regular unity.
Your customizations sound great, they sound as though they address most of my issues.
Differing usage, I suppose.I mostly hold down alt and drag from anywhere in the window when I want to move or resize.
I also use mostly full or tiled windows, without much stacking. I’d use a TWM if any had the HUD and dash.
Ah. I can relate. Since getting used to vim in the terminal only 8 months or so ago, I am addicted.
I’m not sure I’d use it on a non-touchscreen device but I definitely agree that Holo is probably the most elegant look and feel for touch in existence.
…and that 3D bubbles were NEVER stylish.
I’m a bit overdue for bed but I’ll see if I can find time tomorrow morning to pull together some screenshots and links to the resources I’m using.
It depends. I don’t always have my left hand free to hold Alt.
I have actually been wanting to set up a hybrid tiling-floating WM but I haven’t had time to program AwesomeWM or to learn enough Haskell to make Bluetile less GNOMEish.
(I have two kinds of windows. The stuff like Firefox which takes up the entirety of one of my two 1280×1024 monitors and the stuff like MPlayer which tends to be floated over the least important region of one of my applications.)
*nod* Originally, I chose it because it was the least of several evils. (Eric3 was too distracting, Eclipse was too distracting and heavy, SciTE had no command line, GTK+ Emacs felt even less native than gVim and had more alien keybindings, Gedit and nano weren’t programmable, etc.)
However, now, I can’t imagine programming without it. It’s light, fast, there are a ton of useful plugins, and no other editor I know of lets me type “:gui” to release my terminal for other uses.
Speaking of which, if you ever want to check out my vimrc or plugin load-out, they’re up on GitHub and .vimrc contains a reference for most of the keybinds I actually use:
https://github.com/ssokolow/profile/blob/master/home/
I would also appreciate it, especially the tray icons stuff.
That would be greatly appreciated, thanks.
Thanks again. I’d share mine in return, but it is very basic due to the length of time I have been using vim.
Here’s what my taskbar looks like.
http://i.imgur.com/sZMPH.png
The upper one is basically its default state though I believe, if the taskbar were actually that short, lxpanel wouldn’t be wasting so much space inside each task button. (I screenshotted my 2560×1024 desktop and clipped out the middle of the taskbar to shrink it)
The lower one is a composite image showing all the on-hover and tray icon notification states. (Except for Deluge, which uses blinking rather than an alternate icon to grab your attention.)
The chat icon is Pidgin and mail icon is Thunderbird with the FireTray extension.
It’s all accomplished either by using the LXPanel config dialogs or by putting files in ~/.local/share and ~/.trinity (KDE 3) to override or extend the system Elementary icon theme.
https://github.com/ssokolow/profile/tree/master/home/.local/share/im…
https://github.com/ssokolow/profile/tree/master/home/.local/share/ic…
https://github.com/ssokolow/profile/tree/master/home/.trinity/share/…
For the most part, the normal states are Elementary panel icons meant for a light panel background while the notification ones are their dark background counterparts, but there are a few tweaked Faenza panel icons in there too.
I’ve blogged about the rationale for this approach (back before I polished it up this much) if you want to see more about how I use it: (For example, Audacious’s tray icon doubles as a volume knob via the scroll wheel)
http://blog.ssokolow.com/archives/2012/02/15/polishing-up-the-lubun…
I’ve also blogged about the (imperfect) process I used to adjust the Faenza icons if you find yourself needing more:
http://blog.ssokolow.com/archives/2012/02/14/converting-faenza-pane…
That actually looks great. I would nix the taskbar obviously, as I use alt-tabbing and desktops, but other than that, much better than I thought the LXPanel could look.
Also, I was reading up on the KDE release plan, and 4.10 (out in January) has a folder drawer widget, which would seem to be exactly what you wanted.
Makes sense. You might have noticed I don’t have a desktop pager. That’s because I have a Logitech G3 mouse and bound the fourth and fifth buttons to previous and next desktop.
I find that, with pixmap-themable UIs, what often happens is that the developers build a theming system with a lot of power but then don’t know anyone with the artistic skill to properly show it off.
The default LXPanel look and feel is definitely dated but, as those screenshots show, all it took was some better skinning (from the Lubuntu guys) and a little tweaking the tray icons for better consistency (me) to get a really nice-looking panel.
(I was actually the one who got the ball rolling for themability in FireTray. I didn’t have the time to learn the codebase and provide a patch, but I did file the bug and include a detailed explanation of how to use GTK+’s icon lookup API)
Nice. Do they give enough detail to know whether you’ll be able to configure it to interpret subfolders as submenus? (Because I often go three or four layers in before I choose to open the file manager)
Edited 2012-11-10 00:46 UTC
Can’t say I did; I barely touch the mouse, so I don’t use pagers when there are easy key combinations for swapping and moving.
Definitely true. FOSS seems to have devs galore, but not many people working on documentation, and even fewer on art.
Unfortunately not, but given that the folder-style one allows it, it would not be unreasonable to imagine so, considering it may well be a remake of that, with list instead of icon view.
menu thing:
right click on the KMenu and click “Switch to classic”
it still have a few more items, but it is very close
You obviously didn’t read what I said.
“Switch to classic” won’t give me a menu that lets me quickly navigate the contents of my /srv/fservroot.
That’s a different plasmoid.
I’m sure KDE and Qt developers could do lightweight apps to shave some imperceptible time off for you, but then they would take longer to develop and wouldn’t have any features anyone using a desktop would care about. LXDE is basically just a window manager. If it wasn’t more people would be using it, but they aren’t.
This kind of brain damage is what has killed the Linux desktop. I can only laugh in amusement and despair that in all the years I have seen Linux desktop articles on this site that these comments are still prevalent.
You obviously don’t know much about LXDE. I suggest examining the categorized (core, accessories, system, configuration) list of components at http://wiki.lxde.org/en/Main_Page before making that claim.
