Web GL
category: code [glöplog]
Personally, I've found WebGL to have severe compatibility problems... they tend to be (in my experience, which might not be a lot) way worse than nVidia/ATI problems, since they include BOTH the nVidia/ATI ones, and the browser ones.
Plus, it's for some reason assumed by the general public (not by the sceners) that any hardware capable to run a browser should be capable of rendering webpages (that includes webgl ones). Once you start asking a browser for "minimum requirements", the advantages of WebGL as a distribution platform start to fade...
Add that to the obsolescence of its features, and I don't really see the benefit for it, unless they start working hard and -fast- on bringing it, both up to date, and compatible.
Plus, it's for some reason assumed by the general public (not by the sceners) that any hardware capable to run a browser should be capable of rendering webpages (that includes webgl ones). Once you start asking a browser for "minimum requirements", the advantages of WebGL as a distribution platform start to fade...
Add that to the obsolescence of its features, and I don't really see the benefit for it, unless they start working hard and -fast- on bringing it, both up to date, and compatible.
I don't think there's any point (or possibility) of us taking over other scenes, them taking us over, or merging or whatever. They're different, and the differences will keep them separate.
I think more connections between the different scenes would be very helpful though. Our scene could learn a lot from the outside world - in fact it'd probably help stop the scene stagnating. Other scenes could definitely learn from us :)
I think more connections between the different scenes would be very helpful though. Our scene could learn a lot from the outside world - in fact it'd probably help stop the scene stagnating. Other scenes could definitely learn from us :)
And on the webgl thing: so it's new, still buggy, and the capabilities are limited. So make something simple (but still awesome). Or don't make anything at all. Some people like it, some don't. Moving the whole scene to webgl would be just as stupid as moving everything to PC.
But it's important to keep things open to new platforms too, and the web is kindof important these days ;) So is mobile, but we still see some demos on ios and android without huge amounts of bitching over compatibility (which on android at least probably puts webgl deep in the shade), outdated capability etc. It is what it is, and it's possible to make cool shit on it.
But it's important to keep things open to new platforms too, and the web is kindof important these days ;) So is mobile, but we still see some demos on ios and android without huge amounts of bitching over compatibility (which on android at least probably puts webgl deep in the shade), outdated capability etc. It is what it is, and it's possible to make cool shit on it.
What Jcl said, however not even the scene understands why some demos don't run on a tablet, or why they don't work in the obscurest browser, or on their smartphone. Hell you even get different results from the same browser on Linux and Windows.
And not even using THREEjs shields from those problems, especially when one does not only rely on WebGL, and does fancy things with divs.
It's just a fuckton more work to put out something nice, then to fix things on the various browsers, platforms and fight of people who start it in Netscape for the LULZ.
And not even using THREEjs shields from those problems, especially when one does not only rely on WebGL, and does fancy things with divs.
It's just a fuckton more work to put out something nice, then to fix things on the various browsers, platforms and fight of people who start it in Netscape for the LULZ.
Quote:
You say that as if it's a given that sceners fix bugs or compatibility issues in their current demos -- mostly, we don't. We release it, it runs fine for most people, usually. If it doesn't, well, there's the video on YouTube. Move on to the next prod.It's just a fuckton more work to put out something nice, then to fix things on the various browsers, platforms and fight of people who start it in Netscape for the LULZ.
I don't see any reason why we should get bogged down with compatibility issues just because the platform happens to be the web and not Windows. Also: have you ever tried running a Linux demo? :)
Quote:
You say that as if it's a given that sceners fix bugs or compatibility issues in their current demos -- mostly, we don't.
Well, I usually try to fix the bugs, with the last one we even updated some effect without even having to re-up the prod somewhere. Because if I don't fix the issue there, it'll occour in the next one again. Seems like I'm not part of your scene then, ey?
mog: you seem to be describing the exact same flow (distribute, error, possible fix, re-distribute) that would occur if you got feedback on a web-based demo then, except you would get better error-reports faster in the latter case, plus that it would be much easier for you to both benchmark and silently collect test data. Therefore I don't really see any validity in your argument -- oh, and that childish attitude there? Jeez.
Those webgl compatibility issues? They're often fixed by changing browser - which generally means 15 seconds of typing in your current browser and a 2 minute download.
Compare that with the usual compatibility issues with demos, and the usual fixes: upgrade your computer, switch your GPU for an nvidia one, update drivers, download the latest DX runtimes.
