Here are my comments about this:
Deadline: 2 months is not short, but it might be too short to yield acceptable results in at least the "original game" category (which I am referring to in the next paragraphs).
Sure 24-hour game jams sometimes have interesting and incredible results, but that is not the norm (LD48 has a few thousand submissions and at most 50 of those would be considered "good".. comparing to our 50 total developers now...). The main problem with that is misinterpreting how long 2 months actually are and how much work has to be done (it's always less time and more work than expected).
Looking at all past competitions we only had a few really good entries and most of those weren't even done, aren't done at this point and might possibly be never done (including some of my own projects). That's not a good result taking into account how much money was given out (LD48 gives out nothing, except the feeling of accomplishment).
So I would either like to see a "no unfinished builds or tech-demos" rule (which is unpreferable as just trying to get rid of the symptoms rather than the problem) or an extension of the deadline - maybe even just for the "original games" category.
Voting/Community: Community vote... well meh. I see how hard it is to organize a proper judged based voting, but to me that is highly preferable.
The main problem with a community vote is that the community does not vote. Looking at the rebirth competition: 153 people out of over 5500 registered users and over 4000 owners of a Pandora (which is not a subset of registered users) voted - in any case that is less than 5%!
Some people mentioned that there wasn't enough time to vote, though I have to disagree - 2 weeks should be enough to try entries and make up your mind (you don't need to play a game to 100% completion to form an opinion).
The main problem, I think, is that most people simply aren't aware. There basically was no advertisement for the compo, besides some official news posts. I recommend putting up a big banner (linking to a hub site or the announcement thread) at the top of the forum for the time the compo runs, with an additional massage once the voting starts. Put up banners on the repo, file-archive and dragonbox shop, too. Have a sub-section for project leads to post about their projects, have a hub site for them to post pictures and videos (so people don't have to search through 10 pages of discussion to get to the interesting bit). Have someone make videos about the projects (so the leads do not have to, since they should be busy working on their projects anyway and can't spare too much time). Just try to get the work out there with as many means as possible and as consistent as possible, so it's constantly on people's mind and not forgotten until the voting is "suddenly over".
Touching on the first point again (short deadline): Have some experienced devs make Keynote videos, where they talk about managing your time, what you should think about when leading a project, what your goal should be, etc. (see LD48 keynote videos or this one for examples:
vim eo.com/53137719 ).
The compo should not be "fire and forget", there always has to be some effort involved from the organizers to keep it running and interesting to outsiders while they wait for projects to finish. This also generates more publicity and more encouraging comments on the dev threads, so more devs are likely to put time into this, which in turn also makes more devs put up dev threads and have an open development. I am pretty sure people in the community would be willing to do this. I an willing to help organize and lead a group of "journalists" (maybe we could reactivate the pandorapress site for this - I tried covering a few compos, but I failed to stay on the ball every time. There is only so much one guy can do...).
Categories: Good, I like how they are split
Logo/Splash: Good, too. I hope we still get an awesome logo and banner to put in sigantures and link everywhere.
Sourcecode: I personally don't think people should be forced to open-source stuff, but I don't really mind. In any case clearly state that the code needs to be "open-source" or specify a license as with the current rules one could just legally include the source code, but disallow any kind of use or modification defying the intended goal of this.
Some general thoughts: I drifted over to the idea that compos aren't the best idea to spark devs and the community, it hasn't really worked before, it might not work this time. Though I think a few things have been done wrong and with a little of constant coverage this might be more of a success...
In any case, I would rather like to see certain promising projects be rewarded once they are completed (or maybe funded for completion), instead of a shot in the dark (which a competition basically is).
Host a "project slam" in which a few people have to make pitch videos for their project, maybe a first prototype and then let the community decide which one of those gets the money. Then pay devs upon hitting certain milestones until the project is finished. It's similar to a "standard" competition, except there is no deadline and only one project is actually finished instead of a few going half-way.
Obviously you can't actually pay the devs to work on their project full time, but a little bit of money should provide some extra motivation. Also the money is partially payed up front and the devs don't develop for a chance at winning, rather they create a quick prototype, so not much time is wasted in case nobody actually likes it (a great idea in your head can always turn out to be not-so-fun in reality).
For examples of pitch videos see this:
https://www.humblebu...com/double-fine
Thanks for taking the time to read this.