…and even if the combination of Plasma, KWin, Dolphin, and friends didn’t induce noticeable slow-downs that doesn’t change the fact that the maintainership of GTK+ is the problem and the UI design philosophy for things like Amarok and Gwenview does an inferior job of meeting my goals compared to applications like Audacious and Geeqie which happen to be written using GTK+ (and which also happen to be more lightweight).
Edited 2012-11-09 17:36 UTC
I’m afraid I do – and XFCE. It looks and feels like what it is – a mid-to-late nineties post-CDE, barely-better-than-a-window-manager, not-quite-a-desktop that even a small minority of Linux desktop users use. Now that is small. Ditto XFCE.
There’s a reason they are used by a minority and only by those who believe they can see miniscule time delays on extremely modern hardware and believe that if they run it on modern hardware it will be blazzzzing fast. Or something.
People are not going to trade features and functionality for your perception of ‘slowdowns’ I’m afraid.
Many Linux desktop people, and especially those around Gnome and GTK, take the 80/20 rule as to why you just don’t need all those features and why you should choose something ‘lightweight’. Alas, the problem is it is never the same 20% of features that people use.
They don’t even define lightweight either, but they don’t tell you what it is – software that looks like arse and does less.
What I find ironic is that you clearly love ‘lightweight software’ but GTK itself has chosen to be more and more ‘lightweight’ – and you don’t like that.
That’s good – for you.
Ever see any software companies marketing ‘lightweight’ software, ‘lightweight’ word processors or ‘lightweight web browsers? No? That’s because there isn’t a market for software with less features and with the tagline ‘lightweight’. You just don’t see it.
Reviews of such software generally go along the lines of “Well, it seemed to load up fast but it just doesn’t have this feature that I need. If it implements that then I might use it”. Stories like that are a penny a dozen and are as old as graphical user interfaces themselves.
In other words, you’re redefining “barely more than a window manager” to include any desktop which doesn’t provide its own re-invention of every application users commonly want? Nice way to move the goalposts.
I’m honestly curious how you reconcile the fact that so many people pretend Epiphany and GOffice don’t exist and use Firefox and LibreOffice instead.
I’m willing to consider LXDE “minimalist” compared to GNOME and KDE but I think we’ll have to agree to disagree on what is barely more than a WM… especially since I strongly agree with the LXDE view that good applications rely on standards to be featureful without tying themselves too strongly to one desktop.
GQView isn’t “part of” any desktop but, for my needs, it’s superior to every DE-provided image viewer in existence.
Balkanizing desktop components just works to kill off one of Linux’s greatest strengths.
It’s very obvious that you have an impression of who I am and you’re just twisting my words to fit it.
There are several reasons I run LXDE:
1. most importantly, the only reason I have a shiny, fast computer is because my motherboard and video card failed earlier this year and I got a bundle deal on a new mobo, CPU. and memory.
If my time hadn’t been more valuable than the money I paid, I’d probably have sourced a replacement mobo, 1GiB DDR2 DIMM and some capacitors and still be willingly remaining on an Athlon64 X2 5000+ with 4GiB of RAM (I use VirtualBox) and a GeForce 7600GS… both bought at least five years ago when they were state of the art.
When I get better hardware, I expect to be able to do more with it, not do the same things with a few new animations and fewer features because the developers lack discipline. (Hence why I actually test my desktop on an old 2Ghz Celeron every 6 months and make any adjustments necessary to keep the system requirements down)
I think the only upgrade which I actually considered warranted was the GPU upgrade and only because the old GeForce 7600GS couldn’t use VDPAU to allow the Athlon64 X2 5000+ to play 1080p video.
2. with Trinity (the KDE 3 fork) being apparently too under-manned to implement things like the XDG Icon Naming spec, PCManFM is the most comfortable file manager I find practical. (I’ve already mentioned that Konqueror 4 gained some issues in the process of being ported to Qt 4)
3. Plasma actually cannot do certain things I liked in Kicker. (eg. Having one big activity span both monitors so my taskbar widget can span smoothly across.) Why should I run a heavier desktop when I don’t use the features it adds and do want a feature it removed?
4. Aside from the Aero Snap clone, KWin 4 doesn’t offer me any advantages over Openbox and I’d already written and grown used to a basic WinSplit Revolution clone by the time they added it.
5. I use Vim for my programmer’s text editor and LibreOffice for my office suite. Kate/GEdit, KDevelop/Geany, and KOffice/Calligra/AbiWord/Gnumeric are not applicable.
6. Vim loads faster than KWrite or GEdit, gives me more of the functionality I desire, and, when I don’t need syntax highlighting, Leafpad loads in under 500ms.
7. I play my videos in bare MPlayer with no GUI beyond the playback surface. Why waste space and CPU on GUI widgets when I do everything with keybinds anyway?
8. Beyond that, my motto is “right tool for the job, no matter where it comes from”. I run K3b because it’s the best combination of stability and comfort, Filelight or ncdu depending on what I want since Baobab doesn’t know the meaning of “cached results”, etc. etc. etc.
That depends. My mother runs LXDE because she never saw the point in GUIs like Unity and my brother just switched to LXDE in the last few months for reasons similar to mine. (It does everything he needs of it… more quickly)
Exactly why I couldn’t even stand GNOME 2.x. Their 20% of features people “actually need” are very different from mine.
Ok, now that’s just an attempt at provocation.
I’ll admit Vim can look ugly at times, but everything else I use looks great and is, if anything, MORE featureful than the DE-bundled stuff.
For example, I challenge you to find ANY Linux media player that’s still being maintained that’ll play my esoteric chiptune and Amiga module formats as well as Audacious.
…or Geeqie. Sure, things like Gwenview look very fancy, but try quickly spinning your scroll wheel to skim through a bunch of images until you find the one you want. I’ve never found anything with such well-written asynchronous image loading.
While you’re spinning away, Geeqie may or may not be able to finish loading the images, but you’re guaranteed to at least see a glimpse of the top of each image.
Gwenview fakes that kind of performance by skipping images if your scroll events start to pile up.
I could go on all day.
Don’t conflate being lightweight with being unnecessarily rigid.
