Le 19 Décembre 2005 16:54, Mathieu Bouchard a écrit :
In practice all branches and subprojects in pd/src are based on very small teams of size comparable to that of DesireData.
That must a "desirable" pattern then.
Yeah. Well, adding features requires adding code, and get this, getting a piece of software (Pd) to use another piece of software (MsgCat) requires adding code too!
I already mentionned that Msgcat is harwired in Tcl/Tk since version 8.1. I think that using Msgcat would not add code or a library.
Also I don't see MsgCat as saving me work right now, and most importantly, I don't see my current system as difficult to migrate to MsgCat whenever you finally do convert me to holy MsgCat.
I'm not on a mission ( that would be a waste of energy, n'est-ce pas?)
MSW is a platform, not a branch or a fork...
When both are implemented as #ifdefs, what's the difference?
Why is a "platform" more #ifdef-worthy than a "fork" to the point that you deem acceptable that the "platform" has 8 times more #ifdefs than a subproject which is in the same branch and that you call a "fork" ?
You work on a radical new feature, almost a new software, it's not like a port. That's why the idea of a branch, which is less extreme than a fork, could be considered; that was the only important part of my message; I was mentionning the "#ifdef" as an example, not to be taken too litterally.
OK, I'm going back to safe lurking mode, rejoindre la majorité silencieuse et la plèbe gluante des utilisateurs occasionels qui suivent aveuglément la minorité active des vrais développeurs de PD. With joy and confidence. -- Marc