Some questions about tmwAthena.

[grobb]

Hey there!

I used to administer a small/medium eathena server some years ago, and lately me and my friend (who is a graphical designer and does sprites and tiles in his off-time) has been talking about setting up a small game we can play with as a hobby. At first we thought of actually setting up a ragnarok online private server, but we got to thinking it would be cool if we could find a decent 2d client for eathena... and so after some searching I wound up here. And I must say I'm pretty impressed by what I'm seeing so far.

I do some programming in php and python, I can read code. But I'm no real programmer which is why we are dependent on having a framework that is being developed by someone who is

So, I got some questions relating to your projects here.

First of all, is the tmwAthena server development going to continue or are resources being shifted to manasource? I mean, will there be bugfixes and suchlike down the line?

I get the impression that the tmwAthena server is being rewritten at the moment? If so, will there be breaking changes in the custom server data structure? In essence, would it be sensible to wait for it?

How stable is tmwAthena (or manasource?) Has it been tested with a decent amount of clients? If we got parts of our old userbase from my old eathena server interested in our project, let's say 2-300 clients online at peaks, could this be expected to work nicely?

Manasource looks interesting, it has a nice documentation, but development seems a little sporadic, are there any eta's for it?

I don't require much for starters. Custom maps, graphics, monsters and items. Trade. I'd like to be able to customize skills at about the level you can do it in eathena, but restrictions are acceptable. I'd also like to be able to customize classes, but it's ok if this is just a cosmetic change at first (like the name and sprites.)

Well, now you know a bit about me and my situation (feel free to ask too.) Any advice in general relating to what I've said here will be appreciated.

Thanks in advance.

[Nard]

Actually the TMW project has been splitted on two parts:

• TMW itself keeps a server based on eAthena (TMWAthena) and develops actual game content.
• Manasource is in charge of developping a new server (Manaserv) and the client for both servers It also gave birth to a new game, Source of Tales ; to participate the LPC contest (http://lpc.opengameart.org))

[Crush]

Disclaimer: Although I was a developer for several years, I am no longer involved in the development of TMW or Mana, so not everything I write might be factually correct.

First of all, is the tmwAthena server development going to continue or are resources being shifted to manasource? I mean, will there be bugfixes and suchlike down the line?

TMW has a programmer who is focused completely on tmwAthena and who actively continues its development. TmwAthena won't die so easily.

If so, will there be breaking changes in the custom server data structure? In essence, would it be sensible to wait for it?

The refactoring of o11c already has some non-backward-compatible changes. The most important one is definitely the conversion of the syntax of the magic system (which hasn't got anything to do with the Ragnarok magic system) to a Lisp-like syntax. He should be able to tell you more about how much he intends to break compatibility in the future.

How stable is tmwAthena (or manasource?) Has it been tested with a decent amount of clients?

We used to have over 200 users at peak times. It didn't cause any notable problems I am aware of. I don't know if the Manasource team ever did a serious stress test with Manaserv.

Manasource looks interesting, it has a nice documentation [...] are there any eta's for it?

It is finished in a half year - for over five years now . No, there are no ETAs. Software development is already hard enough to estimate when you have a team of professionals working 40 hours a week. But when you have a team of volunteers who only work when they have time and feel like it, it is completely impossible to make any estimates when something will be completed.

I don't require much for starters. Custom maps, graphics, monsters and items. Trade. I'd like to be able to customize skills at about the level you can do it in eathena, but restrictions are acceptable. I'd also like to be able to customize classes, but it's ok if this is just a cosmetic change at first (like the name and sprites.)

Neither tmwAthena nor Manaserv have support for classes, because the initial idea was to have a classless system. Manaserv, however, has various tools to fake it pretty well. I had "skills" (called "special actions" in Manasource terminology - skills are experience levels in weapon classes, crafts or magic disciplines) working partially on Manaserv when I left. But there were substantial gameplay changes afterwards. I don't know in which state the system is now.

[Ablu]

I can tell you the status of gameplay critical features (that blocks building games with it) of the manasource project:

• Equipment / item handling is working
• Trade not 100% bug free
• Channels + guilds working -> lacks proper UI
• Fighting system is a bit buggy (We plan to rewrite it to a more action way of fighting)
• Scripting is quite good already I think (usuable but of course extendable)

However we currently look into rewriting the client with the base of QML/Qt since the old client is a bit hard to maintain (especially with supporting two different servers [tmwAthena <-> Manaserv].

We have a lot of cool ideas for Manaserv/Mana but we cannot tell how long it takes to implement them

About waiting for Manaserv:
You *can* build a game with it when you tweak around some stuff (mainly attack system) and help us to fix the bugs in it

Regards
Ablu

[o11c]

Brief overview of tmwAthena's current and future status:

• Parts of tmwAthena are not very scalable - in particular, the number of total number of characters in the "database" and the number of simultaneous server connections . I can fix these without breaking compatibility, although you will likely have to run a simple script during the upgrade.
• At some point in the future, there *will* be breaking changes to the script system. However, since there is a large body of existing script work, I will have to provide a semi-automated upgrade path. Avoiding esoteric features and functions of the scripting language will make this easier.
• tmwAthena only works on x86 (it uses -m32 for x86_64). Supporting other little-endian platforms is an early goal; supporting big-endian platforms will happen eventually but will necessarily be less efficient.
• platforms other than Linux will not be supported unless someone writes a bunch of implementations. In particular, epoll(7) and inotify(7).
• tmwAthena is tightly coupled and uses lots of global variables, so it's hard to figure out what to change without affecting other things. Cleaning up the existing code has to have a higher priority than adding new features, or even making things more efficient (although often using proper C++ types may make things more efficient as a side-effect).
• Python is an awesome language, and can be used to write pretty-printers for gdb. This would be an easy task for someone who only needs to be passingly familiar with gdb and C++.
• Maintaining compatibility with gcc 4.4 and gcc 4.5 is hard, and leads to me writing absolutely awful code. Even gcc 4.6 has quite a few quirks, but I've committed myself to supporting that indefinitely.
• clang 3.1 is awesome and *should* be able to compile all my code, but it has lots of bugs.