I have seen that the mgdev wiki and svn URLs are "classified information" for an unknown reason. Why is this so? Shouldn't open development be encouraged? I am sure work on mgdev would go much faster with community support. Some bugzilla-like system should be installed so that bugs, feature requests and patches can be posted by developers outside MG. This is Trem 1.2 after all... why shouldn't everyone be allowed to contribute just like in all previous versions of Tremulous?

This isn't trem 1.2
Its balance and other testing for norf, which may end up in 1.2.
And the links are hidden to stop the retards getting in the way.  Coders are welcome.

Furthermore, you are somewhat able to look at the code. We're not distributing, so it's legal.

At the very least, this has meant that the aimbot developer that had the nuts to come to MGDev and ask me where he could get the code for testing his latest version went home empty handed.

If you were truly interested in coding for us, Amanieu, come to #mercenariesguild and talk to us about it. We'll set you up with access, probably.

If the patch in the pk3 isn't working, you need to tell me so I can fix it

I actually mention this because I am also working with some French devs in their DnC (http://dnc.tremforges.net/) project, which is a fork of trem. We currently have a beta version of DnC based on tjw's balance changes, called madcat, but with different features. We are interested in joining the MGDev team. Would it be possible to arrange that?

You're distributing: if someone go on your server, he download your pk3.
There is patch on pk3, it's legal, but you're distributing.


I don't really understand. If you are afraid for aimbot, why do you contribute to a free game ? It's easy to program an aimbot on it, and there is no good way to counter aimbots. More over, compatible aimbot under 1.1 will be easly compatible with MGDev, and the patch will probably be enough to fix little changes to make.

But patch is not enough for over developpers to re-use your modifications.


If you read opinion of major part of the team Amanieu, we are not interested in joining MGDev team, but to contribute on it and work together

In fact, we manage a développement like MGDev, with the help with french and italian community. We try gameplay rules, set it on servers (european servers like cygg.fr or thcteam.org) and wait community feedback. Then, we make modifications on gameplay rules with their feedbacks.

I talked to Lakitu and other MG members about zone buildpoints, it is one of things we made.

We also plane projects like VOIP implementation in tremulous client, lua script for some callbacks (to make scenarios) and config, log rewrite to make it more easy to parse, rewrite admin system with merging / and ! commands (and make custom commands with lua configs), and we have some other ideas.

With have about 7 developpers in the team, and we are active. I think it's useless to MG and DnC to work on *same* things because a team or the other don't want to share their stuff (or don't give a way to get it easly, it's impossible to analyse a 6000 lines patch each time someone need a feature), so if we can get your gameplay features easly, you will probably retreive faster features we plane.

If a feature I descrive upper have any interest to your mod, we agree with the idea to make it as a priority, just say what you need. In any case, our svn has anonymous read access, our task list is free and you don't need a registration to post on it, our ML is free to register too, but we will be very hapiness if sharing could be done in two directions.

Friendly,

Hmm, good point on the qvm downloads. I suppose we are distributing part of it. Hmmmmmmm....

But what I meant with the aimbot developers, if we can put off having them be able to port to the new version as long as possible, that would be nice.

As for the rest of this post...

Disclaimer: I cannot speak for everyone on the team. I have attempted below to represent the wishes of the other developers as best as I understand them. Sorry if I have not been altogether accurate.

I am sure that we would love to have some help on the team, especially organized help. You say you're working on a fork? From the site, it looks more like a mod or a patch compilation rather than a fork. A Tremulous fork is not something that we want to happen, really Or at least I don't. A lot of the developers that work with us have put a lot of effort into Tremulous 1.1 and 1.2. Though competition can be good.

When you say zone control, you really more or less mean Domination, as written by risujin about a year ago as a Tremulous mod instead of a Tremulous distribution. By all accounts, the core developers do not want domination in the distribution of Tremulous.

The 1.2 release is meant to be a "Tremulous as it was meant to be in 1.1" release. The current work mostly has to do with balancing the gameplay according to the 5+ year Tremulous contributor, Norfenstien. Other work we do is slight graphical enhancements and a few other things. We are, at the moment, not trying to create a different game, just fix the current one to be stable and clean.

