Refactoring and compiling with Lazarus

[Shoozza]

Hello Community,

you might have wondered about what was going on since the last devlog post.

Well not much sadly.

But stuff is moving again so lets see what happened lately.

As you might know the server and clientserver code aren't really in sync (if not you do now ).

One reason for this is the outdated server. We had to update it to work with the new client version.
The other reason is that the clientserver never got lots of love.

You might remember that MM gave EnEsCe the soldatserver source in 2006 or so but not the clientserver source.
Thats why soldatserver has scripting and the clientserver doesn't.

But that is not the bad thing. The bad thing is that we have to keep both versions up to date and we actually can't share the code between server and client because they are that much out of sync.

That's why I started to refactor the netcode stuff so we could finally share code.

In the other news: Soldat server compiles with FreePascal/Lazarus again.
For some time the added hardware id stuff was preventing that.
But as it turned out it wasn't as difficult to fix as I though so now it compiles again.

You could ask: "Why do you care about Lazarus?"
Well, we want to get Soldat opensourced someday (if MM agrees) and FreePascal/Lazarus is available for free (and opensource too).
It also helps us to find bugs because it's a different compiles which gets other errors.
When our code stable enough it will allow us to compile the server for 64bit Linux and Mac too.

Though currently it's compiling for windows only.


Ok that's enough for now.

Sorry for making you wait so long for the news! Cya!

[Stuffy]

Wooo nice

[L[0ne]R]

Derrrp herrrp... I hate to admit it, but I didn't get what most of this post means. <_< Are you saying that you're completely rewriting parts of the netcode? And what about problems with the anti cheat? Were they finally solved?

Anyway, nice to hear that things are progressing again. Wishing you more good luck, in case your supply of it is running low since your last post.

[Fryer]

Basically, he's doing huge changes to the code structure to make it easier to maintain. That will, for people playing soldat, mean that stuff will be easier to fix and updates will probably be quicker.
As all devs have other stuff to do atm, not very much is currently being done to the anti cheat, but when Shoozza gets back from rl he will probably work hard on getting it done asap.

Happening right now: I've gathered lots of bugs from stuff people said on the forums and irc and I'm trying to fix as many of them as possible, hopefully before Shoozza gets around to finishing the anti cheat stuff.

[jrgp]

I think perhaps the most significant thing here is the ability to compile the soldat server natively for, as shoozza mentioned 64bit linux/mac, but also FreeBSD and other unixes. Not everyone only uses Linux and I think it's time people started giving that fact more attention.

[Clawbug]

I think perhaps the most significant thing here is the ability to compile the soldat server natively for, as shoozza mentioned 64bit linux/mac, but also FreeBSD and other unixes. Not everyone only uses Linux and I think it's time people started giving that fact more attention.

Edit: Duh, I missed the point. Proprietary software on FOSS platforms should be outlawed.

Why would it NEED more attention though? As long as the OS follows the SUS there should be no problem. The only thing which could disrupt this which I can think of is incompatible libraries but that's something which can't be solved easily(unless the libraries are statically linked, and even then there might be a need for full library recompilation due to ABI issues). ELF is terribly flexible format.

Edit: There are many differences between UNIX-like kernels which are not defined by standards but those should not(in ideal case) infer with the program(unless the program relies on something platform-specific, bad bad coder!). We can only hope that the devs are familiar with portable UNIX programming and the vital differences(which are eliminated mostly by the standards though).

[VirtualTT]

I hope Shoozza won't end up "compiling for macinux" again like he did with oS... -_-'
It has been almost a half of year since "new dev team" came into view but half-dead soldat community still haven't received any clues on when new Soldat version is going to be released (or even is there going to be a new release?). You guys should've  been doing Soldat a CPR, not broadcasting your weird "open-source" and "mac" fetishes on it. It sure needs a lot of cleanup and maintainability improvement, but at this rate soldat would drop dead completely.

PS i hope client server would be cut off. It sounds more promising if soldat client could just automatically launch a normal dedicated server (with separate exe) when user creates a new game.

[jrgp]

I hope Shoozza won't end up "compiling for macinux" again like he did with oS... -_-'
[...]
You guys should've  been doing Soldat a CPR, not broadcasting your weird "open-source" and "mac" fetishes on it.

Spoken like a true windows troll

In the case of the dedicated server, supporting other platforms is not a gigantic task since it just involves slightly changing things around and using a different compiler.

