GSOC 2014

[Cervator]

It's that time of year again - applications for organizations to enter Google Summer of Code are open!

Last year I went through the (at times somewhat heavy) paperwork to submit us and in the end didn't find the time to make an actual ideas page - which is sort of important. But I'm happy to have gone through it once and expect round two to be much easier to get through. Plus - we now have an ideas page!

https://github.com/MovingBlocks/Terasology/wiki/GSOC

Admittedly it is just a draft and could use a few more ideas sometime. Some of the ones listed also could use some more details (or the corresponding issue on GitHub or thread here in the forum)

https://github.com/MovingBlocks/Terasology/issues?labels=GSOC&page=1&state=open

It is near 1 am and I've been at this all evening, will add formatting and lots of links tomorrow then get the app submitted. Ideas page isn't actually reviewed until sometime after submission so we can keep adding to it. Not to mention prep for next year

Feedback, edits, more details, more ideas, volunteering to mentor are all heartily welcome!

Edit: Oh, I was tagging possible people left and right on the ideas, and describing some things from memory. Nothing on there is set in stone and I might have made mistakes. Just because I listed your name doesn't obligate you to be a mentor or anything!

 

[Cervator]

Updated page formatting a bit and started the application. Hooray paperwork!

 

[Cervator]

Paperwork is done! We're in the running

The GSOC bot has some stats as well:

In 2013, 177 of 417 mentoring orgs were accepted; 4144 students submitted 5999 proposals, of which 1192 were accepted

Considering that some organizations (like us last year, heh) don't even have proper ideas pages those odds aren't terrible!

Will look to go over the ideas page some more, maybe split out suggested skills into another bullet point. And moar links

 

[Cervator]

Final tweaks were made and we have a very presentable ideas page, reviewed twice in the #gsoc channel by the main organizer

There were 371 org apps submitted, while only 217 had been there about 24 hours earlier, so some newcomers might have been rushed. Acceptance rate is roughly 50% - not bad odds!

Now we wait, accepted organizations are announced Feb 24th

Thanks msteiger and Skaldarnar for helping out with the ideas page!

 

[msteiger]

Sadly, we didn't make it this year.

Accepted mentoring organizations are listed here:
http://www.google-melange.com/gsoc/org/list/public/google/gsoc2014

There's also an IRC meeting with rejected organizations:
28 February 16:00 UTC (#gsoc on Freenode)

 

[Cervator]

Yep, I'm waiting for the meeting to find out the exact reason. It could be as simple as not enough space - there are a lot of noteworthy orgs in the list.

 

[PrivateAlpha]

https://docs.google.com/spreadsheet/ccc?key=0ArgC47eQVqr6dG1DY0FpdUJqNzhiNUtCb1F2TEt0bHc&usp=sharing

A Listing of all accepted orgs from 2009-2014. Duplicates not removed so that we can see how many times they have been accepted.

Anyone want to look through it to find possible orgs to vouch for us?

 

[Skaldarnar]

He, thanks PrivateAlpha

Maybe we can reduce the list to all orgs tagged with one or more of these: "game", "java", "game engine" (and similar). I don't think that the OS specific orgs would have any reason to vouch for us, for example ...

 

[Mike Kienenberger]

Did we get a response as to why we were not accepted?

 

[msteiger]

Yeah, they said it was mainly due to the lack of an accredited organisation (such as the ASF ) that would vouch for TS. The ideas page was fine ("please keep it just the way it is") and also the fact that TS is in "pre-alpha" didn't matter to them.

 

[PrivateAlpha]

He, thanks PrivateAlpha

Maybe we can reduce the list to all orgs tagged with one or more of these: "game", "java", "game engine" (and similar). I don't think that the OS specific orgs would have any reason to vouch for us, for example ...

Actually I avoided doing that because there are oddballs like "Genome Informatics" which has nothing to do with us.. and yet we started the genome module yesterday and it could result in an interesting linkup and sponsor

TL;DR: filtering the list could result in missed opportunities.

 

[Mike Kienenberger]

Yeah, they said it was mainly due to the lack of an accredited organisation (such as the ASF ) that would vouch for TS. The ideas page was fine ("please keep it just the way it is") and also the fact that TS is in "pre-alpha" didn't matter to them.

Too bad I didn't know ahead of time. I will ask the GSOC folks at the ASF if the ASF could vouch for Terasology in the future. Or is this something that should wait until next year? When does it make sense for this to happen?

