PB2 Version Discussion

[XtremeBain]

I just wrote a huge post that outlined our current versioning format and why I don't like it.  There were lots of wikipedia links and even an urban dictionary one to keep the readers on their toes.  I even cropped a screenshot and circled things in red with mspaint.  I was doing a final preview and accidentally clicked one of my links in the toolbar, instead of my second tab, and everything was lost.  Just be thankful you don't have to spend 30 mins reading this post.

Here is the synopsis leaving most of my extensive reasoning out.

As far as OSS is concerned, Paintball2 Alpha Build 19 as a software release comes off as a very immature and early release of a project:


I'm proposing a version system like Ubuntu's (mostly because jitspoe hates Linux) which is almost yy.m.  That would mean that Build 19 would have been 7.06, if Build 20 comes out in December it would be 7.12, Build 21 next March would be 8.03.

I had a lot more to say explaining why we should move away from Builds and stop chasing a 2.0 version, but I don't have enough time right now to rewrite it all.

Feel free to discuss and I'll rewrite some of my points as they come up.

[KnacK]

I like that versioning schema as well, sicne it scales up and down gracefully. ie.. 7.06.13 would be June 13, 2007.

+1 for Bain's idea.

[Zorchenhimer]

The "alpha" status never really made sense to me.

+1 for Bain's idea.

Me too.

[XtremeBain]

Me too.

My original go at this hinted at the idea that jitspoe may include menu widget variables in 8.03.

[Zorchenhimer]

My original go at this hinted at the idea that jitspoe may include menu widget variables in 8.03.

What are we waiting for?!

[jitspoe]

While I don't deny that the versioning has issues, I don't know that a date-based versioning system is the best way to go.  I don't like the idea of going from 7.x to 8.x when there may have only been a minor change.  Plus, you have no idea how many versions happened in between.  Changing the "major revision number" makes it sound like there was a major revision.  Like the game was somehow rewritten from the ground up or ported to a new game engine.  It's deceptive.

As for why the game is still called "alpha"... it's because it is.  In my book, alpha = not done.  Beta = done, but needs testing.  Final = after bugs in the beta are fixed.

I just don't feel the game in its current state is "done".  It's still a heavy work in progress.

- Most of the sounds need to be replaced, as they're low quality and contain copyrighted material.
- New model format is implemented, but needs LOD support and many meshes need to be re-done.
- Lots of low-res and even 8bit textures need to be re-done.
- Map format needs a complete revamp, allowing for more detail, better outdoor support, embedded meshes, LOD, etc.
- Sound system needs a complete revamp to better take advantage of the 3D audio most cards support these days.

Also, you have the game code which has its own version system going on (hit 1.9x when the global login system was implemented).

My plan was to get some of the key things done, like the new map format and a media revamp, then switch everything over to a uniform 2.0 beta, after which I'd start a 2.xx versioning scheme.

[XtremeBain]

Fair enough, my intended post touched on the gamecode versioning and potentially adopting its.   I still don't see this project as Alpha/Beta software.  With every new release we treat it as the Latest Stable.   I just think that this will remain a game which is constantly in development.  At the(your) current capacity, we're still 2yrs off from the five main things that you want done before syncing up to a 2.00 Beta, but then we'll still have 40 heavily voted ideas that everyone would love to see implemented.  Are you planning on freezing the 2.x feature set and starting a 3.x branch?  Or do you want everyone playing on the latest and greatest release?  I see this project as something that is and will be constantly evolving and it's also why I think a date based version (or even the existing 'Build X' version) suits it a little better.  Your latest releases have been really solid.  I'm not a huge fan of your pretest model for beta testing, but it goes to show that it's working.  I agree that this project is missing a lot of features (especially your gamecode), but given the approach that you've taken to developing the stand-alone over the past few years, I'm not sure why we're waiting for the far distant 2.00 before we have a "released product".

PS: In the future, don't feature vote any threads I start with "discussion" in them.

[y00tz]

Is there a documented roadmap for 2.0, or just the blurb Jitspoe posted up there?


Edit:

I kind of like this version system Xbain is proposing, and I'll definitely implement it at work, however it doesn't seem appropiate for this game, it seems more inline with a program that doesn't alter greatly between builds, and something that is built nightly. 

I realize you guys are talking about the game code, so this would not affect the build numbers that people are used to?

[jitspoe]

The "pretests" are basically just to make sure there isn't anything blatantly wrong with the release, like debugging information I forgot to take out, missing files, etc.

[Fullmetal_Steeb]

Jitspoe, are you ever planning on making any money with maybe a much later version of the game if you take out copyrighted sounds, etc...? Would that ever even be possible? I would definitely be willing to pay for a later version of the game, provided there were some major changes of course.

[jitspoe]

A bit off topic.  You may want to read this: http://dplogin.com/forums/index.php?topic=2093.0 (it's an old thread, so don't bump it).