Slashdot Mirror
Y: A Successor to the X Window System
impto writes "Whenever someone brings up the topic of replacing X, everyone always says that's nice, but where's the code? Well, Mark Thomas put his money where his mouth is and produced a replacement that maintains network transparency while adding many of the features that people desire from X such as alpha blending and a built-in toolkit. It still needs a bit of work to be as featureful as X but it's a fresh start that takes advantage of current technology and ideas. Read the paper here in PDF (1.7MB) or PS or grab the source and start hacking."
The website (available by clicking his name) should really include more information. It tells us that Y is the replacement for X, but offers NO reasoning. What problems does it solve? What features does it have? Can I install it alongside X? With this little bit of information, it makes me not even want to download it. Sorry.
Also, it should be noted development has stalled on this project. He says it should start up again in November.
I don't really like this... X is mature and popular, replacing it will take a *lot* of time, and most people will not switch unless the major distributions do.
And anyway, alpha blending and an official toolkit? No. The unofficial X toolkits work well enough, and alpha blending is not very hard to hack in - it's also quite useless for anything other than eyecandy.
I doubt this project will get much highlight after this initial slashdotting.
1. If it can't run existing X windows applications it's useless. Additionally if it can't run anywhere it's useless.
2. If you can't use existing XFree accelerated drivers it's useless because you're not going to make gfx card companies produce new drivers for another environment. Asking them to produce X drivers was hard enough, and some of them haven't even got that right yet.
3. I don't want unified toolkits because it means the entire KDE and Gnome desktop projects would need to be scrapped, unless QT and GTK can be re-created as "wrappers" to this new, unified toolkit.
4. It's a final year project. Sorry, but this guy's just an undergraduate student, no offense but I find it highly unlikely he can come up with something superior to X, QT and GTK (all of which this system supposedly replaces) in a year of work.
I agree to some extent.
However, the state of toolkits under X is now quite a mess. How many of them are there again? 15? 20? All with their own look and feel, and all with their own pain in the ass dependencies. It's not enough that GTK and QT is somewhat of a standard. That's still one toolkit too many.
Ideally, there should be one standard toolkit api that is easily extensible by developers (ie a very flexible widget system), easily reconfigurable by the users (one standard look and feel, that "power users" can change).
How small a thought it takes to fill a whole life
Everybody says X sucks, that it's bloated, that it's slow, ... and everybody wants to replace it. The best effort I've seen so far it Qt (especially Qtopia for palm devices).
I think X is like Unix : it was inadequate and bloated but computers have caught up with their demands, in terms of power and disk capacity.
Computers get more and more powerful, networks are faster and faster, and X is more and more lightweight comparatively. Combine that with decades of testing and millions of developers who have experience with it, and I can guarantee X is there to stay and evolve.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
There is no way to get consistency in a window system. People will port their favorite window managers and toolkits to whatever window system you create. MS Windows runs many of the same toolkits that X11 does. Apple is even worse, officially supporting OS 9, Carbon, Swing, and Cocoa-based applications on the same desktop, and now also X11; and in addition to all that, toolkits like Gtk+, FLTK, Swing are also being ported to native Quartz backends.
If you want consistency on your desktop, just choose to use a consistent set of applications. Most non-computer experts can't even tell the difference between an MFC, Gnome, KDE, and wxWindows application: they all look equally flaky and confusing to them. And most people incorrectly think that something is an OS X native application if it has shiny gumdrop buttons. In short, most people neither know nor care.
Read the paper. It is of shockingly good quality, both in the writing and the completeness of ideas. The writer is a college senior!
This is the strength of the Open Source model. The source is out there for everyone to see; think you can do it better? Step up to the plate and take a swing!
And if his Y takes off, it can replace X withoutht causing you any problems at all at all.
I'm not really sure, in truth,how your post is relevent to what's going on, and i'm a bit put off by your sarcastic whine about gamers etc.
All to be followed by a (rather poor) backhanded attack on windows. Except this isn't about Windows with a capital w.
In short, what are you talking about, and who on earth modded you up?
Old truckers never die, they just get a new peterbilt
This guy seems to know what he's talking about and as far as I can tell he's got a proof of concept to show allready. Along with solid research and design.
I wouldn't be to fast at hand with bashing this guy - he lists all the other XFree replacements and for some like Berlin/Fresco he can clearly state why they failed and what you have to aviod to not fail the same way. And he also acknowleges XFrees benefits and sees no point in overthrowing them.
Keep an eye on this project, this could be something really interessting.
We suffer more in our imagination than in reality. - Seneca
Come now. Windows has shipped with a standard widget toolkit since its very first version, yet it has definately evolved since then.
Don't assume that had X got a built in toolkit, it would never have evolved. Given the extensibility of X it probably would have evolved nicely, in fact.