As I said before, I’m not satisfied with my desktop look and feel being limited to “big fat widgets: light” or “big fat widgets: dark” just because the GNOME devs think everyone either wants a touchscreen or has money to burn on ever bigger monitors.
I can personally vouch for the fact that both GTK+ 1.x and Qt 3.x are capable of high-performance theming and GTK+ 2.x had no problem offering a stable API, so whoever’s developing GTK+ 3.x is either irresponsible or incompetent.
Given that you leaped on my mention of the word “lightweight” and brushed off my comment about them fitting my needs better, I get the impression you misinterpreted my phrasing.
When I said that Geeqie and Audacious fit my needs best, that was the primary point. The point in parentheses was simply that tools manage to exist which do that while also being lighter than GNOME or KDE offerings.
Do you see me as some kind of threat to your preferences or something? Because all I’m trying to do is argue that, since GNOME can’t possibly please everyone, the diversity fostered by stable GTK+ APIs is a good thing.
I’m perfectly happy with the tools I have. All I want is for them to stay alive, well, and part of a healthy ecosystem.
I used to think I wanted one large panel that spanned multiple monitors so there would only be one taskbar. Then I actually looked at the settings for the taskbar plasmoid and noticed it featured an “only show icons for tasks on this monitor” checkbox.
Now, I live quite nicely with separate panels on each monitor, with separate taskbar plasmoids on each panel, each configured to only show tasks that appear on that monitor.
And, I can now stick the system tray and the menu on the inner sides of the monitors (so the bottom-right corner of the left monitor and the bottom-left corner of the right monitor), which would be impossible to do on a “single large panel that spans both”.
And, when I finally get a three-monitor setup working, the menu and systray will only appear on the centre monitor.
Doesn’t that make them harder to aim/click?
Per Fitt’s law – if they would be in “really corner” the target effectively has a sort of unlimited size.
(but seriously, your PC, your business how things are done
)
It’s more about making it less travel compared to where my eyes are. I’m always looking near the middle of the screens, swapping from screen to screen depending on the app, so it makes it easier to have all the information near the centre.
Hm, it almost seems like what you’re really saying is that, right now, the most looked at part of your desktop is… the “combined” bezel between monitors ;P
I can see why you might want to get a third one ;>
(but seriously & BTW, if the central monitor will be larger, with its upper bezel at the level of two smaller ones, then Fitt’s law can be still nicely applied to tray/menu in two bottom corners)
It’s not so much that I spend all my time looking at the bezel between the monitors, but that I’m switching from “app on left monitor” to “app on right monitor”, that when I need to see what’s in the systray or access the menu, my eyes are already near the middle.
I actually have 4 monitors, two connected to one station, and two connected to another, with synergy sharing a keyboard/mouse between them (and configured to wrap the mouse cursor around).
I’ve played around with various configurations for the location of the menu, systray, taskbars, etc, and this is the setup that works best. For me, of course.
I didn’t migrate off KDE 3.5 until I found that option. However, despite it being superior in theory, I could never get fully comfortable with it for some reason.
Switching from Plasma to LXPanel was actually a bit satisfying because I regained the ability to have all my windows pile up on the left-hand monitor’s taskbar (regardless of which monitor they were one) when there are no more than five of them.
I’m afraid you’re trying to split hairs now in order to avoid the main point – it’s a pile of crap that does very little that a minority of the minority use.
Ergo, you don’t matter.
I have no idea what that means, but put simply Firefox and LibreOffice have features people want, Epiphany and and GOffice don’t. Very simple.
I’m afraid you made that very clear when you started talking claptrap about being able to shave milliseconds of response time off by using LXDE – having spent quite a bit of money on hardware and 16GB of RAM no less. You just don’t matter I’m afraid.
Don’t care. You’re a minority of a minority.
Ahhhhh, there we have it.
I’m not talking about your needs here. You’re a minority of a minority. If that work for you great, but you’re a very small minority nonetheless.
I hardly think so. My point is that this is the same claptrap that has been espoused by users and developers of Linux desktops for ten years.
GTK has made a decision that they can’t possibly please everyone. They just don’t have the manpower to maintain it properly.
Great. However, you’ve missed the point that you are not important here.
And now you’re resorting to the adult version of putting your fingers in your ears and saying “na na na na na”… coupled with an ad hominem dismissal.
I can still see a lot of avenues I could take to further my argument, but you’re obviously too attached to your view of who I am, what I want, and the state of GTK+ so I’m not going to waste time talking to a brick wall.
All linux desktops are used by a small minority – much closer to themselves than to the big boys.
Anyway, I’m not using desktop environments, I’m using applications; DE needs to just get out of the way.
Apps should be loosely integrated, so I can choose what I consider the best full-featured combination, and not on the basis of their relation to particular DE.
^it’s perfectly valid approach, /me thinks…
Also, there are older computers out there. For example, a mini-webcafe at one institute was set up with old, surplus machines – just for browsing and ~office. For quite some time, they were running KDE 3.x (and OpenOffice, not KOffice…). But KDE 4.x was a bit overkill; now they run LXDE just fine – BTW, similarity to Windows is a desired trait.
And you know, there’s one saying about opinions and arses :p
~Marketing surrounding for example Opera often goes in that direction. BTW, it’s a rather full-featured browser…
(though yeah, the market seems limited; but Opera does see decent success in CIS, where machines tend to be older/slower; and a significant success in mobile, where the handsets are often very slow)
But Chrome and Firefox were also marketed similarly, at one point or another. And browser is one of most significant apps used by people nowadays.
But it is still C++ bloat.
GTK+ was reasonable, but what normally happens is that 2.0 is supported through 4.0, and maybe even 6.0 and 7.0 to allow for a transition, and the differences involve critical design and operational differences, not cosmetics.
Worse, Bluetooth PINs never worked, and the Network Manager would display every AP on my drive to work, but take 2 minutes or more to find my local AP. That was a “feature”.
I seriously disliked KDE, but Gnome managed to screw up, down, and sideways what used to be a straightforward, flexible, robust system.