However, having said that, it is probably a good idea to get you people involved if you choose to do so. As an open source organization, we should be open to developers from all over the Internet. However, it seems to me that you people are interested in making a fork, or a mod at the least. This does not really coincide with our aims for Tremulous 1.2. I do encourage you to join the 1.2 development staff, though, if you are willing to work on our aims.

One last thing to mention:

A lot of the core developers agree that this upcoming release should fix balance issues. Issues include Tyrant being far too good, dretch having a massive hitbox, barricades getting in the way, and a variety of other nitpicks. However, after 1.2, game changes are likely to start happening. One of the things I personally want to see (and I do not speak for the others) is a gametype browser, to let you choose what form of mod/game you want to play, and support for those mods in the core distribution.

So in short, you're welcome to stick around if you like. We'll be in #mercenariesguild on irc.freenode.org if you want us.

Sorry for the misunderstanding here. We don't plan on make a completely new game, although rewrites of some major parts of the game are planned. (mostly internal: new admin system, scripting support, replace GUIDs with public-key identification, internal VoIP support, etc...) We are just distributing a customized version of the tremulous client and server, with patches from a lot of different places, including some of ours. We currently have 3 versions: eggy (bugfixes, minor new features, applying existing patches, etc.., currently being worked on), kitty (major rewrites mentioned above, will be based on eggy, will start working on it when we are done with eggy), and madcat (basicly all of the eggy patches + tjw's gameplay modifications, kitty modifications will be brought in too when they are written and bugtested, also currently being worked on, we should merge this with mgdev). The only problem that is preventing us from collaborating is that most of the DnC devs are against having a "closed development" system, in which users are not able to provide feedback and other coders to contribute. This is why I suggested a bugzilla-like system, like Trac on our website, to open up the project. Also please enable anonymous SVN to make contributions easier.
Same here
I don't believe this is a valid argument, because you already publish the code in your pk3s and if the aimbot dev really wanted to get the code he could just use the patch in it. Instead I think MGDev is not popular enough to attract the attention of aimbot devs, and whether now or later, it will only take them about a week to port the aimbot to the new version, which in both cases will not be significant enough to justify this kind of closed development. As Azrael07 said before, if you wanted an aimbot-free game, then go develop a closed-source one. DnC, as well as all trem mods(tremx,arcade,balance) aren't well known enough to attract aimbot devs so why should MGDev be, at least until it becomes trem 1.2.
Wrong: http://dnc.tremforges.net/trac/wiki/ZoneBuildPoints
A more important (from my point of view) addition would be a filter to filter out any server without unlagged (no unlagged with 300 ping sucks ).

Amanieu make a confusion about the fork. Call it with word you want. It's come from a patchset, and it's now a kind of mod. Some month ago, we hadn't see any real activity from developper and we didn't know MGdev project (I think it was not started), so we said DnC *could* become a fork, if we no longer see any activity from developpers. It seems to have activity, so it will not became a fork.

We have two main objectives:
- First, gameplay improvements, with the madcat branch. We use TJW patches, and we try stuff and get players feedbacks, to make better stuff and get a very balanced game. We don't hope only a balanced game, but also a fun game, so we reject balanced improvements if there are boring.

- Second, game improvements, with the kitty branch. The main focus is to add to tremulous better support to many things, about client, about server, about masterserver. I gave some exemple upper of features we plane to provide.

If our second objective is not compatible with MGdev, the first is.

But, as say Amanieu, and as other members said on our Mailing list, we don't want to join a closed developpement. The fact is we don't want a private svn access to MGdev for our members, we hope a anonymous read only access for everyone. We develop for tremulous because we like this game, and also because it's a free game, and we want to develop with the spirit of free software. So, we don't want to join a team where wiki page can't be given in public "to prevent from spam", and where svn anonymous is denied "because hoster don't allow it".

So, sure it could be a very good idea to work together about gameplay improvements. I don't care if the result will be named "tremulous", "MG", "DnC" or whatever, I don't care if my name will be in credits or not, but I really don't want to develop under the secret without sharing.

Most part of DnC team have same opinion.
If you need help to find a free and stable svn hoster, we have ideas. If you need ideas/help to set up a trac or any developpement scheduler tool, we will be happy to help you.

