Window manager variety
Posted Apr 11, 2012 21:40 UTC (Wed) by alankila (subscriber, #47141) [Link]
Window manager variety
Posted Apr 11, 2012 22:00 UTC (Wed) by apoelstra (subscriber, #75205) [Link]
Would you like a window manager with 10-pixel-high title bars, which automatically positions windows in a tiled layout, which can be rearranged using the vi keys?
Perhaps you would. I wouldn't have it any other way.
But given the relative popularity of metacity and kwin, I suspect that most people would disagree -- and that is why we have 20 different window managers. Because nobody agrees on what "good" even means, let alone "good enough".
Window manager variety
Posted Apr 11, 2012 22:12 UTC (Wed) by fluxion (subscriber, #62978) [Link]
Window manager variety
Posted Apr 12, 2012 5:33 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link]
Window manager variety
Posted Apr 12, 2012 12:06 UTC (Thu) by sorpigal (subscriber, #36106) [Link]
Window manager variety
Posted Apr 12, 2012 16:11 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link]
It's simply that nobody cares much. I've seen a tiled window manager for Windows, btw.
Window manager variety
Posted Apr 12, 2012 14:31 UTC (Thu) by alankila (subscriber, #47141) [Link]
Window manager variety
Posted Apr 27, 2012 5:58 UTC (Fri) by Duncan (guest, #6647) [Link]
FWIW, my kde is highly customized as a sort-of hybrid tiling/floating mixture, with floating dominating, but various windows configured to specific sizes and/or locations, including two dominant themes of maxed-Y/half-maxed-X for side-by-side for things like browsers and terminal windows, and maxed-X/almost-maxed-Y, lacking only titlebar height (and sometimes no-border so I get full app height as if it had the titlebar anyway), for things like tri-pane mail, news and feed-reader clients. Combined with (sloppy) focus follows mouse, click-to-raise and with semi-transparent inactive, that lets me work with either the two side-by-side windows or with the almost-maxed and a half-maxed window concurrently, or with all three windows, the back half-maxed one viewable when focused (but not raised) thru the semi-transparent inactive almost-maxed window above it.
It sounds horrible I know, but there's a productive workflow there that I've grown to depend on, and I was VERY unhappy when an earlier version of it broke with my transfer from kde3 to supposedly ready but still in reality VERY early alpha quality kde4. Fortunately I was able to work around or find alternate solutions for the broken bits, and kde4 itself has improved dramatically since then, such that at least with semantic-desktop not only turned off but actually compiled out (gentoo, USE=-semantic-desktop), I can honestly say I'm enjoying kde4 as much if not more than I did late kde3, now.
Window manager variety
Posted Apr 12, 2012 0:16 UTC (Thu) by nybble41 (subscriber, #55106) [Link]
Anyway, there is no such thing as a single "good enough" window manager, or any other program for that matter. Different people have different requirements, which in turn demand different solutions.
Window manager variety
Posted Apr 12, 2012 6:46 UTC (Thu) by niner (subscriber, #26151) [Link]
Things we would lose are having extremely lightweight window managers and the choice about the programming language it's implemented in. Personally I don't think that the first one is really a loss. Window management is no performance critical task and I guess that even a completely new behavior would be just a couple of 100 lines of code. A device which is able to handle a graphical UI would already have enough memory that this doesn't matter at all.
Window manager variety
Posted Apr 12, 2012 7:33 UTC (Thu) by daniels (subscriber, #16193) [Link]
IMHO, it's a bit hypocritical for a Linux user to argue against the possibility of multiple, competing alternatives.
Window manager variety
Posted Apr 12, 2012 15:31 UTC (Thu) by nybble41 (subscriber, #55106) [Link]
However, were it not for the freedom to make unpopular choices, including the choice to develop an alternative to an existing "good enough" solution, we wouldn't have Linux in the first place. Anyone who works on Linux is already exercising that freedom, which makes it hypocritical to talk about denying the same freedom to others.