Considering that most desktop applications seem to be being made in python these days, is c++ really too bloated? The overhead compared to c is far less than most things, and when no stable and well supported c-based alternative exists…
Now I understand why so many applications in GNU/Linux have become dog slow.
A bit slower to start, maybe, but, as a PyGTK user, I can honestly say that the problem isn’t the language. PyGTK does a pretty good job of implementing anything not I/O-bound in C.
The problem is that PyGTK makes GTK+ accessible to know-nothing novices who used to give up because writing a stable application using the GTK+ C API was too much work.
They’d write applications just as badly if they used Vala, which compiles to pure GObject-flavored C.
Edited 2012-11-09 14:10 UTC
It’s not just Gnome, it’s Ubuntu with Linux. In some cases it’s the Linux kernel too.
I grew tired of “updates” breaking things and having to spend days “tweaking” things after an upgrade or in some cases a just major updates.
With Solaris/OpenSolaris, with rare exception, I was able to update and upgrade more or less with impunity. And programs kept working.
With Mac, at least until Lion, upgrades weren’t as in-your-face with required work flow changes (i’m pointing at changing scroll direction defaults, for example) most of the time.
At this point my primary get work done environment is Mac for one very simple reason: I turn it on and can be productive. I don’t have to tweak things practically every week like those I work with on Linux. I dislike the direction Apple is going as a company, so I am looking at options again – vote with my wallet – but until I find a similarly productive environment I’m sticking with my Mac for better or worse. This ability to be productive was critical when I was working for myself – it meant virtually no non-billable time due to screwed up systems requiring tweaks or fixes just to be able to work reliably.
Sure, the main thing that drove me away from Linux on the desktop is squarely at the feet of Ubuntu. An update seriously borked my system and because my system wasn’t “new enough” it wasn’t considered a real bug. At that point I looked into options and landed on Solaris, and then later Mac. I’ve been much happier. Tweaking my environment now is because I want to, not because I have to; it’s on my schedule and not whenever something breaks.
Sorry but people who excessively tweak their computers should consider spending more time outside and less time playing with their computer… Gnome 3 is great because no tweaking is required. If you don’t like it then use something else, don’t try to gain attention from people by starting flame wars.
There does seem to be a pretty extreme rift in the OS world. There are those who “just want to be productive” and for them Windows, OS X and (to a certain degree) Ubuntu meet that need. It seems the Gnome devs want to join that party too.
Then there are the folks on the other side who crave total control over their systems. Often you’ll find them starting out with something like Slackware and KDE, then getting into Arch or Gentoo with a minimalist window manager, or in some extremes foregoing X altogether and doing everything in screen sessions and unplugging the mouse.
There are, of course, a few of us who stay in that tenuous middle ground, and I feel that’s the worst place to be. It’s always in flux; one day the kernel is broken because some guy upstream managed to squeeze into the release channel a performance patch that works for the tweakers but throws everyone else out of whack. Another day the devs of a very large and popular desktop environment project decide to throw out almost all the code and ideas, and start from scratch with something alien, just because they can.
The relative stability* of Windows, and to a lesser extent OS X, the BSDs and Solaris, makes for a pretty compelling platform for those who lean towards using their computer as a tool or appliance to get things done and/or make a living. For those of us who view the tech sphere as a hobby first and a career second, well sure we probably would benefit from some time outdoors.
Maybe I’ll go for a walk while my tweaked kernel is compiling for the fourth time today…
*Speaking purely of update release schedules
Edited 2012-11-08 23:18 UTC
That sounds about right. I was perfectly happy on KDE 3.5, where I could get work done when I wasn’t geeking around for fun’s sake but, now, it seems to be getting harder and harder to find something that’s stable, performant, and doesn’t try to force an alien workflow on me.
In fact, I think that’s the real distinction. I come into a desktop knowing what I want and, if I can’t get it, I’ll dig in my heels and fight until I do before I’ll consider alternatives, superior or otherwise.
I switched off Windows cold-turkey onto Mandrake Linux 10.0 and it still took me several months to feel safe on an OS where I hadn’t been a power user since my age was measured in single digits but I was a KDE 3.5 user from the start because I knew exactly what I wanted and no desktop was going to tell me differently.
(And I was willing to put up with eye-searingly glossy/glassy icons if that’s what it took to get that level of control. Remember, this was before the XDG Icon Naming Specification.)
If you’re not going to upgrade the system, I suppose this is true, but once you factor in upgrades and updates (and you are going to need to upgrade at *some* stage), I’m not sure it holds. Ubuntu updates often break things so that you then have to spend hours making simple stuff work again, and OS X now seems to want you to adapt to new ways of working with each new release. I went back to Slackware on my main work machine mainly because I was tired of change and wanted stability. Slackware 13.37 was the first version I’d used since 9, and I could essentially pick up where I’d left off.
Bang on right, though for many this is a matter of productivity rather than control. I do this more and more when I’m working, for example, largely because I’m sick of desktop environments that want to reshape my work habits or worm their way into my personal identity. I just want them to get out of my way and fade into the background so I can get on with work. A simple window manager like Window Maker or Openbox in contrast for the most part does just this. And yeah, I do work in X-less sessions – I started doing this as a reaction to where modern desktop environments seemed to be going, but I’ve found that if all I need to do is do a test-run of a model, or write a paper in LaTeX, a purely text-based session is wonderfully distraction-free.
The real problem is much larger than the predicament we’re facing with arrogant developers. The younger generations are getting more and more totalitarian – they get easily offended and have less empathy. I dunno why it’s like that. Perhaps in an overcrowded world with billions of egos, people have to fight harder to stay on top…
It’s the whole coddle approach they’ve been through.
Honestly, when you can’t lose, have to be given a trophy even if you came in last, always have to have your self-esteem measured and propped up, when you get 1+1 wrong and told you’re right….what do you expect?
A me-first, I’m never wrong, easily offended generation.
You know, we have examples of ~”moral decay of youth is a threat to civilisation” since the dawn of preserved written word… (some Babylonian tablets, for one) which were definately more violent times than our (post?)modern world.
