[RandomTiger]

I think there is already a general rule on this kind of stuff but I would like to discuss and come to some kind of solid agreement.
Note my own believes on this are somewhat scattered so its not like Im trying to lay down the law.

Most importantly I believe that anyone using our fs2_open exe should be able to play FS2 the way originially intended. I think all new features should be optional and by default off.

You could say, use the original exe but I have a GF4 and I would like to play with the fix and bumpmapping when it goes in, but everything else the same as before.

Fortunately there will be a lot of things that can be used in fred but not effect old missions so thats cool.

So then I think about the new graphics engine, I guess that would have to be compulsory feature since we cant really provide support for both engines. So I think, well as long as everyone prefers the way it looks that OK. Then, hang on! I really hate the beam titling but apparently thats really popular! Crap. Im in the minority and it sucks, I dont want to ever have to play with beam tiling.

Note this is an example of the problem that I am using to illistrate my point, please DO NOT fill my thread with crap about beam weapons.

OK, I could use my infleuence to keep it optional in a new engine or plain just make my own version without it but what about other people (who cant code) and other features.

Then I think about AI, if someone does a complete overhaul and makes it great, I even like it, great! But the changes (would almost certainly) effect old missions and mean they dont work in the same way, the well planned balance between V's code and level is broken.

But we cant keep every code diversion or every option. It would be a coding nightmare and hinder progress.

So what is to be done?

Abandon the FS2 and FS1 missions in favour of the mods which (please dont flame me) most likely, for me anyway, wont be as good as the freespace campaigns.

Or

Do we have different code branches, release increasing numbers of exes with optional parameters for that go on for miles.

When I started this rant I knew what I wanted, now Im not so sure.

AAAaarrrghhghghg!

Im sure theres a solution, but it escapes me. I believe this issue requires some thought. We sort it out now and identify these problems, we save time later.

[Inquisitor]

Case by case.

THings like the beams are meant to be optional.  They aren't in THAT build because we wanted to get a build on. Stop obsessing over them.

Search back in the forum though, this is the philosophy we should apply. Don;t break the existing game, provide additional functionality thru sexp's, pof attributes, command line switches, etc, and occasionally as a build option that builds a special exe.

Extend the functionality of the game, preserve the original masterpiece, only make those changes compulsory that improve (defined by consensus here) the overall game, and document the changes sufficiently to let anyone with access to the code to make something custom, removing functionality if they feel compelled to.

Just because we have ONE feature that CURRENTLY is not optional, does not mean we are going back on that philosophy

[WMCoolmon]

IMO, backwards compatibility isn't a huge deal, but it is something that should be taken into account for things that will affect it. For example, if we consider adding a little light on the HUD that blinks when you're out of ammo, but it will break backwards compatibility, it's probably not worth it.
On the other hand, if there's no way to add bump mapping while keeping backwards compatibility, it's probably better to drop compatibility in that case, since bump mapping would be a MAJOR improvement.
I think at most there should be two EXEs. One that keeps compatibility with Freespace and one where that isn't the main goal. The former would contain only bugfixes and perhaps some vital features, while the latter would have both bug fixes and new features.
And though the Freespace 2 campaign may be the best campaign so far, it's getting old and it's a pain to deal with because the art and such cannot be redistributed. A mod/campaign does not have that problem.

[penguin]

Originally posted by WMCoolmon
IMO, backwards compatibility isn't a huge deal, but it is something that should be taken into account for things that will affect it.

...

And though the Freespace 2 campaign may be the best campaign so far, it's getting old and it's a pain to deal with because the art and such cannot be redistributed. A mod/campaign does not have that problem.

Many excellent points here.  I've always held that backwards compatibility was paramount, but your post is making me reconsider.

The "two EXEs" idea might be ugly, but it's workable -- prudent use of #ifdefs around the surrounding "incompatible" code would be fairly easy.  Similar to what's in the current FS2 codebase, with the FS2_DEMO define.

Maybe this should be a separate thread, but here's what we would need for a fully-distributable game (I'm probably missing some stuff here, this is off the top of my head):

• campaigns / missions
• ships and weapons (no :v: models or TBLs)
• UI -- backgrounds and buttons
• HUD(s)
• sounds and voices
• music

And all the other TBLs: strings.tbl (although that is generated from the source), the ones that deal with personas (forget which file), Species, etc...

[Martinus]

Considering that music and voices are for the most part expendable assets why not suggest a competition? Simply ask people to submit new material and the assets that get the most votes are integrated into the game as default.

There are loads of excellent models, missions etc. already available, you have an excellent range to choose from.

[Nuke]

this is an issue that needs to be looked at from different angles. im not sure how many of you who have already are modders, but i am. this should bring a different perspective to the argument at hand (this is a debate thread, right?).

i recall how quake 3 was being moded from the begining. everyone who made mods had to update their code to be compatable with the latest point release. despite this henderance, many mods were released. this means that it is possible to manage 2 (most likely more) versions.

the engine improvements is the first, which is the base for the others. this version should be fully compatable with old freespace. there will most likely be many a version of freespace that is specificly designed for a particular mod or campaign. the b5 mod version for example. worst cas scenario is that we will have so many versions of freespace floating around that do different things.

after observing the way code has been developed so far, it seems that each developer has his own interests. some perfer to work on graphics and performance changes to the game, others are more focused on mod development. there are many versions of freespace out there (this is the same thing that happened with the quake source), and each does something different, one lets you use fighter beams and one has beem tiling. i just hope that this is due to the early state of development with the code.

all the ideas i have thought up can be implimented without breaking the original freespace.

[Bobboau]

and if you don't like a feature or the implementation of it, go to the thread dedicated to it and ether express you're disaproval or make sugections on how to make it better,
the beam tileing for example was comited shortly after I made the most primitive build of it, as of now it is fully customisable, and fully cool, but there are still bugs