Window manager variety
Posted Apr 12, 2012 15:40 UTC (Thu) by daniels (subscriber, #16193) [Link]
However, were it not for the freedom to make unpopular choices, including the choice to develop an alternative to an existing "good enough" solution, we wouldn't have Linux in the first place. Anyone who works on Linux is already exercising that freedom, which makes it hypocritical to talk about denying the same freedom to others.
Yes, it would be if anyone had said that, but you're arguing against a strawman. Anyone who wants to work on an alternative desktop environment is more than free to do so; no-one is denying them that right. However, it doesn't mean that everyone around them has to care about it, and/or pander to their needs, requests, and demands of other projects.
Window manager variety
Posted Apr 12, 2012 16:28 UTC (Thu) by nybble41 (subscriber, #55106) [Link]
> Yes, it would be if anyone had said that, but you're arguing against a strawman. Anyone who wants to work on an alternative desktop environment is more than free to do so; no-one is denying them that right.
Sure, it's not literally _denying_ the freedom to work on alternatives--that was a bit of justifiable hyperbole on my part--but the original comment was clearly disparaging the idea of developers working on anything but the "1 good enough" solution, which, if consistently applied, would include anyone working on Linux.
> However, it doesn't mean that everyone around them has to care about it, and/or pander to their needs, requests, and demands of other projects.
I agree completely. However, while we're on the subject of strawman arguments... who said anything about providing support? The original comment was simply about developing alternatives to Weston. No mention was made of any expectation of interest or "pandering" by others.
Window manager variety
Posted Apr 12, 2012 8:06 UTC (Thu) by Seegras (subscriber, #20463) [Link]
And not only that, different people have different hardware. On the desktop I use WindowMaker with sloppy focus and autoraise, because a) I've got the screen real-estate and b) a mouse. On my netbook, I'm running xfce with click-to-focus, because the screen is very small and usually there's no mouse attached. And on my phone, there's an entirely different window manager running -- I'd be very unhappy with xfce or WindowMaker there ;))
Window manager variety
Posted Apr 12, 2012 14:56 UTC (Thu) by alankila (subscriber, #47141) [Link]
It's also a valid point that a different hardware setup requires a different paradigm. Whether Weston can adapt to, say, mouseless and keyboardless window management is a separate question in itself, but at least different interface devices undeniably pose different requirements.
Window manager variety
Posted Apr 18, 2012 19:13 UTC (Wed) by kleptog (subscriber, #1183) [Link]
I'll admit that progress is not very fast, but there is progress.
Window manager variety
Posted Apr 18, 2012 22:33 UTC (Wed) by khim (subscriber, #9252) [Link]
Your definition of progress is kinda strange: tiling window managers were tried and explicitly rejected! And it happened decades ago! They were tried and rejected before X Window System was born!
Tiling window managers were invented before anything else: remember that both Xerox Star (in 1981, no less) and Windows 1.0 (in 1985) employed tiling window managers! They were neat, but people rejected them.
Today they are in position similar to other old rejected technologies (such as Acme or FFM): some old-timers still use them and even few newcomers are choosing them but most users don't know about them and don't want to know about them.
Window manager variety
Posted Apr 19, 2012 0:14 UTC (Thu) by dlang (✭ supporter ✭, #313) [Link]
and therefor (according to you) such choices should be eliminated so that NOBODY is able to use them.
By that logic Linux itself should never have been created, the iPhone should never have been created, nothing new should ever be created because everyone is perfectly happy with what they have.
Window manager variety
Posted Apr 19, 2012 17:47 UTC (Thu) by khim (subscriber, #9252) [Link]
and therefor (according to you) such choices should be eliminated so that NOBODY is able to use them.
Probably. It really depends on the popularity of said things.
By that logic Linux itself should never have been created, the iPhone should never have been created, nothing new should ever be created because everyone is perfectly happy with what they have.
This is strange conclusion: we don't know if something will stick or not unless we'll try. Any new creation starts from one user and grows from there. Or not. If it becomes large enough then it gets enough resources to survive, if not then it's time to finish it.
Linux succeeded: it basically killed Unix and took it's niche, iPhone succeeded, too: it (along with Android) killed RIM and Symbian and took it's niche, webOS failed (and it's now time to write it off).