Surely demo compatibility issues are worse, no? And if you're about to disagree, why doesn't your latest demo run on my mac? :)
The "it's out of date and seriously limited" argument is valid, definitely, but only for those who like pushing the hardware to the limit. A lot of people just like to make something cool instead. I mean how many here have picked up a high-end calculator and thought "cool, lets make a demo on it"? Or fired up some new programming language, and made a quick demo effect?
There's a whole range of opinions, and for a lot of people webgl is as valid as a high end PC. It's just another platform, with its own pros and cons.
Compare that with the usual compatibility issues with demos, and the usual fixes: upgrade your computer, switch your GPU for an nvidia one, update drivers, download the latest DX runtimes.
Surely demo compatibility issues are worse, no? And if you're about to disagree, why doesn't your latest demo run on my mac? :)
The "it's out of date and seriously limited" argument is valid, definitely, but only for those who like pushing the hardware to the limit. A lot of people just like to make something cool instead. I mean how many here have picked up a high-end calculator and thought "cool, lets make a demo on it"? Or fired up some new programming language, and made a quick demo effect?
There's a whole range of opinions, and for a lot of people webgl is as valid as a high end PC. It's just another platform, with its own pros and cons.
gloom: You made a point and I explained that I see it different - and your "mostly, _we_ don't" is like forcing everyone else to nod along to your opinion. As stated earlier, I try to make the stuff work were necessary, if it doesn't run on a friends Opera - I'll surely fix it.
And I'd rather not spy on fellow demosceners to get information i can get easier, and more throughout on IRC. Especially when one can't measure if people see all models one placed on the scene.
And you're very much welcome to try the world of JS/WebGL demos, were all the memories you had from building websites will come back to you: "doesn't work on IE6" - it's fun, sometimes(tm).
And I'd rather not spy on fellow demosceners to get information i can get easier, and more throughout on IRC. Especially when one can't measure if people see all models one placed on the scene.
And you're very much welcome to try the world of JS/WebGL demos, were all the memories you had from building websites will come back to you: "doesn't work on IE6" - it's fun, sometimes(tm).
Lmao, this thread..
Here, have some WebGL stuff!
http://apps.playcanvas.com/will/doom3/gangnamstyle
Here, have some WebGL stuff!
http://apps.playcanvas.com/will/doom3/gangnamstyle
oh my, lol.
also, not a demo - it's totally out of sync!
also, not a demo - it's totally out of sync!
Ripped graphics, ripped music! Only one effect! *thumb down*
okkie: thats an awesome demo! just missing some shadows!
Fun fact, it has shadows :D
not on my browser. :)
qed.
lol
lol
that gangnam style is clearly missing some sine scroller and rob is jarig music :)
haha that gangnam style track is really funny, where is it from?
ok bad trolling attempt was bad but srsly, im so fed up by that song.
this thread is now about gangnam style ripping typical benassi's basslines.
Oh god, filipe, why did you have to show me that? WHY?
Hmm... Different views and opinions aside, I see mostly speculation about pros and cons of WebGL as a worthy demo platform. Why not do it for real instead? As with most technologies, WebGL (and more generally, the web platform) requires real applications for it to mature. People like Mr.doob have done a great job for WebGL in the non-scene arena, raising awareness, innovating, pushing limits, and generally helping/forcing browser vendors to stabilize, standardize and optimize technologies such as WebGL. I think the time is right for the demo scene to embrace the technology as well.
Speaking from experience, the web platform is great fun to learn and work with, and it has a lot of attractive features (crossplatformness and accessibility being two of my personal favorites). Sure, you struggle at times, but it's part of the challenge. As for compat issues, I find it quite easy to test/develop for different browsers (it's about as hard as alt-tab+f5), e.g. compared to testing on different gfx cards etc. Also, you pick up the differences between browsers quite quickly and learn to deal with them.
So, I guess I'm trying to say: Try it! Who knows, you may come to like it...
Speaking from experience, the web platform is great fun to learn and work with, and it has a lot of attractive features (crossplatformness and accessibility being two of my personal favorites). Sure, you struggle at times, but it's part of the challenge. As for compat issues, I find it quite easy to test/develop for different browsers (it's about as hard as alt-tab+f5), e.g. compared to testing on different gfx cards etc. Also, you pick up the differences between browsers quite quickly and learn to deal with them.
So, I guess I'm trying to say: Try it! Who knows, you may come to like it...
