Post a Comment
For the actual talk:
http://ktown.kde.org/akademy2005/unprocessed/BeautyMagicKDE-ZackRus...
As the URL says, it's unprocessed, so the file quality/size ratio is not optimal.
For the demo, check the xgl_wanking.avi linked in the news here.
That stuff looks pretty iimpressive. I wonder what graphics card he's got in that thing? It certainly looks more fluid than the effects on my Mac mini.
Like he says, the trick is to make USABLE cool effects. I don't like the transparency stff, generally a very bad idea. But the wobbling of the windows could be kinda useful if toned down.
ooo this is my favorite thing to say:
When it's ready.
Plasma is still pretty young, I wouldn't expect code from them for 6-12 months. And I wouldn't expect a product for I think they estimated 18 months?
http://wiki.x.org/wiki/X11R6970ReleasePlan#head-0a8adac9d6d5f4d14ee...
But that should give you a decent idea of what to expect from x.org. R6.9 and R7 are the same thing btw, just that R7 will be built differently than R6.9 which will be built the way Xfree has been since forever...
This was great. It may not be of any use, but people will find a way to use that power to create something that is both usable and good looking.
But what's the deal about the <em>filename</em>? I had to delete the browserhistory just to avoid anyone from seeing that. On the other hand it might be the correct name. Just people getting off on cool effects.
I have seen similar demos and the cpu doesn't get any more use than it does normally moving a window around. It makes things much more usable because there is very little hit to the cpu performance. The other cool thing is that XGL is not doing something like taking a picture of the desktop during all the animations. For example, on the workspace flipping, if a movie is playing, it keeps playing fluidly the whole time throug the animations. Very cool stuff.
One thing to note is that EXA has nothing to do with many of the effects shown in xgl_wanking.avi. The really cool stuff is all Xgl. EXA will not bring rotating desktops and whatnot to a computer near you.
EXA will, however, help with dropshadows and transparent windows (now the composite extension can be properly accelerated). EXA will be ready for (many drivers) in the next Xorg release.
Two more things, David Reveman is the primary authour of Xgl and X[e]gl is probably some way of yet before it's ready for prime time.
I don't know when the X extension will be part of xorg
-----
EXA should be contained in the next X.org release.
Which will be within a few weeks from now.
Which version will be 7.0.
Which will be of a modular design.
Which will allow updates of video drivers without updating the *complete* X.org climborium.
Which will allow Zack to release EXA-enabled drivers independently from future X.org releases.
Which will bring us dozens of EXA-enabled "old" drivers within a couple of days after the X.org-7.0 release.
Which Zack promised.
Which I believe is true.
I know some older people who really lean on their computers to communicate with the outside world. Email is huge for them. It's very hard to overstate the importance of family emails to a woman in her 90's who can't really go out and about.
The problem is that older people often have trouble reading text on the screen. Tech like this that can zoom in and out fairly arbitrarily would change their lives, if someone could intergrate it into the interface in a reasonable way.
"The problem is that older people often have trouble reading text on the screen. Tech like this that can zoom in and out fairly arbitrarily would change their lives, if someone could intergrate it into the interface in a reasonable way."
VERY important part that's often neglected in free software development. The developers tend to be young, and although developer/artist collaboration is a good start for free software GUI design, the artists tend to be young, too.
But many parts of the world (US, Italy, Spain, and more) have aging populations and accessibility features (especially for low-vision users) are critical for a growing number of market segments. Magnifiers, screen readers, virtual cursors... it all has to be there, not just for government contracts but for the home market as well.
If desktop linux ever hopes to have any sort of market penetration in the general consumer market these effects are needed.
This is however not just "eye candy" as many of you would call it, A good looking and easy to navigate desktop increases productivity, something that is easier on the eyes and easier to read is not as difficult to focus on for an extended period of time as something that is less visually pleasing.
Smooth antialiased fonts and drop shadows do believe it or not increase productivity. For instance, a well implimented drop shadow will draw your attention to the maxamized focused window better then the same window with no drop shadow this in turn keeps down eye strain and keeps attention focused on the work at hand.
just a little food for thought... I read an article about this some months back and I implimented it to see if it really made a difference (I was always a plain Jane no special desktop stuff kind of guy) and to my amazement I really do now seem to be more productive and I navigate my desktop faster than I ever have before.
"This is however not just "eye candy" as many of you would call it, A good looking and easy to navigate desktop increases productivity"
I disagree. I think most of this IS just "eye candy." In my experience, if you are looking to increase productivity, you have to find a way to pull heads out of asses. Nothing short of that really works.
Eye candy is a side effect of the underlaying techology on a OpenGL driven desktop.
Besides the aforementioned eye candy, people can develop new clever ways to achieve new functionality like realtime smooth screen zooming, as someone pointed out anyone with vision defects will seriously appreciate the stuff.
I'm pretty sure it's like that in a lot of computing- there's always a faster computer or better graphics card or better 3D graphics engine... Windows just happens to have not had a release in four years. But Pixel Shaders 2, all the new Longhorn improvements, the Doom/Quake engines, the speed limit on the Pentium 4 they ran into... all those happened since Windows XP was released. It's just you may not have seen it.
In Linux there's more hoopla because the graphics engines and all manner of new ideas are being done by actual people and groups amongst the community.