However, likewise you shouldn't assume that if X had a built in toolkit everything would use it. Of course, this would not be the case. Mozilla would still use XUL. OpenOffice would still use the VCL. Wine would still use its clone of win32 widgets. Some apps would still reinvent widgets for whatever reason (just as they do on Windows and MacOS X).
Toolkit compatability is best dealt with by standards, IMHO, rather than just moving them around....
The choice in a tradeoff between flexability and marketshare in a product that doesn't require marketshare to survive should be obvious.
What you're suggesting is that instead of having complaints and working toward a solution, we should simply ignore the faults.
No, thats definently NOT what I'm suggesting. I'm stating that X is NOT slow. I've meet quite a lot of people that have heard some place, that OSS requires less CPU power than Windows, so they load RedHat 9 on some poor old Celeron 300a, turn on all the eyecandy on KDE, and expect it to perform like the previus Windows 98SE installation. Then it goes: X is slow and bloated. Let me see... It has network transparency?! hmmm.. windows dont, so that must be it!! lets remove network transparency, then it'll be all better."
The fact is, X is an old protocol, designed to work on slow machines, and KDE3 is not supposed to be compared to Windows 98.
X does not need to be replaced.
True, but Linus didn't plan it as the next big thing that replaces Unix and Windows. It turned out quite well after years of work by many contributors, and the same thing might as well happen to Y. It's too early to predict anything, so in the meantime it's probably best to do as Linus says: forget about competition and just focus on writing good software.
Escher was the first MC and Giger invented the HR department.
The arguement here seems to be not that the inconsistency between who puts what tools onto a given system puts off the casual users, which tbh is a fair point.
Frankly, OSS is generally not about making Joe Sixpack happy. Most OSS software is far better than the closed-source equivalent...for techies.
I'm not really broken-hearted about the state of affairs, either.
May we never see th
You are making a common mistake. Just because someone is not skilled in the areas that you are, doesn't mean that they aren't skilled in other areas. Just because they have a hard time learning what you find easy doesn't mean that that don't easily learn many things that you find difficult to impossible.
E.g.: My wife has, after years of sporadic effort, finally learned that files are not stored inside of the programs that create them. I think. But she can pick up a new musical instrument and with a couple of hours practice play reasonably advance music on it. Not just scales, and not just strings. She specializes in ethno-musicology. Some things she handles well in an hour would take me years to do as well.
But with a bit of guidance she is able to handle ordinary WordProcessing, Graphics, and Music Composition programs. (The only problem is that she tends to save files in random places, and not understand why. Or where. I'm still working on trying to get her to understand disk folders.)
People have radically different skills. Learn to enjoy this. Or at least accept it without shouting. Its the people with different skill sets that have the most to offer each other.
I think we've pushed this "anyone can grow up to be president" thing too far.
> That's still one toolkit too many.
Both Gnome and KDE are based on this "One True Toolkit" ideal, which as everyone knows is bogus and extremely unlikely to happen anytime soon. It's just not going to happen.
It's time for some pragmatism on the Unix desktop. Accept the fact that there's going to be 20 different toolkits and start standardizing 'themes', keyboard commands, clipboard handling, and everything else.
Windows users don't know or care if their app was written in MFC, XUL, Borland, or Eclipse. That's because all of those toolkits agree on the same behavior (MS UI Standards) and share the same configuration (registry).
On UNIX, however, the developers are "in your face" with their toolkit choices as part of the marketing effort for their "one true desktop". gUsers don't Kare what gToolkit you Koded too, asswipes.
unless they are such complete morons that they shouldn't be allowed within spitting distance of a computer...
It's offensively mindless ramblings like that that are a) keeping Linux out of majority desktop share and b) keeping the sterotype of the arrogant geek strong in the hearts and minds of millions.
Computers are not important enough to serve as the deciding factor of one's intelligence. I would argue that social skills are a much better yardstick, and by using it on your post above, I see that you shouldn't be allowed within spitting distance of another human being (loosly assuming that you are one yourself).
Computers are just pieces of equipment that some people understand, and some don't. Could you rebuild your car's engine from scratch, or do you take it to a mechanic? Does he make you feel like an imbecile because you don't know how to tune your own fuel injectors?
The norms need some kind of consistent interface, and Linux (or *nix in general) just doesn't provide that. Yes, they can click through a Redhat install (doubtful as one of the first things to do is partition the drive, and I can't see my mother doing that), but it'll be different enough from what they know. And, then, what if they want to install a new piece of software? Do they have GTK? Probably. QT? Again, good possibility. OCaml? Ahhh...I don't know...and they sure as hell aren't. (I know, not a toolkit, but an example). All they know is when RPM tells them they need it, they won't know how to get it. And, considering how well written 90% of the install docs are, they'll never find out.
Which is more painful? Going to work or gouging your eye out with a spoon? Find out!
http://www.workorspoon.com
The thing is that you're assuming that all widget sets provide the same functionality (or that you can go with the least-common-denominator). In reality, that'd make for some pretty awful applications. You'd have to give up your KDE-draggable-panes (not supported under Win32), your sliders (not supported under classic Mac OS), gtk's easy layout system that provides automatic resizing support (Win32 and classic Mac OS use a pixel-based layout system). Some widget sets use infinite progress bars (XUL), some use animated icons (Windows). The two can't fit in the same space.
