Slashdot Mirror
Y Window System Project Started
cuppm writes "Y, Mark Thomas's final year project for his masters degree, is back in active development (outlined here). Here is the email I received: '...Y development is about to start up again. If you are interested in participating, the website is at: http://www.y-windows.org/. There are links to mailing lists there, and you can download the latest development snapshot, which should compile this time :o). I apologise if I did not respond to your email personally. I was on holiday in Japan when the story broke, and by the time I got back I had over 80 emails about the subject, many of them in depth. If you had specific points that you'd like to raise, I suggest re-raising them on the y-devel mailing list.' So for all those who think it's time for a X replacement, here's your shot. And for those X lovers, use Y's extensibility to make it X compatible." See our previous story for more background.
I find it highly unlikely that I would consider another future desktop additions. It would be more prudent to patch and hack on the labors that have already been provided ensuring both stability and security before adding other extensions. The whole damn planet has gone desktop happy.
I may be new here, but will someone give a quick rundown on what exactly Y windows is and how it is different or is an improvement over X windows or any other windowing system?
"When the only tool you own is a hammer, every problem begins to resemble a nail." - Abraham Maslow (1908-1970)
Y-dash would be more internationally accepted, Windows being trademarked and all. I'm surprised X can be used internationally, given the recent Lin---- news.
Would you care to explain why the new X licence is better then the GPL licence (which is what you imply)? I don't feel I know enough about the subject to comment either way, but when making a contentious (on /. at least) statement like that it'd be helpful if you could explain your reasoning! (And I'm curious anyway ;-) )
Thank God. Finally, someone has decided to quit bitching about X Window and finally implement a system of their own.
For any programmers out there that are even remotely interested in getting Linux On The Desktop, consider this a call. A super-awesome rock solid kernel cannot be the end-all be-all for Linux. We need to have a good windowing system, one that's faster and more reliable than the competition. From what I know, X Window could use a great amount of improvement in those areas. This is your chance to make things better, and Get It Right The First Time.
--Stephen
Did you ever notice that *nix doesn't even cover Linux?
Fantastic. New users find the selection of different toolkits for X confusing and inconsistent both in appearance and behaviour. One standard toolkit will help with newbie usability greatly - though whether it will stand the test of time remains to be seen. Windows seems to be doing just fine with it's standards though, so I rather suspect the same will apply to Y.
There is nothing like a little competition to hot things up - perhaps this will also give the languid Xfree86 project the kick up the backside it needs.
I wish the Y project the best of luck!
biopowered.co.uk - catalytically cracking triglycerides for home automotive use since 2008. Just say no to big oil!
So why is it going to succeed where these failed? :
fresco
YAX (Y Ain't X)
The Y Window System
Oh never mind. What's the point?
Stick Men
Apache2's default configuration doesn't scale gracefully to the load generated by a slashdotting.
I've upped various magic numbers in poolsize.conf and it appears to now be responding much faster.
Cheers,
David
Actually, I'm guessing it was a Masters. The way it works at Imperial (where Mark, and I, went), you can do either a 3 years course and walk out with a BEng (Bachelor of Engineering) or do an extra year and get an MEng (Master of Engineering). Both are "first" degrees, and so might be called Bachelors in some parts of the world, but you get a Masters certificate, so it's a Masters. The more traditional way of getting a Masters in the UK is to go back to university some time after completing your original Bachelors degree and do a short (1-2 years) "conversion" course, usually in a different subject. At Imperial, people who do this can be distinguished because they get an MSc (Master of Science) rather than an MEng. In my experience they also tended to be french. But I'm not sure why that was ;)
---- Den ene knappen er powerknapp, den andre er Bender voice knapp "Bite My Shiny Metal Ass"
I think a lack of standardization/uniformity is what's hurting the Open Source movement. You can't get that many people together, and have them agree on -one- way to do things. Everyone is out to push their own little twist. The result is a bit chaotic to anyone looking from the outside. (And to some from the inside as well, I'm sure.)
Hate 'em as much as I do, the one thing MS has done well is ensure compatability. Obviously there's problems; but the basic principles of windows applications are near uniform. I don't think you can say the same for a lot of OSS. Chalk it up to people being sheep, if you want, but until there's one clear leading force, Linux (sadly) won't succeed on the desktop.
About a year ago, I had started work on something I called YX (yes, the pun was intended). It didn't get very far, I'm glad that someone is working on such a project. I definitely intend to help with this project, though.
I have discovered a truly marvelous
So we are not happy with the X licensing hmm?
And now we have;
o Xouvert
o XDrive (sp? rename?)
o Berlin (rename?)
o DirectFB
o Y
Could someone Please enlighten me why y? and none of the other?
If someone is so knowledgable to do a comparison I'd glady appriciate it.
