[freeciv-ai] Re: time table for ai restructuring

[Petr Baudis <pasky@xxxxxxxxxxx>]

Dear diary, on Tue, Jul 16, 2002 at 02:27:51PM CEST, I got a letter,
where "Per I. Mathisen" <per@xxxxxxxxxxx> told me, that...
> On Tue, 16 Jul 2002, Petr Baudis wrote:
> > I don't think "_move_ the AI sources physically" is a good idea. IMHO it 
> > would
> > be better to rather shortly build up the AI in client/ai/ once more from the
> > original AI, keeping it working properly all the time and adding original AI
> > stuff as it gets ready to live in a client (that WILL be painful; happy
> > settling ;). This way, nice review to the Syela AI will be given as pieces 
> > will
> > get copied and more importantly, in CVS there will be one AI which will be
> > still working, possibly evengetting some bugfixes inside, while the second 
> > one
> > will grow up hastelessly and also actually _working_, though not completely.
> > Also, if the development of client-side AI would stop for any reason (see
> > Syela), just the AI client can be disabled in release, still having the
> > server-side one.
> 
> Do you mean building up a new AI in client/ai/ _as a client-side AI_?
> Having two simultaneous server-side AIs won't work.

If it will live in client/ai/ and _won't_ be client-side AI, don't place it to
client/ at all. It's terribly confusing, I'd say :). What's wrong on just
keeping living in ai/ then? It won't even reuse client/common etc, just
brain-teaser "why the hell they planted it into such a silly place" ;) for
everyone coming along.

> My only beef with this is that unless it happens fast (you say both
> shortly and hastelessly so I am not sure which you mean;)

It will depend on people how long it will take, and one can only very hardly do
predictions about time spent on an opensource project. But the point is that
there won't be any *haste*, which can encourage developers, but it can also
drain and kill them.

> there will be an additional problem of either sync'ing the two or halting
> development on the server-side one.

Natural selection.

> Potentially we might be capable of "outsourcing" some code that might be
> shared, though.

-- 
 
                                Petr "Pasky" Baudis
 
* ELinks maintainer                * IPv6 guy (XS26 co-coordinator)
* IRCnet operator                  * FreeCiv AI occassional hacker
.
You can get much further with a kind word and a gun than you can with a
kind word alone. -- Al Capone
.
Public PGP key && geekcode && homepage: http://pasky.ji.cz/~pasky/