Well, this may be a good idea, but ...

I don''t think that most Internet projects fail just because people haven''t CVS or something. They fail because people start working without having appropriate experience.

I, personally, have been in four Internet game projects. The first one was insanely successful... (Well, it''s still going strong from what I know.) But I had to drop out because I couldn''t keep up - yes, it was my first project. The second and third... didn''t work worth a damn. Why? Because the other members were like me: didn''t know the first thing about game development, but thought they could make a game any way.

After those two, I was all depressed and figured I''d make simple games alone and then try again. After making tetris, pacman, minesweeper, another tetris, a few other simple "board" games kinda like minesweeper, and countless other old arcade clones, I figured it was time to try again.

I joined a one-man project. But he knows exactly what he is doing. We''ve been working together for almost a year, and have achieved more than I have ever done in all the other projects combined. And that even though we didn''t have a web-site, or CVS, or forums, or anything. Our only medium for communication was ICQ. Well, as we progressed, we got money for a site and CVS and all those nice things. But we are still working as a pair. Noone else. Why? Because the guy has the vision and I have the skills. And we can work together. That''s it.

So... My point is not to make anyone jealous or anything stupid like that, but is this: no Sourceforge for games is going to help you unless you know exactly what you are doing, and you have the skills and experience. That''s what''s lacking in most "teams" now. Not that they don''t have access to CVS. Such things are solvable when you''ve got everything else. It''s like when you are building a house. Skills, and experience are your foundation; vision and design is your walls; and forums, cvs, and stuff is your roof. You''ve got to start building with the foundation. But most people start with the roof, and so it doesn''t work...

It does sound like a good idea, but I think some people (including me) would be very hesitant about putting their closed source up on a public high-traffic website like that. If someone managed to hack into the system or steal your password, they could download all your source. Plus there is the fact that the moderators and owners of the site can probably automatically see all of your source. SourceForge doesn''t have this issue because all of it is open source so nobody is really concerned about who gets to see the source. A lot of the more unprofessional game developers would have no problem ripping off someone else''s source and putting it into their own game if they could get away with it, and a site like this will almost definitely attract that sort of crowd, unfortunately.

But I think something like this might still be useful even if it focuses on open source. SourceForge is great, but its focus is more generic for all types of applications. I think a site dedicated to games would be very useful, especially if you could provide digital asset management like some of the game tools such as alienbrain, for example versioning and tracking of models, textures, music, etc., instead of just a generic binary backup like SourceForge. Things like screenshots of the day and lists of available open source tools and libraries on the site to be used in game development would be very useful.

quote:Original post by carb
That''s very true. I mean, if you don''t posess the skills to begin with, CVS isn''t going to materialize them. Of course, I think CVS is just one aspect of the idea, but certainly a central one.

The idea came to me when we were discussing potential projects in a 4th year independent research course at my university. As it happens, a group of students will be porting G-Forge to the school''s server for certain CS courses, those that involve group programming assignments, for example. It''s an interesting application on a number of levels, because a) TA''s and professors can see who checked in what (who did what work) to CVS prevent cheating and assign marks to those who deserve them and b) groups can more easily communicate among one another and merge their work easily.

Now, being in 4th year, I can see where such a service would have been helpful in an academic setting. I remember doing assignments (and personal game programming projects) which relied on just a slew of emails going back and forth hour by the hour. That was not very effective ... Everyone had different versions of the code at any given time.

Hey, I''m the AP from above here...

I see what you mean better now, and certainly it might be of some use. Now, I am a second-year CS major, and at our university, we all have access to a CVS system (granted, not for personal use), and writing a bug-tracking system is actually a semester-long project in one of the classes, so basically, everyone has one of those, too.

So, I certainly see how a Sourceforge-like piece of software can be very useful in academics. But, isn''t Savannah pretty much all you need there? It has the CVS system, bugtracking, coding assignments, and such, and a rudimentary forum system. So, you can just put Savannah up on your server and work from it. Of course, that only applies to a university setting.

When it comes to organizing teams over the Internet, a whole other problem comes in: being able to motivate all of the coders and keep them working. It is a bit easier when you can contact the member of your team in person, but when they can just disappear in the Internet, there is no sense of accountability for the work you need to do for a project.

But any way, I digress. You say that you would like to keep such a server clean from people that we shall for brevity call the "Help Forum posters", meaning those that maybe have ideas but not the necessary experience. You can certainly monitor requests for hosting, but I think there is no way you can keep your server clean from such "teams". The trouble is, you''ll want to be able to host teams early on in the development process, otherwise the hosting might loose certain appeal, and people might become suspicious: why do you want to see actual progress before accepting our project? But then if you only trust the words, anyone can say: oh, I''ve got the team, and everybody''s got the skillz, and we got such cool ideas... blah, blah, blah. There is no way to tell the real teams that can achieve success from the hundreds of HFP''s.

So, you''d have to constantly weed out the current members based on the progress done. Like, remove accounts of those that have done nothing a month after they registered, say. But then you run the risk again of hexing legitimate teams. For instance, the very first Internet project I worked on, the one that I couldn''t keep up with: when I just joined it, there was a tremendous slowdown. Their team leader''s computer had fried, so he could only access the net from the library in his town for like a month. And then, at the same time, their lead coder got hit by a car and broke pretty much every bone in his body, so he spent over a month in the hospital and also had a concussion or some such, so he couldn''t code for a while after he got out due to terrible headaches. So, for at least a month, the team was a bit disorganized and nothing much was going on but bug fixing. Well, if you have tens or hundreds of projects to weed through every day, you might just accidentally take and delete such project as inactive, while might certainly not be right.

So, I guess my point is: while it might be a good idea to set up a server like Sourceforge for "serious" game developers, I am not sure how one would actually go about maintaining the "seriousness" and weed out the "other" teams...