Re: [Super-tux-devel] New admins!

[Ricardo C. <ri...@ae...>]

Em Quarta, 16 de Mar=E7o de 2005 14:30, o Marek escreveu:
> Ricardo Cruz wrote:
> >Hello there,
> >
> > This is quite a suprise. I've been asking for a new maintainer for so
> > long, so this is definitively good news. Although, I think this is just=
 a
> > too bloated team of admins, so I'd suggest to attribute functions to
> > people. You know, webmaster, release coordinator, etc, and maybe even a
> > leader that should have the last word on subjects and have a view of the
> > road map.
>
> I agree that assigning functions and even a leader is a good idea. The
> best thing, however, would be if we could manage without one, that is,
> with lots of discussion in the team. I suggest we assign functions to
> team members in a meeting and see how much we need those functions.
>
> > Anyway, what the project needs right now, in my opinion, is a sense of
> >direction on the coding part. I personally don't like the way coding is
> >heading, and I currently very much prefer when the code was C. It is
> > getting too complex and, from what I've seen from the last re-code, less
> > flexible. Small games like Alex 4 [1] kick SuperTux's ass big time, in
> > terms of playability.
>
> We currently need direction on every aspect of the game, not only the
> coding. We need major decisions on looks, story, gameplay, features,
> code etc. that will be included in milestone2.

 I agree. I guess I find the coding part more important right now, because=
=20
this is where I can help and this is what's more active.
 But yes, especially the graphics are important to define, so that they loo=
k=20
a-like.

> As for the complexity of the code: I think you can't really compare
> SuperTux to games like Alex4 anymore as their goals are completely
> different. Compared to Mario games, SuperTux might be about to become a
> SuperMarioWorld or even more than that, while Alex4 will probably stay
> at a SuperMarioLand stage. Not that this is a bad thing, though, it's
> just completely different; and complex games require complex code. IMHO,
> SuperTux' playability is great; it can still improve, which requires a
> lot of care when implementing new features.

 I really don't think they're not comparable. They are both games and of th=
e=20
same genre. SuperTux might be more ambition yes, but it's still a Jump'n ru=
n=20
game. Something like a world map could be easily implemented in Alex4. If y=
ou=20
look at SuperTux's code that part is pretty much isolated in one file and a=
s=20
far as I remember it is not much bigger than the 1000 lines.

> On the other hand, I agree that the code is confusing in parts; it's
> really hard for me to get to fully understand it. There are probably
> still a lot of hacks that need to be replaced by proper functions (i.e.
> lots of things I wrote because I didn't fully understand the code :) ).
> However, the code in it's current state is rather buggy. I guess Matze
> left some major bugs unfixed after the Big Commit (some of which eat up
> a lot of resources); that is what I meant when I said before that
> currently, only Matze fully understands the code. But I'm sure he'll fix
> those as soon as he's got the time again. (*wink* :-) )
> However, I guess you noticed that some of his changes greatly improve
> handling the code, especially when it comes to implementing badguys,
> level objects and triggers.
> Anyway, the code should be a major topic in the meeting.
>
 I'd prefer to call it complex rather than complicated. I mean it doesn't u=
se=20
any magic stuff, it's just that things are so spread and rely on each other=
=20
and stuff. I don't like this kind of coding. Maybe I've spent too much time=
=20
coding C this year. :)
 About the eating resources thing, that might be because of the ammount of=
=20
STL, virtual methods and other expensive stuff use. I've recently developed=
 a=20
algorithm equation calculator in C, and it's amazing how optimized you can =
be=20
when you're so limited.

> > That said, I wouldn't like SuperTux die, because somehow it has a big
> > base of players -- I was even asked some information about the project
> > for an article about it for a quite popular Portuguese magazine (PCGuia=
).
> > So, what about organizing a few days in August or December for a meeting
> > and some hacking?
>
> About the article, that's great news! SuperTux seems to have a lot of
> fans, and we shouldn't disappoint them. I don't think it's going to die
> soon, as it seems we still all enjoy coding (and playing) it. And
> although we took a rather long break now, that doesn't mean SuperTux
> caught the DNF-Syndrome already. :-)

 Yeah, I also find it fascinating with the ammount of popularity SuperTux h=
as.=20
But notice that this popularity is actually derivated from the fact that it=
=20
is a SuperMario based, and is not completely ST's merit.

> Matze and I already thought about having a SuperTux coding session in
> Karlsruhe, during the LinuxTag days end of june. If anyone else is going
> to see the LinuxTag this year, it might be a good opportunity to meet.
> But first things first: We need a date for an IRC meeting soon. I
> suggest some day between March 29th and April 6th. Suggestions, anyone?
>
 I was looking forward a meeting a year ago, I'm not that excited about it=
=20
right now, and I don't think I would make it in the Summer holidays.

 About a IRC meeting, I'd prefer if it was in a working day. At the end of =
the=20
afternoon (say 6 o'clock GMT) or at night (9 h).

See ya,
 Ricardo

>     Marek
>
>
=2D-=20
The whole world is a scab.  The point is to pick it constructively.
		-- Peter Beard