-mo
a)it looks like the only reason development started again was because of all the Xfree86 licensing hubbub(which isn't going to be around much longer, because Xfree86 will most likely cave). If the project did not have the merits to succeed before, I do not see how things have changed in such a way that it will be successful long-term, and this was a blatant "look at me" attempt. Y was dead, FreeDesktop was humming along quietly.
Did you read the bit at the top? "I was on holiday in Japan when the story broke" means he was in Japan on holiday, and so couldn't start dev then. Also, he wrote it for a "final year project for his masters degree", which doesn't sound like a look-at-me-attempt, but more of a I'd-like-to-pass-my-course attempt.
c)We already have an interesting, viable alternative(FreeDesktop)...and it's got heavy involvement with the major developers of Gnome and KDE, the two most popular desktop systems. Everyone is playing Chicken with Xfree86, while hedging their bet(and strengthening their position with Xfree86) by starting work with FreeDesktop. Y is nowhere to be seen in all of this, especially if it's only got one guy- versus a whole group of some of the best Linux programmers around.
How can you play chicken with an OS project??
Anyway, Y is nowhere to be seen since there was no centre for dev until yesterday. The mailing list has only just become active, and now it is there are lots of people interested in helping develop this. It's worth mentioning that the closed nature of development for X means lots of people are looking for something where they can have more of an impact and really get involved, which is exactly what Y can provide them.
Okay, it's not an X killer yet by a long shot, although Linux never started as a Unix killer, and look where it is now!
Howdy.
You make some reasonable points.
A development restart has been planned for months; the only reason it hasn't happened sooner is that we've all been settling into new jobs and simply haven't had the spare time to get this going properly until now.
X Windows *does* have issues; I think we can all agree on that. But by the same token, we're not trying to argue that X is not useful; I'm using XFree86 on my production machine right now to good effect. But we think it can be done better.
Linus was just one guy when he started work on Linux. Other people then joined in, and made Linux what it is today.
Mark, myself, and the other chaps who were in the room when the Y concept was born are doing this because we enjoy it. Whether lots of people will join in on our little project remains to be seen.
Sure, it'll be gratifying if we become popular, but that's not what we've set out to do -- write good code.
Cheers,
David
Is it just me, or is their logo surprisingly similar to that of GTK+?
You are simply citing the differences between OS and any company.
In Open Source Development there is a Naturally driven variations. Think if it as leaves driven before the wind. Eventually most of them end up in the same place.
With any company, you do as the boss says or you're toast. Any questions?
I think there is a lot of merit in having variations in WindowManagers. I will fight that to the death. But when you have to apply layer upon layer of Glue Code to get some really useful, it implicates a problem exists. And when the various solutions are all inconsistent and independently parallel to each other, you have another implication of a potential problem.
If done correctly, most of this new code implimentation wouldn't require a visual (user aware) change to any of the existing Window Managers. However it might provide for a more consistent approach so that all buttons, labels, etc. appear the same. Today that doesn't exist unless you choose to use only a certain base library for your graphics (eg: Qt)
For Y to be remotely usable for me, it would need good support for OpenGL on nVidia and ATI graphics cards...for which (annoyingly) we only have binary drivers.
So - my questions would be:
1) Can Y use GLX protocols and work with existing (binary only) OpenGL drivers?
2) There is mention that Y can use hardware accelleration on 3D hardware. My concern about this is how much of the valuable 3D resources such as texture map memory it consumes. Generally, X runs plenty fast enough without using those resources and I wouldn't want to impact my 3D capabilities in order to make the 2D windowing system run ten times faster than it really needs to run.
Certainly X needs updating - it's old and it shows it's age.
www.sjbaker.org
As long as it's network transparent I don't care either. If I can tunnel apps over my SSH connection from one box to another then it's pretty useless for me no matter how fast it is.
Maybe each of us is thinking something different, but your comments disagree with what I'm seeing.
Current remote X server: 272 MB. Remote (tight)VNC server: 60 MB.
Local X server: 161 MB: Local (tight)VNCviewer: 4 MB.
vnc here runs faster than running X remotely. my guess is that it's because VNC is mostly client-pull/server push oriented.
To-do List: Receive telemarketing call during a tornado warning. Check.
Which is a fair critisim because even though that was written 10 years ago, there's been no effort made to move optional extentions into the "standard".
One extra thing to consider would be configurability. My XF86Config file is a beast, since I have a laptop with three pointer devices, and S-Video and RGB connections out to the tv and projector. I can't reconfigure my video on the fly like I can with Windoze. Help!
>... Further, with Y's in-server knowledge
> of widgets, applications run over a slow network
> can appear almost as responsive as local
> applications (especially when compared to an
> X application).
Great idea - is this the same thing the are doing with Fresco? Fresco also has SVG.
> Y widgets use the currently loaded theme to
> render themselves. Since all server widgets
> are using the same theme, all widgets appear
> consistent throughout the desktop.
