Guest post by “I’m writing to inform you that the release team discussed Drop or Fix Fallback Mode yesterday. We’ve come to the conclusion that we can’t maintain fallback mode in reasonable quality, and are better off dropping it.” Gnome-fallback has been my refuge, as I find both Unity and Gnome 3’s shell unusable. Yes, we have been warned this would happen. I thought the cost of maintaining gnome-panel would be so low that it might never need to happen. But as it appears, it is deemed necessary. As for me, I’m bound for something Qt, as I am very angry at Gnome for abandoning its ‘classic’ users.
You once were the best when you knew what a desktop was.
What is a desktop, exactly?
Good question. A better one would be “what is a chair?”
For me, Trinity Desktop Environment is the only viable desktop. Yes the GUI looks a bit dated (like XP) but at least the basics of desktop computing have been met. I hope Timothy and his team port the whole desktop to Qt4 while keeping/reimplementing the existing functionality of the current TDE.
Edited 2012-11-10 04:58 UTC
not to mention that when installed bare, it just consume about ~60MB (tried it with opensuse 12.x KDE 3.x version).
Holly shit!
With today’s computers, RAM usage is not an issue, that’s why we have 64-bit however of course, let’s not go insane.
Anyway, the point of my post is simple things such as drag and drop work flawlessly without a hitch in TDE. I was able to drag/drop GTK and other toolkit icons including the URL icons of web sites from Firefox, Opera, Konqueror (and Chromium I think) from the address bar to the panel in the quick launch. I was also able to drag any icon (folder or non-folder) from any file manager I tried (Qt and non-Qt) to the quick launch. Regardless what toolkit the program I used is written with, (definitely not Qt), drag and drop just worked happily with TDE. I cannot say the same with some other desktop environments that we have today.
These features to me are basics for desktop computing and should not be omitted in anyway. Of course, re-arranging icons by left clicking an icon on the panel and dragging the cursor also worked without a hitch while the space between icons is equality adjusted and allocated for you automatically such as other icons being repositioned equally. Also, the size of the icons on the panel is always the same – width and height. You have all this with TDE.
Today’s other desktop environments lack aesthetics and common sense and whenever I use any of them, I feel like I am sitting upside down and try to use a computer.
Also with the latest TDE, you can convert the panel launch menu into the same one as in KDE 4. It is truly disappointing that major distributions ignore TDE superiority by not creating a TDE based version of their distribution.
Edited 2012-11-10 06:19 UTC
Sorry to burst your bubble, but it is the same reasons they dropped XMMS(1.x) or any other top GTK1 apps. It is not because they were not getting the job done anymore, but because of how hard they were to maintain. The TDE guy have to support over 10 million lines of code. Everytime something change and break compatibility, it have to be fixed. Every GCC version that change default settings and break compilation, it have to be fixed. Everytime upstream libraries bump API levels it have to be ported or removed. Scale that to 10 million lines of code and maintenance/packaging/support is not cost effective. It is that simple. It also serve a very small niche. Unless the maintainer is very motivated, TDE will get to a point where it will be impossible to maintain without forking even more libraries and making the problem even bigger.
TDE should have forked KDE 3.7 or 3.8, not 3.5. At least 3.7/3.8 were ported to Qt4, but it was before KDE4 features were merged in. It would have allowed to keep only the KDesktop and Kicker (possibly the old Konqueror and Kontact too) while enjoying upstream support of the base libraries and ability to blend in KDE4 apps when the new version is simply better, aka, Kate, Konversation, Digikam, KDenlive.
Well you see, Timothy should not have to do this but unfortunately he does because there is no coherence between toolkits. The foundation of every toolkit is different! There should be like a central protocol where all developers can turn to and follow.
If what you said is true, it means that the functionality I specified is hard-coded to handle various different messages from different toolkits but this is not Timothy’s fault and he should be praised for this because he tries to at least cover, lower the extent of the mess created by other developers!
With Windows, based on my own experience, regardless if the program looks like Office XP (super flat buttons), Office 95 (3D), Office 97 (flat) or is using the Ribbon UI, they all follow the same rules.
– Every window has a Window Procedure
– Every window receive the same messages: WM_ACTIVATE, WM_LBUTTONDBLCLK, WM_KEYDOWN … you name it because they all derive from the same object.