In related news: past tends to be often seen as better than when we lived through it (but, notably, we don’t seem to have now that many problems with totalitarianism as we did in the past)
PS. And on the general topic of “those damn youngsters!” – I must note that boomboxes and shouting are a thing of past – teens nowadays prefer the privacy of headphones and text messages ;P
Edited 2012-11-09 22:05 UTC
Ekk, when Men were men, Mounties always got their man and coders er…
I get the impression that you didn’t fully comprehend the article linked.
The complaint is that GNOME has started treating GTK+ like an internal GNOME component rather than the foundation of 3rd-party desktops like Xfce and LXDE. SpaceFM is a fork of the legacy branch of PCManFM, the file manager for LXDE.
Let’s pretend I do use GNOME for a moment. I don’t know about you, but I like having more choice in themes than “too much padding: light” and “too much padding: dark”.
I must admit I didn’t read the article I was put off by the headline & posted snippet. If that is true then I’m sure it could have been presented in a more mature way. What will *really* kill a community is disrespect & childishness, not API changes. If there is a valid complaint to be made then you don’t need superlatives (“rotting”, “every turn”), and you don’t need to make heated comparisons (microsoft-like). When you add emotion to a discussion like this you lose all respect and therefore you lose your message. (I mean the original author here)
There are plenty of themes available anyway just because there isn’t a really nice way to apply them yet doesn’t mean that all the developers should be shot and burned seriously programming takes time & effort nothing happens overnight but it doesn’t mean it will never happen.
Apparently, you still haven’t read the article.
In it, gnome developers are quoted, with their opinions being that themeing should not be possible.
Haha woops anyway opinions can be always changed but only via the correct manner.
Except for the need to remove and replace the abomination that is Gnome Shell… or the need to add a new file manager because Nautilus 3.6 has been completely gutted of functionality.
if you look at how well computer interfaces have done for long periods of time without changing, and if you look at how insignificantly different even cutting edge ones are today, you have to wonder if too much is made of linux GUIs.
the GUI guys can go crazy if they want doing all kinds of crazy stuff. but I think a successful distro would sit with the same thing and innovate in other areas unseen to users.
Dear whiners:
GNOME devs made it clear that they weren’t interested in maintaining the theme engine for multiporposes, just for the needs of the default theme, they said this since day one. If you wan’t to customise it and make themes and whatever, you will have to waith to the API settles down or update your theme with every release.
The engine still needs optimization and they are working on it, if you don’t like it, you can always fork it, or contribute to other DE like KDE or XFCE, cause, it is a free world.
BTW, the Linux kernel break a bunch of stuff with every release, and I don’t see you bitching about it.
KDE 4 deprecated all the python, ruby, C#, etc. plasmoids for the sake of QML and javascript and I don’t see you whinning eather.
Edited 2012-11-09 00:59 UTC
XFCE is based on GTK+, as is LXDE. The only DEs which aren’t being forced to choose between dead code (GTK+ 2) and having GNOME devs yank them around like marionettes (GTK+ 3) are KDE and Razor-Qt and only when used without programs like VMWare and GIMP which use GTK+ for their GUIs.
…because, if said breakages do exist, I never encounter them. This stuff actually affects the end user.
Because I was unaware of it… Great! Now where am I gonna flee to when Weston implements client-side window decorations?
Maybe I’ll still be able to use KWin with no other KDE apps when that eventually becomes an issue.
Edited 2012-11-09 01:07 UTC
Because because because, GNOME devs also have becauses you know.
Edited 2012-11-09 02:21 UTC
Yes, but opinions don’t hold equal weight merely because they are equally fervent.
By welcoming GTK+’s rise to popularity as a general toolkit used by many different desktops, the developers also implicitly accepted responsibility to manage GTK+’s development wisely.
As a user, my only responsibilities are to be polite, to avoid acting entitled, to report bugs, and to choose a desktop where the developers views on good UI design are not diametrically opposed to mine.
I have done all of those. I am not a GNOME user and my issue isn’t with GNOME. It’s with the GNOME devs abusing the power over other desktops (like LXDE) that maintainership of GTK+ grants based on those other desktops’ views of their formerly responsible behaviour.
Edited 2012-11-09 02:37 UTC
Easy solution to this: Some developers from XFCE and/or LXDE contribute some man power to maintaining GTK?
No because I think everyone outside of the Gnome/GTK projects now knows not to deal with Gnome/GTK development. Anyone who has tried in the past has been met with a campaign of misdirection and misinformation as well as being marginalised via Gnome/GTK devs going ‘radio silent’.
You simply cannot work with them.
Dude, XFCE and whoever uses GTK is free to fork it or use another toolkit, so, what’s ur point?
BTW, I haven’t hear any one from the XFCE team complaining, you complaine more than them.
Edited 2012-11-09 03:11 UTC
First, there’s always selection bias to consider. Just because you didn’t see it doesn’t necessarily mean it didn’t happen.
Second, I’m just clarifying the parts of my viewpoint which match those from the post.
Had you not begun your original comment with “Dear whiners:”, I probably wouldn’t have felt strongly enough about it to have replied.
The point is that you don’t want to duplicate manpower simply because developers are douchebags and don’t understand what the meaning of a point release is. The Gnome and GTK devs are worse than Ulrich Drepper. However, in that case you simply have to make a decision as to when it’s got beyond the joke and you fork or move to something else. That will be coming.
I love it when people write stuff like ‘they are free to fork it’ as if they are writing that from a position of strength. In Gnome’s case they are simply bleeding into total irrelevance through lack of manpower.
No, they didn’t. This also goes way, way beyond the theming engine because it’s become very clear that GTK is a rapidly moving target.
We are several point releases into GTK 3.x now. A Minor release means you maintain backwards compatibility. The API will never ‘settle down’.
It’s OK. That’s happening.
No, they don’t. People in userspace outside the kernel moved from Linux 2.4 to 2.6 with barely a murmur. Stuff worked.
While that kind of thing is regrettable it had to be done because QML could be better supported, Plasmoids made more reliable as a result and more could be integrated into Qt itself as a stable target for everyone. Divergence in that sense is just not beneficial.