This may be an SFQ, but shouldn't the application choose its own look and feel? I have various Java apps, some use the Metal L&F and others use the Windows one. Why force me?
Here I sit back, reading slashdot on a pentium 166MMHX, with 80M of memory, through Galeon and the X Windows System on a OpenBSD machine.
I read the posts that say X is slow.
X is currently using about 5% - 7.5% of my processor. It jumps up to about 15% when I change windows. MPG123 consistantly uses more CPU then X. Galeon tends to use more CPU then X as well.
I read the posts that say X is bloated.
X is currently using 15MB of memory/8MB resident. Galeon is using about 16MB / 27 MB resident.
As for hard to set up, linux distros usually set up X for me. There are even several configuration utilities shipped with XFree86.
I also tend to use the network transparency of X, which is easily accomplished through ssh -X.
Don't know why you guys keep having problems, but may I suggest bloated OS installs and bloated WMs?
FVWM + XFree86 works for me!
Your statement is based on an implied statement something like this:
"Most people do not often need network display in their windowing systems. Most X users dont use network display."
To which I respond, "Most of the peolpe I know who use X, use remote display daily. I personally use remote display daily. One of X's biggest strengths is remote network displays."
I am not necessarily a valid statistical sample and I am well aware of it. For MANY people, your statement is true, but there are a LOT of other people for which it is false.
On the other hand, your point that having a local-only core with a remote module for those who need it is ok. I agree with you as long as the protocol directly supports it and I dont have to have special software on the client but rather I just have to add a local module to use it.
Free Online Dark Fantasy RPG - http://www.blackmud.com
I was at Stanford during the development of W/X, as a user of early version, but not a developer of X. It was a time of rapidly evolving technology, so some standards they guessed right, while others were kludges. The technology was the workstation, a computer small enough to put into the office (the size of half-height file cabinet) and enough power to run UNIX. PCs were way too underpowered to run UNIX and bit-mapped graphics. Apollo (absorbed into HP) was the UNIX king, but all its standards were proprietary. Sun was just a couple years old and its standards were half-open, half-proprietary- a practice they continue to this day. DEC (absorbed into HP) was willing to tolerate UNIX on its min-VAXes, but not write all the missing parts- especially window graphics. So they essentially delegated that to Stanford and MIT with hardware and R&D grants. So there was a lot of R&D then on how to do client-server computing and graphics.
The primary problem at that time was the availability of a suitable object-oriented programming language. Everyone knew that was the future of software. The UNIX crowd preferred something related to C. C++ was very unstable, while ObjectiveC, based on on SmallTalk, was good but proprietary. The fledgeling company NeXT (in the Stanford industrial park, later absorbing Apple Computer) decided on ObjectiveC. The Stanford W/X group decided to use neither of these but invent a quasi OOP extension to C in the Xt Toolkit. And XWindows has suffered ever since.
I kind of like the way that X calls are low level, and that frees people to build nice widget sets like GTK and QT on top of those base Xlib calls. That's something that X does right, IMHO. And having the network transparency at that level is good too. However, I find Y's idea of a network transparent widget set intriguing, as I don't yet have a high-speed connection at home. I'm not saying get rid of network transparent Xlib, but I think building a network transparent high level widget set could make applications be nicer to run for me over a dial-up modem. Granted, apps would have to be ported to this new widget set, but still, I think it could be worth it. Maybe the X guys could pick up this idea.
TZ
Actually, building AA into the client isn't as hacky as you'd think. What Xft does is implement a generalized compositing mechanism in the X server. One of the many uses of this compositing mechanism is AA'ed text. Ergo, keeping the AA mechanism on the client is actually reasonable, because other uses of the compositing mechanism will also be in the client.
A deep unwavering belief is a sure sign you're missing something...
Does this mean it could be ported to Win32, BeOS and AmigaOS?
Excerpt taken from "Y: A Successor to the X Window System" PDF, Section 10.3 Future Work.
Legacy X Protocol Handler
"In order to support the wealth of X applications that already exist, and to ease
the transition from X to Y, an interpretation layer will need to be built.
This is an excellent example of the elegance of the design of Y. The X layer
can be implemented as an in-server driver module. This module would, upon
initialisation, create an appropriate socket to pretend to act as an X server.
When X applications connect to this socket, the X module would translate the
requests into equivalent Y requests.
One drawback of supporting the X protocol is that many of the advantages
of Y, in particular the lightweight protocol and server-side objects, will be lost."
From what I've read, it's exactly what I want (and have been advocating.) My money was on PicoGUI, but hey - competition is good.
Ah, you kids. Sun did this years ago, with NeWS, Network-extensible Window System. It used DPF and rendered widgets locally. It never caught on because what wins in the mass market is the lowest common denominator, which was X. Sure, it sucked, and everyone hated it, but it was as vendor- and architecture-neutral as windowing systems got, so every Unix vendor picked it up. For the same reasons, the Y project will never be anything more than an academic curiousity.