And if you agree with there simple things, I think DnC team will be happy to join MGdev developpement (it will be probably voted).

I'd just like to point out that what MG has done with development is already more than people got through trem.tjw.org.  The patch for client-side changes was in the PK3, the server balance mods was on a webpage, and I don't even think the two fit together without a lot of hand-hacking from benmachine and others.  And yet I don't recall anyone giving him a hard time about how he handled his own modifications repository for testing and balance changes.

I'm not with Garoth on the aimbot thing. The code comes first, and if we have a choice between making it easy for everyone and easy for no-one I'd take the former every time.

As far as I know, we don't allow svn or http because svn+ssh is merely easier to set up, and for our purposes we've never needed anything else. People have asked for the source, yes, and I've directed them to the pk3. The SVN is publicly viewable to anyone, but we simply haven't the infrastructure to allow people to checkout it easily, and I personally haven't yet felt the need to put Rocinante or someone through the bother of creating that setup.
Of course, with your offers of collaboration, that prospect becomes more interesting, but to be honest if it means creating a great deal of work for ourselves then don't be surprised if we are not overenthusiastic about the idea.

On another note, I'm interested in your zone build points concept, since we have discussed the idea ourselves. Have you solved the problem of the conkit's BP display no longer being accurate for each user? Our primary obstacle with implementing this idea ourselves is finding an intuitive way to inform the builder of how much BP is available for his or her local area.

edit: I've just learnt that STAT_MISC2 has just become STAT_UNUSED. Watch this space?

With the current set up, adding public svn isn't an option.  We would have to migrate to source forge or some such, at the moment I don't see any real benefits of doing so.
ATM mgdev hasn't posted on trem.net or anything, and isn't known about by the majority of people, and as I understand it that the way norf wants to keep it until its a bit more stable.
Due to this lack of knowledge, there hasn't been much need for bug trackers etc, the wiki serves us fine.  (People can post bugs there, although is just easier to say them in IRC).

Both the wiki and svn are public, there are no passwords etc on them.  All we ask is people don't go posting the links everywhere, although I'm sure that will change when public testing is asked for.

Also, IMO the aimbot thing is pointless, they can port quickly, and most of our changes wont effect them.  The patch is in the pk3 anyways >_>.

I'm sure Rocinante can set it up. Since there is a need now, I think we should setup public checkouts.

Read their site. They solved it by limiting the amount of BPS you can put on any one BPS source. So your build points say "30 pts remaining" but when you try to build it says "Reactor has too much BPS on it, go somewhere else".

Come to our IRC and we will give you the links.

That's not entirely true. I have already made several client-server communication changes that limit the amount of information the server will send to the clients. THZ's egg counter and such should no longer be possible.

I didn't say that >_> also, we mean 'the aimbot thing' as in 'making the source annoying to get hold of so aimbot coders have a harder job'.
I don't think it's an issue since at any given time most of the people on the server are admins.

If that is how we feel, I most certainly agree that we should open up the svn access. As a huge open source backer myself (using Linux for years) I fully agree that an open system is good.

The reason I was a bit wary was for the exact reason that MGDev, as the "official" 1.2 development branch, _does_ attract that kind of attention. I _have_ had someone I'm mostly sure was an aimbot developer come and ask me for the code.

However, I 100% agree with risujin that working to make their lives harder is more important than keeping the source closed. I'm all for limiting them, adding metrics, and several other things I'd rather not talk about on a public forum. The one thing I want the authors of those patches to keep in mind is that we can't afford to sacrifice efficiency. Forcing servers to be smaller would be bad.

Anyway, the way this is going, I totally agree. It sounds like the DnC team has a lot in common with us in the sense that we hope to polish up the game, add more power to it, and release some mods on the side for when we get bored (not in the official release).

Two things I'm worried about:

1) Norfenstien is having trouble assigning tasks already. It's hard for one man to keep up, and no one else is qualified.
2) A lot of changes that even we make may not make it into the official release, pending Timbo/TJW's discretion. We have one major dev, but the others may veto some changes. Are you DnC crew willing to live with that (considering the knowledge that your code can still be used outside of the official branch)?