On Monday 12 March 2007, Eddy Petrișor wrote: > I've looked at the site, the game seems to be nice...
Ok I know, MenAreAnts' graphisms are ugly but I am alone on this project (yeKcim wants to help me and begins to make graphisms). > Although the needs might be similar now, have you considered the > possibility that the needs might diverge? This needs are : - classes about screen (actual Surface, Sprite, SpriteCache, etc.). I think this need is similar to needs of all C++ SDL projects.. - classes about GUI (all in src/gui/). Now we needs EXACTLY the same thing. > Note that having an unique library will lead to a need to work in a > more organized manner than it was until now (you'll need to make > releases, tag the library, not destabilize the branches, make sure > game specific code does not enter the library, etc.). Yes, the library will only be used for global classes like Sprite, Surface, Font, Widget, ListBox, etc. Everything around a specific game will NOT be included in this library. > IMHO, the Wormux team still needs to organize itself so that > development is not that chaotic as it is now (look at the history of > the different releases - read tags in SVN - and you will understand > what I am saying). Having a library will mean that even more order > is needed. This is the wormux's problem. > > Still, compatibility needs to be preserved, so that a new > installation will have to migrate the settings from the old XML to > the new configuration. Lodesi told me yesterday he thinks it is a bad thing to have a depend with libxml. But if everybody think we have to preserv XML format (compatibility), I am ok, it was only an idea. > On the good side, seems that the arch independency is kept since the > configuration would be kept as strings. > > What I don't like about this approach is that it introduces > yet-another-new-and-particular-configuration-file-format which needs > to be parsed and *will* need its own library/functions, which, > although you say you have it already, I am sure it didn't got so > much testing as the libxml library and might be vulnerable to > different issues - not as robust, prone to break, not reviewed. Ok ok, you're right.. > I think this proposal is a part of the NIH syndrome often found in > programmers which are young and enthusiastic. I don't know.. > > Après pour l'aspect API des Surfaces/Sprites, je pense que garder la > > votre est très bien dans le sens où la mienne est pas terrible. > > > > Et pour les widgets, il faut faire des compromis je pense car moi j'aime > > bien la mienne :). > > But wormux' look better :-) This is your opinion, you didn't see my code, and I think there are too much classes in wormux about widgets. > > De manière générale, je trouve que l'idée peut être très intéressante, > > mais ça nécessiterait en tous cas un certain travail de réécriture dans > > nos deux projets, ainsi que pour la constitution de cette librairie. > > > > Voilà, maintenant je vous demande votre avis, vous développeurs de > > Wormux, êtes-vous favorables ou non à la création d'une telle librairie ? > > The library could be a good idea, but still a clear line must be > drawn between the code in the games and the code in the library. Yes, of course, I said my idea, not all specifications. > I am sorry if I sounded to harsh, but a library is not such an easy > deal as one might think and the initial analysis is way too > enthusiastic and misses some fine-but-important points. Ok of course, thank you for your point of view. Regards, -- Romain Bignon - http://progs.coderz.info _______________________________________________ Wormux-dev mailing list Wormux-dev@gna.org https://mail.gna.org/listinfo/wormux-dev