This is just a natural selection. It was not as troublesome 10, 20, or 100 years ago: markets grew, population grew, pool of knowledgeable workers grew, it was possible to keep both old and new things alive. Today… there are just not enough people to do that: if you want to continue to create something new then you must be ready to weed out something old, too.
Window manager variety
Posted Apr 19, 2012 13:14 UTC (Thu) by renox (subscriber, #23785) [Link]
Your own definition of progress is weird too..
At the time most users had small screen, so the situation is different now, plus small difference in implementation can make big difference in end-user acceptance.
And "progress" is not something predictable!
Window manager variety
Posted Apr 19, 2012 10:10 UTC (Thu) by mpr22 (guest, #60784) [Link]
I do not see how the "perfect size" for an arbitrary window can possibly be within the window manager's capability to determine, unless the display device is so pixel-limited that the only sane size for any window is "fullscreen".
Window manager variety
Posted Apr 20, 2012 8:56 UTC (Fri) by jezuch (subscriber, #52988) [Link]
Hmm, my only clue comes from Java's Swing UI toolkit, where everything is a tree of nested components, each component having a "minimal" and "preferred" sizes. The containers derive their minimal and preferred sizes from the contained components and the layout manager in use. There is even an operation called "pack" that resizes the window according to the preferred sizes of its components. A tiling manager could use this information but it required cooperation from the toolkit.
Window manager variety
Posted Apr 20, 2012 17:40 UTC (Fri) by apoelstra (subscriber, #75205) [Link]
>Hmm, my only clue comes from Java's Swing UI toolkit, where everything is a tree of nested components, each component having a "minimal" and "preferred" sizes. The containers derive their minimal and preferred sizes from the contained components and the layout manager in use. There is even an operation called "pack" that resizes the window according to the preferred sizes of its components. A tiling manager could use this information but it required cooperation from the toolkit.
So the "perfect size" is determined by the application developer? No thanks.
Window manager variety
Posted Apr 21, 2012 20:47 UTC (Sat) by kleptog (subscriber, #1183) [Link]
That said, I didn't realise the tiling window manager was such an old concept. I hadn't seen it before though, I certainly haven't managed to get my default Ubuntu install to do it.
Window manager variety
Posted Apr 21, 2012 21:14 UTC (Sat) by khim (subscriber, #9252) [Link]
That said, I didn't realise the tiling window manager was such an old concept.
Well, as I've said not only they are old, they were explicitly rejected. Windows developers apparently liked them (as you do) and though it'll be Windows advantage. Ha! The supposed “inability to create overlapped windows” was one of the more frequent complaints about Windows 1.0. It's obvious for the programmer that it was made explicitly (you had the ability to create and use any number of overlapping dialog windows in Windows 1.0, after all) but Joe Average (including Joe Magazine Observer Average) was quite literally unable to even imagine that someone will create something like this intentionally!
Of course the concept was quickly buried never to be seen on mainstream system again.
Window manager variety
Posted Apr 21, 2012 21:30 UTC (Sat) by dlang (✭ supporter ✭, #313) [Link]
In the case of tiling window managers, one thing to remember is that the screen resolution when they were 'explicitly rejected' for windows 1.0 was 640x400, screens are a bit bigger today, so things that wouldn't work then can work now.
Also, you need to break your mindset that Microsoft is the ultimate authority on the Right Way to do everything
Window manager variety
Posted Apr 21, 2012 23:12 UTC (Sat) by khim (subscriber, #9252) [Link]
there are a lot of things that have been 'explicity rejected' by people in the past that are in common use today. It's not unusual for other things to change that make something that used to not be reasonable now become reasonable.
Sure. But before you'll try to resurrect some old concept you need to find out why it failed in the past. If it's some technical limitation then it may be overcome, if it's something psychological then situation is much harder.
For example a lot of guys present iPhone (or sometimes iPad) as reincarnation of Newton, and yes, the devices look similar. But the difference is extreme: Newton was built around the handwriting recognition. It was it's cornerstone, everything else was built around this idea. And it just so happen that this idea does not work. iPhone/iPad looks similar, but they don't include stylus and don't use handwriting recognition. At all. You can as well say that contemporary planes are just a reimplementation of the Daedalus creation.
