Any news about Google Summer Of Code?

[simhar]

Hello,

I read on blendernation that the date for sending in the proposals for GSOC has ended. Will the proposals be presented?
When will the choice be announced?

Thanx
S.

[basse]

Google Summer of Code™ We are no longer accepting student applications for Google Summer of Code 2007. We'll announce accepted student proposals here on Wednesday, April 11, 2007. In the interim, visit us in #summer-discuss on Slashnet or in the program discussion group.

http://code.google.com/soc/

.b

[grafixsuz]

I have been checking in, seems that announcements will still be a little way off. Well it is Thursday 12th here, so I will check again later.

[PapaSmurf]

where's the guy that was here a month ago talking about volumentrics (clouds, fog, smoke, burning gas/fire)? I dont see his proposal...too bad, that would have been awesome.

[kakapo]

http://code.google.com/soc/blender/about.html

they are there now.

[BlackBoe]

Sweet, nice to see them up. :]

Yeah, too bad about the volumetrics, but deep shadow maps and tiling are both pretty slick. Same goes for full 3d window material previews in GLSL, the audio revamp, and the render API.

What I've come to realize is that Blender is now in the transition phase. A lot of the stuff will be more behind-the-scenes than all WOOHOO-RAZZLE-DAZZLE, cause Blender is 10-years old, and parts of it were never properly designed for the massive sets of features that have just been piled on top of the base over the years.

Right now, with the render API and the nodes editor and the new BMesh system and things like the window manager and event system refactor, everything is being properly established now, everything is being given nice, solid, flexible, powerful new bases, and everything is being divided into easier-to-manage chunks (sending instructions through API abstraction layers as opposed to blocks of code mixed in with the base).

All this--while not as exciting--is a very good thing, because when Blender's base is more or less finished, when core work fades to a minimum, then it'll be so much easier and faster to add the bonus 'high-end' artist tools that the industry loves so much(spoiled babies. XP). Right now, all the flash-bang features would just kludge up Blender's already kludged framework. This and the 2.5 release are sort of the 'structures for the future' period. It warms my heart to see these applications accepted. :]

[FuzzMaster]

None of those particularly interest me. Because I don't understand a few of them.

Oops.

[BlackBoe]

Deep shadow maps are good for hair and volumetrics and are pretty much in every way better and more detailed than regular styled shadow maps. Coloured transparency, motion blurred shadows, etc. They take up more space, though, so he's going to make a swapper for them as well, so they don't drain RAM so much.


Render API. Currently, the Blender Internal renderer is tied into Blender in one big, tangled, disorganised mess, all loose wires sticking out. It's hard and unpleasant for anyone to do any work with render systems. The API will take all the loose ends and tie them into a standardized 'plug'. This'll be easier to take care of, and also allow for faster, cheaper, easier and more efficient addons for pretty much any renderer ever, Indigo, Sunflow, Yafray, Renderman, etc.


Blender audio system upgrade. Right now Blender's sound system is like an old toolbox, filled with odds and ends and old bits of junk which might not be useful but which no-one wants or has the the time to throw away. It's a basic cleanup and upgrade job, though it'll take a while, considering the mess.


GLSL Shader Editing and Preview: Basically, a big revamp of how the Blender 3d windows draws materials, this time using GLSL pixel shaders. Will probably be a crapload better than 'textured mode'. :P


Sweet, sweet, quality bevel. Nuff said.

[moh taia]

that's great news
is this projects accepted already or they listed to accept some of them ?

(btw) if i understand right there will be another great project in parallel with SOC this year

see this link

i think it's mean there will be more and more projects to be done this year

[BlackBoe]

These are the five accepted ones.

...Also, re: the Toronto projects: holycrap. I was completely unaware of that. That's awesome.

[Koba]

Before I get accused of wanting shiny features over core code improvements any further, please have a look here at this thread from several months ago where I ask for a roadmap on core improvements and refactoring (which we just got!)):

http://blenderartists.org/forum/showthread.php?t=91396

Frankly, while I appreciate refactoring needs to be done, I am disappointed that Freestyle integreation was dropped. EDIT: Looks like Freestyle integration wasn't submited. In which case, this is about as good a GSoc as could be expected (though volumetrics would have been nice)

My initial thoughts:

- Deep Shadow maps. If this really is much better than raytracing shadows (ie faster) then this *could* be useful to lots of people. Personally, I probably won't need it for the sort of projects I am doing. If hair rendering is 100x faster with shadows then it is probably worth having. Motion blur shadows is an interesting idea.