GTK is currently the complete opposite of that. It is breaking for no good reason every other point release with absolutely no release policy other than ‘We felt like it’. People took shots at KDE 4 but it never did anything like this. They started off with a basic .0 release and moved solidly from there with API and ABI compatibility.
When you have no developers and you are pissing off the ones you have, and you are pissing off your users because of said developer breakage then you are dead. Have fun.
No, they didn’t.
WTF? Go read the mailing list you ignorant troll, segen00b, you always talk just for talking, you are not even worthy of my time.
Read what mailing list you utter twit? Theming is a part of GTK, or at least it is supposed to be. It’s, you know, a graphical toolkit. The default Gnome theming is something else. That’s where these problems are coming from. A toolkit, GTK, is enforcing Gnome preferences on everyone else.
Application developers are whiners. Pure brilliance. Pure ignorance. Troll….. Hilarious.
Edited 2012-11-09 14:59 UTC
Blah, blah blah, You obviously haven’t read the mailing list and you are just trying to spread the dust, it’s ok, i would do the same if I were as ignorant as you, and themes are so overrated, of course, it is fun for 15 years old who like to customise their desktops instead of doing something productive with them, btw, are you even a developer anyway? I don’t think so.
RTFA basically and understand what you are actually talking about. Vain hope I know.
You’ve read the summary, haven’t the faintest idea what the article is actually about, waded in with a ‘whiners’ comment and ended up reinforcing what the entire problem is about. Well done.
What we have here are GTK developers having Gnome’s preferences enforced that have nothing to do with them. GTK, a graphical toolkit, doesn’t support theming? Really? I’m afraid you’ll find GTK’s developers (or is it Gnome’s?) have never said that. In fact, they’ve never said much of anything because application developers have had to go through the commit logs to find out what the hell is going on.
Desktops that tell everyone they’re only going to support one of this or one of that go to die. Programming toolkits for UIs that tell you they don’t do theming (you know, the ‘graphical‘ bit?) is a joke. Like I said, have fun with that sweetheart.
Yeah, whetever you said segetroll, btw, did you read the mailing list? nope? that’s what I though.
And btw, application developers doesn’t need to care about theming, damn you are ignorant, If I make an application using GTK what would I care about the theme? the application will use whatever theme is using by the DE.
btw, have you seen a XFCE developer complaining about this, or is just you?
Edited 2012-11-09 16:46 UTC
Since you haven’t read the article I’m afraid whatever you’re going on about is irrelevant. GTK != Gnome.
Because people make things called…….desktop environments with GTK and it’s also supposed to be, you know, cross platform? Are Windows and Mac OS’s users of GTK apps supposed to look at Gnome’s default theme?
Sorry, I can’t laugh any harder and I will refrain from poking fun further at someone who obviously has issues and where the whoosh has long gone over his head.
Read the article before commenting further.
Edited 2012-11-09 17:06 UTC
desktop environments with GTK
A DE is not an application, damn, can’t you use the correct term segetroll?
When someone makes a DE, it takes one version of GTK and base the development on that version, if they want to upgrate to another version of GTK then they rebuild and make the necessary adjustments and workarounds to use that version, I haven’t seen any DE developer complaining, have you? nope, but as always you make things out of thin air. I have no words to describe your ignorance.
Edited 2012-11-09 17:09 UTC
Read the article before commenting further.
I did read that sorry article, and if you can read, you will note that GTK is still themeable, you should read that the API for themes is evolving and will settle down, mean while all you need to do is upgrate your theme, but as the anti-GNOME troll you are, you extrapolate everything.
Edited 2012-11-09 17:14 UTC
Dear dumb ass gnome developers:
The freshman software engineers in my college learn to ask for user requirements before they start coding. How about doing the same. There is no point wasting your time if no one wants to use the crap you produce. And your stupid bar across the top is a waste of space if I can’t alter it. You fail at UI design.
Dear ignorant troll:
I use GNOME Shell just fine and I’m so productive with it, the fact that you don’t share the same opinion doesn’t mean we all dislike it.
Is called, “Get the facts 101” do they teach that in your college?
Edited 2012-11-09 16:09 UTC
Bullshit.
I maintain a fairly large out of tree kernel project that covers everything from kernel file management to networking and memory / cpu management (and includes a number of out-of-tree drivers).
In the past 10 years I seldom spent more than 1-2 hours porting my code to a new kernel release.
Care to prove wrong?
– Gilboa
Good for you Gilboa, I was thinking the same, but the only it brokes are mostly closed source libraries that cannot be recompiled, yes, I had my doubts in that statement, so I take it back.
I’d give your comment +1 if I could
– Gilboa
Gnome doesn’t worship Apple, if they did they wouldn’t try to reinvent things using GTK they’d just use GNUStep and just remake OSX.
<troll>
No, because they inherited Apple NIH sindrom and wont use something because it is someone else idea (until they copy it, then it’s theirs). So they did their own language and are working on their own toolchain too. As for copying OSX, once they removed everything, they will have a very nice base to start with.
</troll>
Enough trolling, I don’t think the way GTK is done is future proof or developer friendly. I once was working on a GTK application and had a bug. I had read the doc and everything I could find. The function was still not doing what it was documented to do. So I went on IRC. They were really polite and helped me solve my problem quickly. When I asked how I was supposed to know that the function was not doing what it was documented to do and asked about updating the doc. The devs acknowledged that there was a fair amount of what they call “assumed knowledge”. Those facts float on top of the API doc and they confirmed that it does not really represent the state of GTK anymore and that it have to be fixed eventually. I understand they had a major release (3.0) and wanted to clean things up. The problem is, the API un-freeze has not ended ever since, even if it is officially in place.
I have lived through Qt3, Qt4 and Qt5 transitions, I am totally find whit API breakage. I also work on AwesomeWM and no API freeze is one of our core value, even if it drive power users crazy, but we made it clear. API are not frozen until they are perfect and there is no such thing as perfection. The problem with GTK is that dev have to deal with this fog. They can’t really see where things are heading and until this have either settled or documentation/migration policies are well defined, their product is in danger.
