Crossfire Mailing List Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Rewrite (was Re: Classes, Race, Experience proposal)



 Also, a re-write is not likely to stop many releases.  In fact, it might
be prone to create more, as a re-write is likely to introduce all sorts
of bugs.

 Right now, I try to make releases about every 2 weeks to keep intereste
and so that developers are working on code that is pretty much the same
as what I have.  If I did releases every 3 months, several things would
happen.  First, soem people would be frustrated that some bug they have
has not been fixed, or not seeing any new features.  And I would have
more problems if someone sends in a patch near the end of that 3 month
period, as the code they made a diff to is probably different than the
code I have, making it so that I have to do the patch by hand.

 I'll agree that a lot of the code needs to be cleaned up (in the
last release (0.91.1), I think I did a pretty good job on the
treasure.c file)

 A lot of stuff needs to be cleaned up some, so it is easier to understand.
And in fact, I think a lot of the code right now is in pretty good shape.

 So if a re-write is done, what will really be changed?  The topic was
initally brought up making it so that the class specific stuff could
be easily separated.  The thing is, can that be done in a reasonable
fashion under any set of code?  Certainly, many of the structures are
going to change, and in some systems, what certain items do might also
change.  It might just be easier to branch off a separate developement
version, because so many functions would need to be changed.

 So I guess the first questions should be:  What do we hope to gain by
a major re-write, and is this feasible under any situation?

 Also, any re-write teams needs to work pretty closely together.  Several
people working separately probably won't generate anything better than what
we have right now.

 --Mark