- Render API. Again, I thought this was coming anyway with ton working on it. Disappointed that it took up a slot on Summer of code - this seems to have been in the works forever. Good if it gets the job done and LOTS of external renderers quickly get full and proper support in Blender.

- Audio cleanup. Of little use to me because I've never used Blender for audio at all. I suppose this is more for the game engine people than the animators. If the code is *really, really* messy I can imagine why developers would like this (if it hampers work on the visual tools of Blender). Less useful tof the hobbyist maybe, but important for animations like ED where lip sync is needed.

- GLSL. This had better be a *huge* improvement over the current OpenGL system to be useful for normal working (aka use of pixelshaders in Truespace 7.0) - only if the preview *really* replaces a quick renderer will this speed up the workflow. The prospect of having a good preview of node materials is a very good one though.

- Mesh Bevel. I can't see how a modifier that is applied at the object level will bevel a mesh the way I want. While bevel really needs improving, I expected it to be a mesh tool not a modifier. Do I really want to bevel every face on an object? EDIT - It seems to be a modifier *and* a mesh tool so this is a much needed improvement.

In short, this years GSoc is unlikely way draw attention the way sculpt mode did last time round. Still, better a GSoc with one major internal improvement than no GSoc at all.

Koba

P.S> I hope the students doing the Toronto design project will chose to integrate freestyle! There are loads of cool projects listed their though. Hopefully, this will be a popular option every year for those students.

[broken]

Originally Posted by Koba Frankly, while I appreciate refactoring needs to be done, I am disappointed that Freestyle integreation was dropped.

Freestyle integration wasn't dropped, a GSOC proposal for it was never submitted in the first place. All the hubbub earlier was just that the Freestyle developers were saying "Hey, it'd be nice if someone did this for the GSOC, just letting you guys know we think it's a neat idea". Nobody actually committed to doing anything.

Personally, I'm very excited about the Deep shadow maps. Blender's current shadow buffers are quite dated and limited and deep shadows will give much better quality on detailed things like hair, and also should provide the ability to have transparent shadows (currently only available with Blender's slow raytracing).

In short, this years GSoc will in no way draw attention the way sculpt mode did last time round.

Not to take away from the previous sexy projects, but I'd prefer the projects be aimed at improving Blender in needed and tangible ways than just drawing attention. More utilitarian projects in the past like IK and modifiers have been just as useful as the big fancy ones.

[RONIN]

I had my fingers crossed for Freestyle,but it looks like the the code gods were not with us all the way on this one.Here is hoping that the Render API gives it another chance,and with Ton overseeing this, I'm sure it won't fall off like some of the summer coding projects of the past.

Well...back to using the (10 year)old toon egde.

Congats to all the accepted projects.

[yellow]

I thought the Freestyle guys had found a couple of coders? Is it more the reason that they got no support in the blender developer community only a bit of noise here a BA.

Looking at the reception the Freestyle guys got on bf-committers like 'go read to non existent render api then implement Freestyle after theres an api, but when that is no one knows' or 'this ones for the fun board go play there' or 'so what's freestyle some sort of image manipulation plugin' no one was even interested in knowing what Freestyle was before dismissing it.

I'm not suprised they got disheartened and moved on it all seemed to go quiet with them very quickly after the initial introduction.

Ah well before someone else says it, the power of open source, you can always code it yourself, it's a lottery what gets taken up.

[BlackBoe]

Originally Posted by Koba Frankly, while I appreciate refactoring needs to be done, I am disappointed that Freestyle integreation was dropped. .snip. In short, this years GSoc will in no way draw attention the way scuplt mode did last time round. Still, better a GSoc with one major internal improvement than no GSoc at all.

EDITED: Most of my reply has been rendered obsolete by explanation, clarification and editing of other posts. Suffice it to say I have a short temper and all I really saw in the "these are all disappointing, except maybe one is ok" attitude was one of the countless people who sit around, revelling in the shining glory by association of using new features, with no real effort to show for it and no real regard for the work that needs to go into things. I may have been wrong. As it is: Edited.

[grafixsuz]

Well it is good to see these projects given a chance, I think that incorporating too many 'Sexy projects' as broken puts it will just be in the way for the core refraction. Congratulations to all participants!

[Mystery]

Even though it might not seem as flashy as some past projects, these ones are no less important. I think people aren't looking far enough ahead, for example from what I understand things like the addition of the render API will allow easier integration with renderers like Freestyle. Which means after the GSoC you might just get your wish, except it will be a lot easier to do.

Also I think these projects go nicely with the UI work that is planned.