I use many GTK apps, many complex ones, so I really fear where things are heading, aka: nowhere. Application can not survive unless what they are based upon survive too. It is how free software work, we built upon other people work and rely on this trust that if they provided a public API, we can rely on it until the lib****.so.X.*.* change (X). If some low level projects drop this very fundamental aspect, everything crumble. Forking GTK is not an option as many here think for two reason. The first one is that Gnome would have to fork GTK, not the opposite. As everything else rely on the old (theoretically stable) namespaces and when it come to that, there can be only one. The second is manpower. I don’t think C based GTK is competitive (C is not a good GUI language, it wont ever be, C is the best, but only for use cases where it actually is), but removing all the paid for development will make things far worst.
So where are we. We have a team of vision driven developers with a dilemma. Support third party or focus on getting where they want to be. I am fine with vision, I did many project just because I could and looking back I am totally fine with the fact there was no market. But when it impact other peoples work and ideals, is that really acceptable behavior? I don’t think so. There is no solution. The GNOME can’t be forced to care about something they don’t. Even if many of them are paid, they have their own agenda, it’s still open source. Third party devs can’t switch either, it is too late. It is “go with them or go from scratch” senario. Very few are crazy enough to believe in rewrite unless their own codebase have crossed maintainability infection point. So many issues, so little solutions…
I’m afraid I would call that reality. After ten plus years of watching what looks like a duck I am forced to conclude that it is, indeed, a duck.
They are polite as long as you believe that their way is the best way.
So they don’t understand point releases and they don’t understand maintaining compatibility between those versions? GTK 2 certainly got that whether you think it is a good toolkit or not. GTK 3 is lunacy.
Well, if Gnome wants to support only one theme then fine. Great. It shouldn’t stop people should they want to. However, the real problem is that they are enforcing this view through pointless breakages in GTK – a graphical toolkit that should be stable that people who don’t use Gnome use.
The fact that they’re filing bugs with GTK applications to tell people that they don’t support something any more shows an astonishing level of intolerance.
Harsh is good, if it’s true. And it is. About themes, about customization options, about nautilus, about api, branding trials, it’s all there and all who use some of the mainstream Linux distros these days come across these issues at one point. For knowledgeable users most of these issues can be resolved, one way or another, with time spent, but for others they just might break the deal. And I don’t think it’s overly negative, it’s just seems to come from someone really frustrated (which I can understand and relate to).
http://tirania.org/blog/archive/2012/Aug-29.html
Edited 2012-11-09 08:16 UTC
And quite a few people found his post of a very low quality..
And who was that? Mainly the people he was criticising.
I haven’t seen one post rebuttal yet that actually invalidated any of what he said.
Edited 2012-11-09 13:11 UTC
You’re kidding? You can summarize his post as finger pointing: the Linux desktop is a mess (true), but it’s not my fault, it’s because “the Linux kernel has no internal stable API”, really??
He picked THE software in a Linux distribution which has a forward compatible ABI for userspace as the reason why desktops lack stability/compatibility?
I don’t remember that when he worked on Gnome he was especially pushing for compatibility, but it must be Linus’s fault also.
Well you suck at reading comprehension.
He was talking about the attitude that is was okay to break APIs, is prevalent because of Linus. While people could argue about whether this is true or not it really doesn’t matter whos fault it is.
Breaking APIs is the problem with no path to follow while transitioning. Which is what the blog post is about.
Edited 2012-11-09 14:19 UTC
Yes, and one cannot help but notice that when he was leading a major Linux project he didn’t do anything special about this point..
He does say “we” when referring to those that were guilty of it.
Edited 2012-11-09 15:19 UTC
He totally misunderstood that, which is probably why he doesn’t understand a lot of things. To blame his mistakes and that of others on the kernel just shows how little he understands.
Linus has no qualms about breaking APIs within the kernel, but what the kernel did not do was break compatibility with userspace applications – i.e. external interfaces. That’s something Linux desktop components have done time and time again.
Edited 2012-11-09 20:25 UTC
No he didn’t misunderstand that, you guys keep on misunderstanding the fact it is about the attitude … the specifics are not important.
Software Engineering is as much about attitude as it is competence.
Oh goody … well people <insert device> are still breaking after updates.
No, it wasn’t about attitude at all. Miguel specifically talked about the Linux kernel breaking APIs without the slightest idea as to what actually gets broken around the kernel. What doesn’t get broken are the external userspace interfaces, and Linux desktops have done that time and time and time and time again.
Speak to your distributor. I fail to see how Linus is responsible or the updates they put together, but this kind of logic appears to be common. The point still stands that the kernels userspace interface have remained totally stable.
Yes it was. You are having a reading comprehension failure.
No, and it looks as if I’m going to have to point this out otherwise we’ll be in Punch and Judy territory:
Linus pointed out that he was completely, totally and utterly wrong about this because the kernel has never changed any interfaces for anything outside the kernel – i.e. userspace. People switched from Linux 2.4 to 2.6 without any problems whatsoever. People upgrade their kernels all the time without userspace stuff breaking.
He’s completely misunderstanding what drivers are actually a part of and what he’s actually talking about and trying to blame the internals of the kernel for the mess of the Linux desktop that he perpetuated.
The problem with Linux desktops is that they change their external interfaces every five minutes. Even when things are API and ABI compatible, expected behaviour changes. It’s just funny that even now, after all this time, he just doesn’t understand what the problems actually are and is trying to palm them off on something and someone else.
No you are still wrong.
He is actually talking about the decision for internal driver interfaces which is more widely known than the kernel’s external interfaces.
Other developers decided to follow it, without understanding the difference between the two.
I think miguel understands the difference fine.
Anyway back to the initial point, it doesn’t matter who fault it was … it has happened and developers have gotten pissed off with the situation.