This is why everything works with Windows because there is one central foundation to everything that’s running and they are all related to particular objects following the same rules.
From what I have experienced, X.ORG does not have these rules and that’s why everything is broken and nothing is compatible with each and even toolkits themselves are not compatible with each other with different versions. It’s a joke but this can be fixed. Put everything in the past and start-over, create a protocol or if starting over is bad then take one toolkit and make it official and then everyone can use the same toolkit and improve it.
Edited 2012-11-10 07:59 UTC
And with Office it’s a bit notable that it perhaps still has essentially its own toolkit – or at least it had & shipped with it: Office 97 running on NT 3.x looked just like under win95.
But about “one to rule them one” – which one to pick? (plus not everybody will move)
Mmm, wasn’t the classic version had like zero users anyway?
I thought the cost of maintaining gnome-panel would be so low that it might never need to happen
Why don’t you try to maintain it and find out?
Edited 2012-11-09 23:53 UTC
Yada, yada, Gnome, whine, yada, yada …
This has been known since almost… forever?
Seriously though, how is something that has been said so many months (a year or more maybe?) ago news-worthy? If GNOME users didn’t already begin preparing to jump ship already, then… well, I can’t say I feel sorry for them. This could be seen coming from miles away. I don’t even remember the last time I used GNOME 2 (yes, 2… 3 is an atrocity, even with the legacy mode).
Anyone who jumped ship, if they didn’t choose KDE, LDE or Xfce, probably switched to Cinnamon or the MATE Desktop. Those are the logical choices and the most readily available. Anyone who really thought GNOME 3 “fallback” mode was going to last and is still on it… well, tough luck.
Edited 2012-11-10 00:13 UTC
Two years now.
Wow, time flies. I guess I was off… both by a large amount of time and in a bad way.
Ouch. People really have no excuse to be latching on to that thing.
Well, I liked it better than Mate, it worked better for me than Gnome2, and it allowed me to use a (for me) sane interaction model. Gnome 3.8 will be off-limits for me…
Why do all UI projects feel it necessary to change a design proven by 20 years use (Win 8 UI, Unity, GNOME) ?
The assumptions seem to be:
(1) Something’s old, therefore it must be changed
(2) We know what people want, so we’ll make the change without extensive user testing
Dumb.
And I suspect they’re hoping to force fallback users into the Gnome shell fold. I think it will work against them.
I think that’s exactly what they’re doing – trying to force people into their indoctrination.
Not at all. There’s a bug report* about making it even easier to get a GNOME2-like setup under GNOME Shell for the people that still want that kind of Windows 95-inspired experience.
https://bugzilla.gnome.org/show_bug.cgi?id=685744
Why do people keep claiming the Gnome devs don’t do user testing? Because they *did*, even if you don’t like the result…
UIs 20 years ago can be seen more different than the change now…
And Unity seems to work out fine ( http://www.osnews.com/thread?541935 ), also in the sense that it better utilises widescreens.
I think Linus prefers it too.
I’m split between that and KDE/Qt, but I don’t like the object disoriented obfuscation. Actually Qt isn’t that bad, I use it as one of two good class libraries (the other is Apple’s IOKit where things like CD, DVD, etc. and USB, Firewire extend SCSI).
Gnome is broken. The developers make strange, huge, arbitrary changes on a whim. I don’t know why, only that everything breaks or is unfamiliar. I can’t get wireless (Network-mis-Manager, or even freaking bluetooth PINs) to work properly.
Qt itself isn’t bad, it’s great for developing cross-platform stuff, for example, but I personally can’t stand KDE. KDE is just… it’s all over the god damn place, there’s gazillion redundant features and options there, and there’s literally no consistency there: just take a look at how many different places are theming-settings spread over, how many tabs and options are constantly visible in those, and how in some places the system offers you the option for KDE to automatically download and install themes you choose, in an other place the system only allows you to install from local filesystem, and in some places you have to drop to the command-line to install the themes!
I would love a DE with similar approach and values as the GNOME 2 had, but with Qt as the toolkit. Then again, I only use Linux on servers or in a VM, so maybe my opinion just doesn’t count.
Have you seen http://razor-qt.org/ as yet? Might be along the lines of what you’re chasing.
Aye, I am aware of that, but it’s still too raw, and it still takes too many cues from KDE. The theme needs a lot of work, for example, and judging from the screenshots the applications and settings use free space quite inefficiently.