The future of blender looks very flexible

Mystery

[Koba]

the only work you have done on this forum was one thing in 2004, your stance a liiiiiittle pretentious.

BlackBoe, have you ever considered that I may use Blender *extensively* (ie every day) but don't post my work on elysiun? Did that even cross your mind? I may even be saving up for something big. Just wait four months or so and you will see how big.

I don't think you appreciate that refactoring really needs to be done in the slightest or else you'd understand why these are so great.

Did you even read my post? Did I at any point say that the render refactor didn't need doing? Actually, I assumed it was so important it was being done without a Summer of Code project. Wasn't that clear or should I spell it out again?

As for the audio refactor, how about a poll to see how often people (other than those using the game engine) use audio extensively in Blender. I think you'll find *most* people rarely use anything audio. *Most* people use Blender for 3D visual work and only need audio if they are synching to music/speech. You'll find that 90% of animations on this forum have no synched audio at all. Probably more useful for project like ED or Plumiferos. So if a refactor there is needed, it must be to clean up code from a developer's point of view and not the "average user".

Now, I think the fact that you think that just because all this doesn't impress you, the industry as a whole won't notice is just silly, and once again, a little pretentious.

That comment was completely unjustified - I never said or implied anything of the sort. The big changes (apart from the render refactor with ton) are in the first half of your list and they have nothing to do with GSoc.

Oh, also. I like how you say "then (shadow maps) *could* be useful." as if it's all about people trying to get your approval.

When I wrote "my thoughts", I meant *my* thoughts. Not *your* thoughts or anyone elses but my own. That means pertaining to *me*. Did I say the GSoc project should be abandonned because of what I think? I was giving my *personal* opinion - wasn't that clear? I certainly hope these features will be useful to many even if not to me.

I have nothing against you BlackBoe and I hate having to reply to you like this. There was no need to be so aggressive - you could have stated your points in a more civil manner. Such pointless animosity.

Why don't you could the number of posts on elysiun/blenderartists/blender.org where I tell people how great I find this community. Unfortunately, the atmosphere here has turned hostile. Frankly, that is such a shame. If posts can't try to be a little more understanding and civil, I think you will find that many longtime and dedicated members of elysiun/blenderartists will leave.

Anyway, thank you broken for that information and I'm glad you liked the car BlackBoe. That project was abandonned but I have a huge volume of artwork here I plan to show you all one day if all goes well.

Koba

P.S> I have half a mind to flood this thread with my current work after reading that nasty post. Luckily, I'm not stupid enough to ruin a whole *year's* work because of some groundless, unprovoked attack.

[Vylaroth]

Well, I got to say that I didn't care much for the Freestyle integration so this doesn't disappoint me at all. As people have already stated, Blender needs code clean-up before it can have more flashy features implemented into it. And well, to me integrating a sketch-style renderer is more like adding a new filter to an image manipulation program: it just makes your model look neat but when all people start using it, the flashiness of it wears off. I'm not saying that it would be a waste of time, but when compared to the current GSoC 07 projects, it just looks really unimportant.

By the way, is Briggs working on BMesh alone? I just can't wait to get my hands on n-gons and the accompanied tools. I wish this had been addressed last year already, but well, better late than never.

Will GLSL preview implementation make normal map previews possible in the viewports? That's one feature I've been waiting for too.

Blender is already a great indie game modeling tool, I hope this side of it gets some attention too in some GSoC. Stuff like smoothing groups or smoothing angles would be just awesome. Also, I don't think that the Blender game engine is currently very feasible for developing anything major so putting some attention to exporting stuff out of Blender (especially animations) would be really nice. Some proper integration with Ogre and other open source graphic and game engines would be really handy.

But well, it's just me daydreaming about stuff that I want. I bet people who use Blender for e.g. rendering stills for graphic design couldn't be less interested in stuff like that. I just think the gaming side hasn't gotten that much attention lately.

[simhar]

oh, I'm very excitd about this all, although the volumetrics didn't get it...

@broken: why do you say the buffered shadows ar outdated, because there was this huge step forward with irregular and halfway-shadows...? are these based on old code or do you mean just the old classic stuff?

All in all it's great, but I'm also a little bit afraid, that it don't comes like last year with the skygen project... that was very disappointing. Imagine: everybody says: whoa, deep-shadows and then nothing follows *cry*

@blackboe: it's for me the same like for koba- i have a lot of work done with blender, but don't show anything. some of it, because i can't (because of clients).

@koba: don't think you HAVE TO show anything...

Greetz
S.