Edited 2012-11-12 09:38 UTC
you know that a kernel has one ABI facing userspace AND one facing drivers (or in general, modules), right?
and that OTOH userspace applications are all linked against the standard c library which actually implements syscalls – thus effectively decoupled from the kernel – but there’s no decoupling layer between drivers and the kernel?
that on the userspace side, keeping the kernel ABI stable is actually unnecessary (as long as the library is updated accordingly when disrupting changes are made) and it’s really enough to maintain compatibility at the library level – but it’d be necessary (and a sign of good sw design and development practice – ie compartmentization) on the kernel side (which ironically is the one ever changing)?
now, linux actually got it backwards in this regard (and also wrt desktop IPC…)…
and although users can adapt to the current situation for them to choose between using a certain kernel version OR supporting a certain piece of hw, and risk breakage/regressions (not that there arent any, see aspm..) updating the whole kernel, is a suboptimal situation
De Icaza may have been too far reaching in that sentence, but he wasnt completely wrong…
Miguel’s post was nonsense and more than a bit ironic, because as Alan Cox pointed out, he started and caused many of the fundamental problems. Much of the software he was at least partly responsible for coming up with was of such poor quality that it went through the very API churn and develop dissilusionment that he is talking about.
No It wasn’t.
It doesn’t matter even if he was part of the cause of the API churn.
He acknowledged:
* It exists
* He was partly to blame.
It is irrelevant he might have been part of it or even started it. Hindsight is 20/20 as they say. He was a lot younger then and probably didn’t know what he knows now.
Edited 2012-11-10 09:09 UTC
Yes it was.
No, I’m afraid it was a rant from someone who now uses a Mac who has used a Linux distribution recently and realised that stuff breaks and had a go at what he perceives to be wrong. Unfortunately, it was then pointed out to him that he was at least partially responsible for most of the problems he cites.
It’s called hypocrisy, it does matter and no he didn’t acknowledge he was to blame in any way.
So if we live in your world.
* People are not allowed to change their opinions.
* People are not allowed to make mistakes.
* People are not allowed to learn from mistakes.
* People are not allowed to comment on past mistakes.
OKAY!
Edited 2012-11-10 14:39 UTC
1. He hasn’t changed his opinion. It’s only now he is seeing things as how they actually are for an end user. He just doesn’t see the correlation.
2. He isn’t learning from his mistakes because he isn’t going to do anything to fix anything. He’s a Mac user now.
3. He isn’t commenting on past mistakes because he has never admitted that any of what he described was even partially his fault. It was only Alan Cox who pointed that out.
Put simply, his post was really quite pointless.
Obviously you missed the word “we” that was mentioned several times.
I’m using Gnome and GTK since 1998, for desktop and programming custom solutions for medical image viewing and processing.
This moaning about changes in Gnome are there from the beginning. It was very very loud the first few years of Gnome 2 and it is very, very loud now.
And I? I’m very happy and find Gnome 3.0 a breeze to work with, I have no problem with customization and my workflow improved significantly. I didn’t experience any problem to switch my development over to GTK+3 and added CLutter to mix, which by the way, changes it API also quiet often.
I’m really looking forward to what Gnome 3.x will bring in the future and think they are on the right track. Gnome 2.x needed at least 3 years before it stabilized.
You know the situation in Linux Desktop is bad when you have to choose between Gnome 3, KDE and Unity. Cinnamon is three guys hacking in weekends, and LDXE and XFCE are good from a usability standpoint but dated.
Linux Desktop is losing relevance quickly. It’s the ugly truth, but if Linux Desktop was run by business men willing to make money, they would stop the UI madness and divert resources into fixing X.org and PulseAudio breakages in upgrades. For example, OS X is run by people who want to make money, and they refrain from messing with the UI in too radical ways, and instead focus on making upgrading easy. Instead, the Linux world is run by arrogant people, and “if you don’t like it, it’s free, I don’t make any money from you I don’t care”.
As another example, if Metro in Windows 8 x86 tanks, Ballmer will oust Sinofsky and bring back a more traditional desktop. If Ballmer doesn’t do it, the stockholders will oust him and the new CEO will do it. But if you don’t like Gnome 3, they will push with their “plan” till the end, because they don’t make money from you. This is an advantage of the proprietary model IMO
It’s impossible to divert resources to fix X.org or PulseAudio. (there is something to fix on them actually? X.org is on verge to be replaced altogether by Weyland and PulseAudio is quite stable right now)
Why? Because “Linux” isn’t a company, it’s a kernel, one of the many projects that make a full usable OS, each one with their benevolent dictator, and every single one with their own priorities.
Developers in the open source world go to where their interests and most important, their abilities, fit’s in. I don’t expect that the average GNOME/KDE developer know a dime of driver development or hardware level infrastructure to mess around with X.org or Weyland, neither the opposite, that a Weyland developer knows how to design every single application used in a desktop environment.
Linux desktop is not loosing relevancy. It never had any. It is a desktop whose applications are made mainly by power users to power users. A market niche that will never exceed 5% of the total desktop market share.
And for such small market share, it’s quite impressive how alive it is. In Windows what i’m used to see is the same old applications refreshed and nothing new coming along for years in a row. In Linux, in just 12 years (release of WinXP), we left from KDE 2.0 to KDE4.0, from GNOME 1.4 to GNOME 3.6. Many window managers born and died, many new small desktop environments came along, with hundreds of new different apps… and goes on. It’s a thriving environment that clear fills his niche market quite well, and people who really use it are happy.
If someone wants to grab a chunk of all projects that makes what we call “Linux desktop”, fork it and make it a “user happy” OS, with plenty of market share so he can rule the world, they are free to pool the resources needed to do it and go ahead. Google is the living proof that this is possible with Android.
But don’t be so idealistic about the abilities of the proprietary model to fill up their markets. Inside a company there is far more than just money that can decide your future (and the company’s future), and egos play a large part on it. If companies was so fast to adapt and listen to their users, we would never see tech software companies going bankruptcy.
The Gnome dev team seems to be treating their desktop as a single standalone unit, plus optional extensions.
This is not how a desktop works. A desktop has to work with third-party applications, in a way that is conducive to the user getting stuff done. Gnome 3… doesn’t, which is why I suspect it will eventually die off.