I am interested in the project in general, though, and in general I have a good feeling about it. I could definitely see Razor-Qt becoming a high-class DE in the future, and I actually really hope it will do exactly that. Definitely a project worth keeping an eye on!
What OS do you use. Windows or MacOS X ?
I’ll try Razor-QT once more, I heard it’s developed nicely. I also really don’t like the look & feel of KDE4.
What are you using on the desktop, WereCatf?
You’re asking the wrong person. I don’t use Linux on the desktop, it’s simply too broken. But I use XFCE in my VM atm because I just haven’t found anything better. There are some things that I’d want XFCE to do differently, but it still suits me better than the others.
So are you only using a desktop in a VM?? Or some other OS as well?
I haven’t got the vaguest idea what you’re looking at be honest, but it ain’t KDE. The theme stuff is all from the same place because KDE uses a really novel thing called component reuse.
Mind you, I’ve heard this repeated many times pretty much ad hominem over many years.
It sounds like KDE to me.
All the spacing is off on every app … everything is just a little bit wrong.
Edited 2012-11-10 15:15 UTC
That’s the kind of response that just makes any discussion on these things go nowhere. ‘Spacing is off’ and ‘everything is just a little bit wrong’ is the kind of subjective rubbish we’ve seen for many years. It just doesn’t mean anything.
Not that I’m saying KDE is perfect. It’s always tough when you’re trying to keep a lot of powerful features in your software rather than cutting them. However, you can’t sell software based on its ability to do less I’m afraid, regardless of how ‘clean’ you think it is.
Back to the subject of the OP, I just don’t see what she’s describing regarding themes. Downloading themes, then having to install a theme from a file and then installing from a command line in different places? Really? When I access themes from the desktop or KControl it’s the same interface that comes up. That’s component reuse in action.
Yes it does.
Because like it or not, my initial judgement on somethings’ merits will be based on things being aestictically pleasing.
There is a lack of taste in a lot of open source projects, where they do stuff that looks good and adds nothing.
* Transparent Windows Terminals.
* 3D “Cube” rubbish.
* Silly Windowing minimizing effects.
Oh this argument again. I think KDE has a lot of un-necessary things.
Some themes do, do that.
No, he doesn’t. He prefers Gnome 3:
http://www.theregister.co.uk/2012/11/07/passion_of_torvalds/
Edited 2012-11-10 07:30 UTC
I guess you’re a bit outdated. Linus is using KDE currently, he had enough of GNOME already.
https://plus.google.com/u/0/+LinusTorvalds/posts/DbmEE8kXLDA
The article is from two days ago and mentions his KDE posting.
If you follow the threads, he just started KDE again when he wrote on November 2nd. The Reg article from the 7th is just way outdated in a weird way. I don’t think he’s gone back to Gnome3 after a week of KDE in spite of the prolific configuration options or the cartooniness (I agree with both).
Nowhere does he say he’s had enough of GNOME. All he’s saying is that he’s trying KDE.
The second link mentions that Unity, LXDE, XFCE will be affected since they reuse some of the fallback components. When the whole Gnome3 mess erupted I initially tried XFCE for a while, but missed a number of important things, so I went for fallback. Now I’m going to avoid GTK stuff a whole lot more, so XFCE won’t be my first stop.
> missed a number of important things
Examples?
I would say the opposite – other desktops miss a number of important things (except for KDE and Emacs, of course).
I tried XFCE4 a number of times over the last year and a half. I really miss:
– Freedom of positioning of icons (I use my Desktop like I use the top of my desk, I put current files on it). But you can run XFCE with some other file managers (eg. Nautilus) to take care of this.
– The Gnome clock, with the multiple locations, sunlight-over-the-globe and weather reports. (The whole point of this article is, that this will no longer be maintained by Gnome).
– The Gnome volume control, that allows you to set the volume output ^aEURoeover 100%^aEUR, I don^aEURTMt know yet how to do this in XFCE4 (though VLC will do this by itself).
Thanks, that’s helpful. Chances are some of your issues will be (or have already been) addressed.
– Manual icon positioning on desktop seems to work fine here. Did you encounter any bugs or were you simply using an older version of Xfce?
