Quite some time ago it became clear that Las Venturas Playground 3.0 on MTA was not going to work out. The development team continued working on version 2.90 for SA:MP. But it's time for something new.
Recently I've been working with Fireburn on a concept for LVP 3.0 on SA:MP. In the attachment you will find an introduction to what we think 3.0 should be like. More extensive plans and information will follow soon.
Please reply with any suggestions or comments you might have, we'd like to know what you think so we can incorporate this into the plans. We want version 3.0 to be perfect.
Mind that this post and the attachment are confidential, and should not be shared or discussed with anyone out of the development team or the management, not even the crew.
Introduction
About 5 months ago Las Venturas Playground 2.90 was released. After a long time of 2.x versions, it’s
time for something new: Las Venturas Playground 3.0.
The code of 2.x versions is quite the mess, years of developing spread the code into too many files and
there are duplicate variables for various things. Even though some code was cleaned up for the release
of 2.90, it is still too messy. LVP 3.0 will be a complete rewriting, and it’s very important that this code
stays clean; therefore there will be extensive guidelines for developers working on it.
We also think it is important to get the players of LVP more involved into the development process. The
first steps for this have already been taken, by making our bug tracker, Mantis, public. The next step we
want to take is to make LVP 3.0 open‐source. We’ll be using a project manager, possibly Google Code, to
make the entire gamemode publicly available. The SVN will have read‐only access for everyone, but only
developers will be able to commit. Players are allowed to add features or fix bugs by sending us patches,
which are .diff files. After we revise them we can merge them into the gamemode, send them back to
the author for some changes or deny them entirely. This will allow active players that are able to write
decent code to be very involved in the development process without being in the actual development
team or having any obligations. They will however, have to follow our coding guidelines.
An important change in the code of the new version is the use of a pre‐compiler. This pre‐compiler,
which was developed by LVP‐Media, will allow developers to use new features in PAWN. This includes
classes and some other features. More information on the pre‐compiler can be found in the
documentation of the precompiler which is on the current SVN.
To prevent people from ‘stealing’ the new gamemode and running it on their own server, a plugin will
be needed to run the gamemode. This plugin will of course not be available for the public. Details on the
actual plugin are still to come.
We plan to begin working on LVP 3.0 in the near future. We’ll start with making an extensive plan for
everything we want in 3.0, including the guidelines for developers. After we have a good plan the
current development team will start by making a solid base for the gamemode. After we think we have a
base that others can use to develop more features, we will release it to the open SVN. After we have
done this all code will directly go there, no more private code. A test server will be running to find any
bugs and test new features. Its gamemode should be compiled on a daily basis so it includes the latest
changes. As soon as we have something that is good enough to replace the current 2.90 gamemode
we’ll have to test it extensively and decide on a possible release. In the mean time LVP 2.90 will also
need to be maintained, though the focus should be on 3.0. Fixing important bugs on 2.90 should still be
done, though it is not necessary to develop new features.
I'm not too sure about the code being public and the fact players will be able to commit their code. I mean, how many Pawn developers are in the community? (Not those who just copy-paste, but those who actually understand what they are doing).
@ Wesley: I don't have a clue how that would work as I don't know PHP, but I'll look into some tutorials and see how hard it is.
(07-08-2009, 08:01 PM)Matthias link Wrote: I'm not too sure about the code being public and the fact players will be able to commit their code. I mean, how many Pawn developers are in the community? (Not those who just copy-paste, but those who actually understand what they are doing).
There actually are quite some people that know pawn. Also, making LVP open-source would make people more interested. People want new features, with open source they can create those features themselves. It'll allow the players to be more involved, it'll be less work for us, and cool features could be the result. So why not?
(07-08-2009, 08:06 PM)Wesley link Wrote: There actually are quite some people that know pawn. Also, making LVP open-source would make people more interested. People want new features, with open source they can create those features themselves. It'll allow the players to be more involved, it'll be less work for us, and cool features could be the result. So why not?
I'm not too sure, since people can't run the gamemode themselves it might lead to a lot of untested code being committed and we'll have to spend lots of time on it anyway.
As for PHP, I'm completely open to it if we can make it work properly, it's a lot more powerful and flexible than Pawn.
(07-08-2009, 08:18 PM)Pugwipe link Wrote: I'm not too sure, since people can't run the gamemode themselves it might lead to a lot of untested code being committed and we'll have to spend lots of time on it anyway.
As for PHP, I'm completely open to it if we can make it work properly, it's a lot more powerful and flexible than Pawn.
People will not be able to commit directly on to SVN, they'll be able to send us patches which we can merge into the code. As for testing, that indeed might be an issue, we should think of something that would allow developers to test their code on some sort of testserver, without giving them too much access.
The only thing I can see the open source move beneficial to is the SA:MP community. The LVP community is quite large, but let's face it, quite stupid to. If people where going to be interested enough to commit code to 3.0's code base without actually being able to play and test it themselves right away then they simply would have applied to become developers. As Matthias stated I don't think the community has many experienced coders - looking at the scripting section in the SA:MP board of this forum shines a light on it.
In my opinion PHP would be better in the maybe pile. The only four developers that have knowledge of PHP are Peter, Wesley, Pugwipe and Badeend. With Wesley being the only one active of the four, I can see PHP eventually resulting in the projects death. With the Pre compiler already finished and offering a lot more capabilities and advantages over pawn, I don't see a reason not to stick with that. It looks really easy to and wouldn't require fucking about getting PHP to work on a SA:MP server.
I agree with Jay. Wesley is the only really active one, and let's face it you can't learn a scripting language in a week or 2.
I'm for keeping LVP on Pawn.
(07-08-2009, 08:24 PM)Jay link Wrote: The only thing I can see the open source move beneficial to is the SA:MP community. The LVP community is quite large, but let's face it, quite stupid to. If people where going to be interested enough to commit code to 3.0's code base without actually being able to play and test it themselves right away then they simply would have applied to become developers. As Matthias stated I don't think the community has many experienced coders - looking at the scripting section in the SA:MP board of this forum shines a light on it.
In my opinion PHP would be better in the maybe pile. The only four developers that have knowledge of PHP are Peter, Wesley, Pugwipe and Badeend. With Wesley being the only one active of the four, I can see PHP eventually resulting in the projects death. With the Pre compiler already finished and offering a lot more capabilities and advantages over pawn, I don't see a reason not to stick with that. It looks really easy to and wouldn't require fucking about getting PHP to work on a SA:MP server.
Jay
The functionality that the precompiler adds to pawn is default in PHP. And PHP really isn't that hard, you can learn from tutorials and other code quite fast in my opinion. You can start with the basics, that are similar to pawn, such as if and else statements, the only thing that's really different is the use of variables, PHP does not need you to specify the type, especially strings are way easier to use in my opinion, no need for format anymore, though there are functions that are very similar to format and would allow you to keep using that syntax.