In the case of tiling window managers, one thing to remember is that the screen resolution when they were 'explicitly rejected' for windows 1.0 was 640x400, screens are a bit bigger today, so things that wouldn't work then can work now.
Perhaps, but I seriously doubt it. The biggest problem with tiling managers is perception of the lost control. People want to control their windows, not have them arranged automatically. Perhaps it's possible to create tiling manager which will feel natural, but I'm not holding my breath.
Also, you need to break your mindset that Microsoft is the ultimate authority on the Right Way to do everything.
Of course not! Microsoft is not perfect, but Microsoft is what people use. For better or for worse Windows defines expectations - simply because it's the most popular UI.
Even so… Microsoft tries very hard to destroy itself for the last five years or so. People hated (and still hate) their Ribbon - but apparently not enough to switch to something else. Perhaps Metro will finally push them over the edge?
This gives newcomes a chance. But I doubt something as radically different as tiling WM will be accepted by Joe Average even in the wake of Metro fisco.
Window manager variety
Posted Apr 22, 2012 0:46 UTC (Sun) by dlang (✭ supporter ✭, #313) [Link]
you are looking at things backwards.
If the talk about tiling window managers was only from people creating them, saying "this is what people should use" (the type of talk you see coming out of the Gnome or Unity projects for example), then you would have good reasons to point out past attempts and say "No, people don't really want that, here's why"
But when the statements are from people who are using a tiling window manager saying "It's the best thign I've ever used", then you should just shut up because otherwise you are telling people "no, you don't really like what you think you like"
As for Microsoft, If you really like what they do so much, go use it and let those of us who want some variation use what we want. We are used to having features as default (a desktop pager for example) that just aren't available (or only available via a third party hack) on Microsoft desktops.
The great thing is that on Linux, we don't force everyone to use the same thing. Even the distros that have a primary default allow you to switch to one of the other options (and in some cases, like Kubuntu, it's only barely a second class option)
Window manager variety
Posted Apr 22, 2012 11:40 UTC (Sun) by khim (subscriber, #9252) [Link]
But when the statements are from people who are using a tiling window manager saying "It's the best thign I've ever used", then you should just shut up because otherwise you are telling people "no, you don't really like what you think you like"
Nice strawman. I said, quite literally: the concept was quickly buried never to be seen on mainstream system again and today they are in position similar to other old rejected technologies (such as Acme or FFM): some old-timers still use them and even few newcomers are choosing them but most users don't know about them and don't want to know about them.
How exactly this went from this to "no, you don't really like what you think you like" I'll never understand. It's quite obvious that I'm not talking about individual preferences here.
We are used to having features as default (a desktop pager for example) that just aren't available (or only available via a third party hack) on Microsoft desktops.
Right. But mindshare of people like you is shrinking and it's not clear where is the natural limit of this shrinkage. If some tools are only used by people in some small group then it may find out few years down the road that they just don't have the hardware they need and they can't run the software they want (while staying compatible with the rest of the world). History is littered by examples: Lisp machines (they were all the rage back in the day), RiscOS (ARM is quite popular today and you can even run RiscOS today on PandaBoard… but how many former RiscOS users actually do that?), Amiga ST, Atari, etc.
The great thing is that on Linux, we don't force everyone to use the same thing. Even the distros that have a primary default allow you to switch to one of the other options (and in some cases, like Kubuntu, it's only barely a second class option)
This is greatest strength of Linux and also is greatest weakness. RiscOS fate looks more and more real as time goes on. There was time not so long ago when people used Linux en masses in the university. Today they use MacOS instead. Even if it does not have a tiling window manager.
Window manager variety
Posted Apr 22, 2012 17:02 UTC (Sun) by apoelstra (subscriber, #75205) [Link]
Tiling window managers have recently seen a spike in usage, due (I believe) to the sudden destruction of GNOME by folks like you, who argue "users don't want to know about that, so let's stop supporting it" and remove essential features.
There is no question that tiling window managers are more efficient. Nowadays the most common complaint about window management is that wide screens are too wide to read on, and it's a PITA to manually resize windows to look nice on them. So, in Windows 7 you can now drag windows to the sides of the screen, and they'll fill that half. So you can emulate tiling for its most common use case.