– Xfce doesn’t come with a Gnome clock and it does some things differently (more often than not, that’s good). Weather reports are in a Weather plugin. A worldclock indeed is missing (thank you for the suggestion), but you can add several Orage clocks each set to a different time zone – I’m using this method myself.
– Volume control (a mixer plugin) has recently been improved but I still prefer Xubuntu’s sound indicator. Not sure what you mean by “over 100%” output volume (is it some sort of a workaround for a bug elsewhere?)
Of course, there will always be Gnome2 or Ubuntu applets/indicators which are not in Xfce and often there is no reason to duplicate the effort. For these applets you can use an XfApplet plugin (works with Mate’s applets) or an Indicator plugin (currently works with Gtk2 indicators).
Thanks for your helpful suggestions.
1. Are you suggesting that I can now shift icons a few pixels to the right in the latest XFCE??
2. The Gnome2 clock applet was just really great, I also loved the little map that came with it. When I needed to see the time in another timezone, it was always just a click away.
3. Have you ever wondered why you can’t increase the volume beyond some point? Of course you can, if software wouldn’t limit you.
The main reason XFCE won’t be my first stop is it’s dependence/integration of Gnome/GTK stuff.
You can’t really, hardware has clear limits (plus it working at/near them often isn’t the most optimal situation, can introduce greater distortions).
Of course, there are some tricks – with volume control going “to 11” it’s more than likely at the cost of reducing dynamic range, introducing sound compression (NOT the same as psychoacoustic sound compression, like mp3! Check Wiki article about “Loudness War” to get an idea – compressing the dynamic range features there prominently). It perceptibly increases loudness, even while not going beyond the maximum threshold of audio output levels.
In short, it alters the sound somewhat, happens at the cost of quality.
If something is recorded at low levels, surely it can be played back louder than “max”, though I guess at the expense of quality, like blowing up a picture beyond it’s native resolution.
Not only when it is recorded at low levels, that’s the whole point (of dynamic compression, how it’s used in Loudness War, etc.) – but, yes, changing the sound somewhat.
At the very least it would tend to make noise more prominent. If the processing is simple (as is to be expected in such general user playback solutions) it could also introduce clipping of louder moments (really short moments).
Generally, that’s a discussion more for hydrogenaudio.org
(one of the better ~audio resources on the web, with a strict no-BS policy, adhering to essentially scientific method, properly performed blind tests)
I assume the gnome guys know about the rise of linux mint cinnamon and are just passing the torch.
I am personally not so fond of Cinnamon, but more importantly, I don’t like the Linuxmint upgrade model (or the lack of it).
And humans!
They used to make jokes about it years ago. Everyone knew back then that Gnome people hated stupid users and humans. The joke was: The way they’re going, soon all you’ll have is a blank screen. Well it looks like that’s not too far away now. The perfect Gnome desktop for stupid users and humans.
Stupid user/Human: I can’t see my desktop, gnome is broken.
Gnome devs: Yes! That’s by design, marked as WONTFIX.
Stupid user/Human: But how can I use my computer if I can’t see my desktop.
Gnome devs: You’re too stupid to have access to your desktop. For us to do that would mean we’d have to fix crappy bugs and we don’t want to do that. So the best way forward for everyone is if we just remove anything we don’t want to work on. This works well for all as stupid users and humans are too stupid to use anything we made in the past. We have to make new things for stupider people!
Real linux users don’t even need a desktop. Hell, the only reason they use a mouse is for gaming.
Real linux users must be like a true scotsman.
You know I just realized I don’t use a desktop. I guess that makes me a real linux user
But I do use a mouse for more than gaming
I dunno? I’ve been on fluxbox for about 10 years now and none of this really effects me. In fact, I go out of my way to steer clear of any gnome stuff as much as I can. I just found it to be a buggy bloated mess.
I prefer to stay in control of what packages are on my system. It makes it far easier to keep up to date and my system is a 100% working and highly efficient everyday of the year. I suffer no breakage and no lost time. I haven’t done any kind of reinstall for that length of time.
Sure it’s not flashy and shiny, but it works and I get work done with no distractions. Oh and I get to laugh at all the stupid users and humans whining about their broken gnome desktop
So I guess you’re right. I have no icons or folders on my desktop and the only time I click on it is to bring up my menu
It would seem you lose a lot of time to follow …what’s happening WRT your ~desktop and packages.