You're thinking of something simple, like theme engines. These *are* pluggable, and plenty exist for gtk. You can't just drop in a new widget set, though.
Xlib *was* designed so that it's easy to develop widget toolkits, though.
May we never see th
Well, what you're saying is a nonstarter.
When people say "we want one true toolkit" they are users saying they want one user interface. If someone standardizes themes, keyboard commands, clipboard handling, etc. then what they've provided is the one true toolkit.
Of course nobody cares what code was used to generate the app. Nice meaningless observation. What people want are the applications to behave similarly regardless of who programmed them using what and to have a universal look and feel. That's the one true toolkit.
Well, I've only gotten as far as the critique of X at the start of Thomas' paper. This critique is a classic /. "X sucks" troll in academically semi-polished form.
Point by point:
- X has too much latency: See Packard's paper in Freenix 2003: high latency is not an inherent X attribute, even over high-latency connections.
- X requires a toolkit for ease of programming: Duh. As opposed to what, exactly?
- X needs standardized UI semantics: This is moot. You may use X+Gnome or X+KDE if this is what you desire. Either is a fairly good and fairly complete system with standard UI semantics. The existence of two such systems is no more troublesome than the simultaneous coexistence of Windows and MacOS.
- X is "an incoherent mess": When making this argument, it is always useful to confuse the protocol with the implementation. The existence of Kdrive is a nice example of how much the latter can be cleaned up. The protocol hasn't changed in 20 years except for extensions: the argument that the extensions don't work together is supremely unconvincing, supported by one lame example. freedesktop.org has made a lot of progress in a short time in refactoring and standardizing X.
- X is complex: It is. Unfortunately, it is a response to complex application requirements. Again, one lame example involving perhaps the most demanding application running is cited. And again, freedesktop.org is standardizing mechanism for dealing with the cited problems.
All of this would have been easy to avoid. Just talk to an X developer before publishing the paper. Many are quite accessible, and would undoubtedly have been happy to correct the critique.It's perfectly valid to want to write a new window system. I can think of a variety of justifications, starting with "it's nice to try something different" and "I wanted to learn some things". Trolling is hardly necessary, and hardly welcome.
The same is true for pretty much everyone. Sometimes people assume that users who can't understand basic concepts and models in a desktop environment lack some sort of special brain power. The truth is, if they did lack that "brain power", they probably wouldn't be able to handle understanding things in daily life like:
- just because you place a piece of paper on top of another piece of paper doesn't mean the paper underneath is gone
- the type of paper and pen you used to write a letter has no effect on where you decide to place the finished letter, whether it be in a drawer or on a different desk
- if you write a note that says "go get an apple from the fridge", and you throw away the note, the apple is not going to disappear
Point is, it doesn't take a genius to figure out basic concepts used in desktop environments, (after all they were made by people, for people, right?), althought it might take a genius to make a perfect desktop environment. People just need to take 5 minutes or so to explain things, and everything will be clear, to everyone.
The laggyness of Linux GUI is primarily a kernel problem. Expect it to improve vastly with 2.6.x kernels. See the new post on kernel 2.6.0-test6.
I see some flaws in the document posted.
I love research, and more than anything I love the people involved in doing research: those who create, explore and can give us future direction. But I also believe that the research must be truthful if we want to build on it.
The Y presentation paper is interesting on the ideas it introduced (we could argue whether they are new or not, since NeWS did did before, in fact, with a more extensible system) but it fails on presenting X correctly.
The document goes on to show that the X-based approach has lead to major GUI fragmentation, and how the MacOS and Windows do not have this problem.
On the screenshots where X looks bad, the author shows some old graphics program running together:, xpdf and two modern apps: mozilla/xul and gnome calculator. All of those programs have Gtk-based or Qt-based equivalents that would have made the whole experience consistent.
The screenshot should instead be presented as a proof that X can still run applications that were developed 12 years ago.
Then he shows the Mac and Windows. Again, not really honest screenshots, because even Apple is shipping two different GUI views: the brushed metal theme and the aqua theme (this combination kills me) and Microsoft is not exactly known for keeping their GUI look consistent across their product line: Office, MSN and the rest of the desktop use different styles and widgets.
So summing it up: the screenshots are presented to prove a point which happens to not be there.
Now, to make things even more interesting, here is a little bit that the author of Y might not be aware of: widget rendering on MacOS X happens on the client side, and the operation that the server supports is basically "uptade-rectangle-with-this-RGB-buffer", there is no magic of server-side widget rendering on MacOS X.
Also, doing an X protocol translator is not an easy job, but I wish them good luck pursuing this new adventure, it defintely sounds interesting.
Miguel.