As for history, the furthest back I can remember tiling was in Windows 95. You could right-click on the title bar and click "tile". It was crap. Every window got roughly a square inch of screen space, you couldn't control which ones were visible, in what order, how much space to allocate. The default window decorations were still in place, and together they took up around half the screen area. So I wondered what the point of such a stupid feature was, and forgot about it until just now.
But nowadays, I have too much screen space for comfortable full-screen apps, no mouse (since this is a laptop), and I want my windows either 100% visible or 0%. So I've got a tiling window manager, and -now- I think it's the greatest thing since sliced bread. I've showed it off to many people, and they all thought it was very cool (though they were Windows users and didn't have the option to try for themselves). I have never heard of anyone saying they were crap, or restrictive, or useless, until you.
Window manager variety
Posted Apr 22, 2012 17:37 UTC (Sun) by kleptog (subscriber, #1183) [Link]
Fortunately things are not either/or. I have my xterms set to tile, because working with several machines at the same time it's useful to be able to see all the output next to each other at once, tabs just don't cut it. On the other hand, tiling makes less sense for the web browser or email. Hurray for configurability.
In my case since >80% of my windows are xterms it's more useful to have a tiling window manager that floats a few windows, than a floating window manager. But theoretically if a standard window manager had the option to enable tiling for a few applications you'd make 90% of tiling window manager users happy.
Judging by the comments on this thread though, I figure that's never going to happen.
BTW, there are tiling window managers for Windows, see the wikipedia article.
Window manager variety
Posted Apr 22, 2012 17:48 UTC (Sun) by rahulsundaram (subscriber, #21946) [Link]
Attributing to khim the decisions by GNOME developers is fair to neither since khim is not a GNOME developer and doesn't speak for them or even share their mindset in any real way.
"So, in Windows 7 you can now drag windows to the sides of the screen, and they'll fill that half. So you can emulate tiling for its most common use case."
Funny enough, so does GNOME which voids your earlier argument.
Window manager variety
Posted Apr 13, 2012 0:29 UTC (Fri) by robert_s (subscriber, #42402) [Link]
This is utter nonsense. This idea may fly on proprietary desktops where a certain behaviour/layout/theme is provided or mandated from above by "god", but this is a collaboratively developed free desktop where people are able to implement and their preferences more freely.
Window manager variety
Posted Apr 13, 2012 0:45 UTC (Fri) by dlang (✭ supporter ✭, #313) [Link]
Window manager variety
Posted May 10, 2012 15:53 UTC (Thu) by nye (guest, #51576) [Link]
No they don't. Metro could be disabled in the Developer Preview, but it can't in the Consumer Preview. It remains to be seen how the final release will work.
(I've been using Windows 8 for a couple of weeks now; in many ways it's a great improvement over previous versions, but the moment you have to interact with Metro it feels like you're on a truck that just hit a wall. It's a horrifying blunder; I can't believe they seriously thought they could release a new OS with changes like this after seeing how Vista was received.)
Window manager variety
Posted Apr 12, 2012 10:23 UTC (Thu) by cwng (guest, #74460) [Link]
the window managers/compositors in Mac/Win come with and are not designed to be replaced by the end user. It'd be interesting to see if anyone has tried.
Window manager variety
Posted Apr 12, 2012 11:23 UTC (Thu) by cortana (subscriber, #24596) [Link]
Window manager variety
Posted Apr 12, 2012 11:45 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link]
I distinctly remember being able to do things like collapsing windows to title bars by triple-clicking or adding extra buttons to window titlebars.
Window manager variety
Posted Apr 16, 2012 18:51 UTC (Mon) by sorpigal (subscriber, #36106) [Link]
WindowBlinds was not a replacement window manager for Windows. It merely did some hackery to change some of the look and some behaviors and it always ran on top of the existing WM. Think of it more like a very-heavy theme. In the end things like window lists, window position and stacking order were still the same. In addition, it was awfully slow and memory hungry.
Copyright © 2016, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds
LWN