[VirtualTT]

I hope Shoozza won't end up "compiling for macinux" again like he did with oS... -_-'
[...]
You guys should've  been doing Soldat a CPR, not broadcasting your weird "open-source" and "mac" fetishes on it.

Spoken like a true windows troll

In the case of the dedicated server, supporting other platforms is not a gigantic task since it just involves slightly changing things around and using a different compiler.

In the case of the dedicated server, supporting even a single platform is a gigantic task that haven't been accomplished yet.
note that by "supporting a platform" i don't mean current situation like "it runs somehow"

It's obvious that ppl would get more exited if they get news about some bugfixes or new features rather then about server compilation on some crapuntu they never heard before. Soldat (at least for ordinary players) has always been a windows game and i like it that way.

[L[0ne]R]

Actually... I think it would be for the best if 1.5.1 comes out later, but with a LOT of bugfixes (it may even be worth making it 1.6).
This way release of a new version would create a much stronger effect. It would feel more like rebirth of the game, rather than just another small update.

The community would also get more time to do something to freshen up the game - fix more default maps, make some good brand-new ones, write a how-to-play tutorial with pictures, which could then be integrated into the Soldat menu, maybe also improve default and custom interfaces, make some new headgears and hairstyles.

[Clawbug]

Actually... I think it would be for the best if 1.5.1 comes out later, but with a LOT of bugfixes (it may even be worth making it 1.6).
This way release of a new version would create a much stronger effect. It would feel more like rebirth of the game, rather than just another small update.

The community would also get more time to do something to freshen up the game - fix more default maps, make some good brand-new ones, write a how-to-play tutorial with pictures, which could then be integrated into the Soldat menu, maybe also improve default and custom interfaces, make some new headgears and hairstyles.

No.

The ONLY way is to release early and to release often. There are far more benefits with such release cycle. The only downside is that people actually have to install a new version and servers need to be updated. Not much of a problem.

But as always, with Soldat that is just nothing but a mere dream; the next version will be 1.6 and will be released in late 2011.

[Fryer]

The only downside is that people actually have to install a new version and servers need to be updated.

Not a problem if it updates automatically.

[duz]

More two years waiting the next version. Ok.

[Meteorisch]

Actually... I think it would be for the best if 1.5.1 comes out later, but with a LOT of bugfixes (it may even be worth making it 1.6).
This way release of a new version would create a much stronger effect. It would feel more like rebirth of the game, rather than just another small update.

I like this, I mean we're stuck with the words soldat 1.5.1(or whatever) I mean we've expected this version for such a long time now, 1.6 would fit better and awards us a big upgraded version.

[Clawbug]

The only downside is that people actually have to install a new version and servers need to be updated.

Not a problem if it updates automatically.

It should but it won't.

[Monsteri]

The RC should have released months ago.. I do trust the new devs and I believe, that they are working on it, etc etc. I don't want to hurry, but I still do. HURRY!

[Illuminatus]

The RC should have released months years ago.

True, true. Imo it has enough bug fixes and changes (mostly from good-old EnEsCe) and doesn't have major disadvantages/problems compared to the current version.
Waiting for the anti-cheat API makes no sense imo. You have an autoupdate-feature - make use of it (finally, goddamn). There's not really much to prepare for a release-build. We're not Mozilla who do huge (automatic) testing and qa before even releasing an alpha-version. Simply take a week and have no fear to release the shit (v1.5.1). What you're waiting for?? What could go wrong if you do just do it?
Just tell the people that it's not a major huge update (which it actually is, check the changelog, zomg) but a smaller one. Assure them that there will be more small updates. Make them hot about the upcoming default anti-cheat. I don't think they will be disappointed (at least there're would be some changes).
Then you'd have some time to fix some more introduced bugs and go for the anti-cheat API with v1.5.2.

[Shoozza]

the "huge (automatic) testing" made me laugh
Imho the anticheat stuff is very important that's why I want to have it in the next release.

I'll try to get the stuff done the next few weeks.
The dev teams activity was kinda low after the last beta though :<

You might understand that I feel a bit uncomfortable releasing stuff with tons of bugs, but I guess in the end it can't helped.

[Meteorisch]

its sad but i have the feeling Shoozza is the only one doing something despite the number of developers

[Bistoufly]

Could we have a beta before the implementing of the anti-cheat?

I'd like to test the fix for the "mouse buttons get stuck" bug.

Thank you.