(BTW, argh @ mouse being such an element of gaming – sort of promoting the relatively simplistic core game mechanic of… pointing at things)
GNOME will face its final battle when Red Hat sees significant changes in their marketshare when they deploy RHEL 7 with GNOME Shell. It’s not that RHEL depends on GNOME Shell to survive or remain strong on sales; but there will be a great deal of Red Hat consumers that won’t just upgrade to their latest product.
There are many companies who make full use of GNOME 2 user interface through Red Hat current product and changing that paradigm with GNOME 3 would mean a huge distress and added up costs. Since Red Hat is left as the only strong supporter for GNOME, you better believe the changes happening right now on the desktop realm will affect market share. In fact we can see a huge mess going on with the Fedora project.
Since the introduction of Unity, Ubuntu is facing a huge decline. MATE, Trinity, Cinnamon and other fork projects are not convincing the community. XFCE will always be awkward. KDE ensures the desktop customization is to the point the users bleed dealing with so much overkill bars, options and what have you.
Windows 7 not only surpassed XP but also made up for the Vista fiasco, right on time when Ubuntu was coming along as a strong alternative back in 2009. Linux fanboys used to say in 1999 that Linux would be the number 2 desktop operating system by 2003. We are 9 years past 2003 and that never happened. Though Linux is fun, we’re in a middle of huge crisis right now and it will take good sensible developers to fix this huge mess.
Edited 2012-11-10 19:18 UTC
That’s most likely not true – in fact, Ubuntu might be pretty much the only non-Android Linux which sees a significant growth. Now, reliable statistics are hard to come by, but…
1. you can be sure Distrowatch ranking isn’t anywhere near reliable.
2. luckily, we have stats of hits on all Wikimedia services: http://stats.wikimedia.org/archive/squid_reports/2012-10/SquidRepor…
Ubuntu 1,189 M 0.69%
Mint 11 M 0.01% (supposedly “stealing” users from Ubuntu… two orders of magnitude less)
But what about trends, you say? Let’s check a year ago: http://stats.wikimedia.org/archive/squid_reports/2011-10/SquidRepor…
Ubuntu 522 M 0.41%
Mint 17.3 M 0.01%
…yeah (in fact, all notable non-Ubuntu & non-Android distros decreased in those stats over the last year)
I’ve seen already a replicate of your post, about those statistics. They don’t measure up the consequences of Unity yet. There is a lot of people still running old versions – to be perfectly clear, the right statistics should reveal a close look at the versions of Ubuntu being used. Unity based Ubuntu will be the smallest slice.
That is why this trend is obviously wrong to just say that “Ubuntu” as in Unity form, is taking over.
Edited 2012-11-11 15:27 UTC
Do you have any data supporting that? That’s how it works, if you put forward some hypothesis – not just writing “obviously” and such.
And if you’d at least look at those Wikimedia stats, you’d see that there is some data there WRT Ubuntu point releases – 3 of them listed, 12.04 two times (x86 and amd64) at 129M in total (that is already mroe than all other distros combined), and the only pre-Unity that has enough to be listed is 10.4 (LTS, I guess) at 50M.
Plus, you’d have to explain why Ubuntu registered such relatively massive growth in the years when it came default with Unity (and why other distros clearly lost hits in the same time, hmm…)
BTW, a disclosure: I don’t even use Unity, don’t really care for it one way or the other (but, as far personal anecdotes go: I have around some people which use it, and clearly like it)
Your data is coming from one place, and it’s not fragmented. You would need more sources to support the claim that Ubuntu + Unity releases have grown. Market share is something hard to measure and it is necessary that you find more sources. According to other sources, Wikimedia is wrong.
My data is from one of the few most visited serverfarms on the planet; hosting more than a few projects, many language-specific (regional) variants.
You didn’t provide any data.
Edited 2012-11-12 22:18 UTC
+1 Insightful
Oh yes it does. Fallback mode is fast enough to be usable on my netbook, whereas Gnome Shell is like molasses creeping up a hill. Using all this 3D stuff on the desktop is dumb… Especially when you’re on Linux, and the framework to support it properly simply doesn’t exist.
Odd… I run Shell on an elderly netbook (some kind of crap Intel graphics), and it works smoothly enough… no worse than any other desktop I’ve run on it over the years.