UPDATE: I have started a thread on this topic at the ASF. Sometimes things take awhile to filter through 400+ different opinions of how things should be done

 

[PrivateAlpha]

Ideas for partnerships and/or sponsors:
(#x stands for how many times they have been accepted, this year is the 6th)

• 3x 52°North Initiative for Geospatial Open Source Software GmbH - No idea what they actually do, but maybe we can create a scaled down version of earth with their help?​
• 6x Apache Software Foundation - No idea, but I noticed someone said they would approach them, so its here.​
• 4x ASCEND - software for solving mathematical models. No idea if a partnership is possible. Pinging @Immortiusbecause he is smart ​
• 1x ​
  Audacity - Ingame sound maaybe?​
• 6x ​
  Battle For Wesnoth - gaming link​
• 3x beagleboard.org - maybe see if we can fit Tera on the board for truly mobile gaming?​
• 6x Blender Foundation - no description needed​
• 3x ​
  Genome Informatics - we started genome stuffs recently right?​

  ​

Ok, just listing a few that I noticed, there are probably many more though!

 

[Skaldarnar]

Quick search for some more keywords revealed the following:

• 6x Crystal Space - game dev, engine, simulation, landscapes
• 1x PaGMO/PyGMO - genetics, evolution, swarm intelligence, global optimization, networks, AI
• 1x OpenCog - AI

 

[Mike Kienenberger]

I think the hardest part isn't going to be finding like-minded projects. The hardest part is finding someone who will say that you would make good mentors. Based on my experience here, I think you would make great mentors.

I haven't gotten a relevant response from my fellow ASF members yet, but if nothing comes of it right now (and even if something does come of it), be sure that we remember to do this next year. Even if I am not active on the project during application time next year, please send me an email (mkienenb at apache org or mkienenb at gmail com) and I will pursue adding Terasology as a vouched-for organization on the ASF GSOC application.

 

[Mike Kienenberger]

I'm curious. How many organizations were approved which had never been approved before? And of those, how many had applied for the first time?

 

[Cervator]

Thanks for the constructive posting everybody - I got stuck away from this for a little bit

Here's the feedback I recorded during the meeting, where we made an impressive showing IMHO with no less than 6 of us in channel at once, leading to some amusement trying to get us all voiced. We definitely stuck out against most the solo people who made it there, only a few others had 3-4 reps present. I very much appreciate people's time and engagement on this!

• Finding somebody to vouch for you may help (as noted above this is likely the biggest deal)
• Beware of colorful language / humor that doesn't translate well to international university students (like project name "DAMN" with that word in every project component)
• Several orgs (especially veterans) were left out just to make space
• Some fields were highly competitive, such as OSes, programming languages, and bioinformatics
• Doesn't help to prime some known students with extra details and then neglecting your idea page since they already know what to do - doesn't "count" for your GSOC readiness
• One project's idea page had a section for brief mentor bios, looked kinda cool - http://www.dune-project.org/gsoc/2014
• Take feedback well, rather than defend and blame something other than what the feedback points out
• Probably add "Expected Results" and maybe add more structure to the individual ideas - https://github.com/scrapy/scrapy/wiki/GSoC-2014-Ideas / https://github.com/namecoin/wiki/wiki/GSoC-2014-Project-Ideas
• Add pictures to the idea page! http://helikarlab.org/GSoC.html and maybe a slight bit more formatting / color
• Umbrella org might be a nice fallback to consider - seems GSOC is willing to send extra slots to an umbrella corp specifically for any rejected orgs they take under their wing. Probably just need to fit together well. Several other orgs seem to have done this multiple years
• Our feedback: Ideas page already about perfect, just numbers + lack of vouching

In short: Most rejected orgs either had poor idea pages or were awash in a sea of applications in their field and lacked vouching to stand out against the rest. We fell into scenario 2 with our idea page already reviewed and in great shape.

I'm happy to see so many ideas for vouching already. The main thought that came to mind for me was seeing about maybe joining the Free Game Alliance (a few free and open source games including Battle for Wesnoth) which Aperion suggested a long time ago (I suspect I have the site in my browser in the first 100-200 tabs somewhere ...). He has worked on Rigs of Rods, and I poked him a bit here and there about maybe getting some more vehicle type action into Terasology sometime

When I first saw the FGA page I wasn't sure how active it was, but their news page seems active still and even noted that Wesnoth got into GSOC. Their goal is one game per genre and they don't have a sandbox / voxel world option yet. A requirement for joining is an established fan base which we don't really have currently but probably will have by the time the next GSOC rolls around. Finally this would seem like the perfect scenario for an "Umbrella org" GSOC style in case one of the FGA members interested in GSOC doesn't get in

ASF sounds like a great option too since we're linked on a person basis thanks to Mike Kienenberger - I hadn't realized the potential link between that and GSOC before. Just a bit of a pity we don't link well functionally, although I have to keep reminding myself that we do use Apache libs, it isn't just a web server after all Another similar option would be LWJGL, but I can't actually find them in the veteran org list (wat!)

Finally: Maybe we can automate the GSOC ideas page. Grab every GSOC issue then parse out some template fields from the first post, leaving out any details not meant for the wiki. If the first post follows the template it gets a section, if not it gets a mention in a misc pile in the bottom. Could even support sourcing them from different GitHub repo trackers