This guy talks like he is up himself. would be funny to see him humiliated by a MS programmer
Well having read through some of Zack's code in the past and having used some of his software, I think the chances of that are about the same as pigs growing wings. Zack is the real deal, and that's why he's doing what he's doing at Trolltech these days. If that scares some MS oriented people around here then so much the better.
Besides, if you think he's up himself you might want to talk with a died-in-the-wool MS developer someday. You can't listen to the person for more than a few seconds without running off to the toilet to be sick. That's why I don't read any of their so obviously fake blogs. I'd be ill for several days.
I must admit, this is the first time I've seen something from a core linux team that really goes beyond what other desktops are doing. While some of it may seem gratuitous, things like "pullable" windows that deform when you drag them could really heighten the end-user sense of what they are doing. Using a cube metaphor to flip between desktops is both brilliant and should help people to get a grasp on what is going on (normal desktop switching has no real-world parallel, which seems to make it difficult for beginners to understand). The whole zoom-in thing I could see quickly making people lose stuff left and right, but for people with disabilities it could be quite helpful. The little "pop" as things lock into place is a great way to signify a state change. All of it gives the user a heightened sense of what they are doing, which makes interacting with the device easier. Plus, it really does look extremely cool, and will help to sell Linux to users.
Honestly, even though I'm currently surrounded by basically every OS known to man, I'm still a touch jealous. I can't wait for this to go into the distro I use... which is umm, Debian... crap. Xandros, anyone?
All whiz, bang effects aside, I'm eager for the day when the standard, out-of-the-box/tar linux desktop is free of graphical tearing and redraw glitches. I don't think a smooth, double-buffered experience is a lot to ask .... and once it is in place, the combination of Blender/Gimp/Scribus/Inkscape etc will make Linux a seriously attractive design platform.
Now that I think of it; is there a distro targeted at graphical designers?
Looks to me like Archlinux keeps up with scribus, gimp, blender and inkscape very well. If I didn't misread anything, they've got all the latest version in the repos right now...
Arch certainly isn't targeted at graphical designers, but it seems to keep up with their favorite programs and it's all well optimized i686 packages which allow you to VERY easily rebuild with different build options in cause you wanted something enabled which they didn't have.
Not that this is the time or place (nor has this ever been an issue in my posts), but archlinux is a damn fine distribution. It takes a lot to satisfy the most hardcore of gentoo users. But with arch I get everything gentoo gave me plus the convenience of binary packages and a more convenient build system. Something tells me technology like Xegl will be in AUR before it hits the portage tree.
So who's coming with me?
i tried to donwload the avi file with konqueror or firefox. and i can't manage to.
I finally arrived to a page to upload a file whereas i wanted to download one.
what a wonderfull commercial service rapidshare is .. it doesn't simply work !
If you want to try a second time because it has failed, you must have a second IP... sucks ...
Obnoxious isn't it? But it's about the best you can do aside from setting up a torrent or asking sourceforge to lend you the bandwidth. I don't think sourceforge has a limit on package size, but they may pull your 50MB video if everyone is downloading it...
I really wish they'd remember your ip when you start the download instead of when you get to the download start page. But hey, they want people to pay for the service.
It might be smarter of them to force free users to say look at 10 ads; and have the uploader choose the ad category to target. So this guy could have said "linux ads" and then those advertisers would get some good viewership, probably some clicks, the site would get some money, and the downloaders wouldn't have to deal with obnoxious stuff.
Btw, it forgets your IP after an hour
.
Due to several individuals' complaints about the site on which the video is hosted, I made a torrent of the video and will be seeding it for a while:
http://thepiratebay.org/details.php?id=3381636
I hope that this will help a few excluded individuals feel better.
http://kompose.berlios.de/
Try that.
> http://kompose.berlios.de/
> Try that.
That's pretty good, now only two minor details
would be nice to improve on:
1. "performance of Komposé depends heavily on your CPU."
rather: performance of Komposé depends on your GPU!
2. "every window is represented by a scaled screenshot"
rather: every window is running in real time!
(quotes taken from the kompose site)
"
1. "performance of Komposé depends heavily on your CPU."
rather: performance of Komposé depends on your GPU!
2. "every window is represented by a scaled screenshot"
rather: every window is running in real time!
"
I use Kompose and have followed it's development a bit. Basically those things are planned, but the author(s) is suffering a bit of a 'forced' chicken and egg type problem. Can't do "these things" in Kompose because most xservers / video cards would wind up rendering in software, Kompose isn't an Expose killer cause it doesn't do "these things". Kompose is just a smidgen ahead of the state of the linux desktop, I rather appreciate that the author(s) haven't rendered the program unusable for the majority of "Today's X Users" just to be on the bleeding edge... All of these things are in the cards (and probably source repository as well), it just doesn't make sense to force them on people before Xorg 7.0 & updated drivers becomes ubiquitous (well, for those of us that like to use the program today, anyway).
I am very excited about the innovative improvements planned for KDE. More importantly, I would like to see font handling be improved. The implementation of font smoothing is not as good as it is in OS X or Windows XP and I believe that this is a fundamental issue that must remain a focus of KDE developers. I must contend that until they are handled as well as they are in OS X or XP, KDE will not be "state of the art".



