[PD-dev] CVS account submission

[IOhannes m zmoelnig]

Alexandre Quessy wrote:
> Hi again,

> [concat] is different from [list2symbol] because it doesn't need to be
> initialized with a null symbol (!!) in order to concatenates atoms. If
> you don't wnt it to be in the CVS, it doesn't have to, but I find this
> name meaningful. It's just a way not to use [makefilename] for simply
> concatening symbols.

it was pure interest on my side why you would need such an external when 
there is already [l2s] which provides exactly this functionality. (i 
think that the need for one additionaly [loadbang] and [symbol( is a a 
fair price for having the possibility to choose any "separator")

by no means i want to be a censor of what goes into the CVS and what not.
and i don't basically object to your submission
though i don't think that "learning C" is enough of a motivation to need 
write access to the pure-data CVS - but again this is not an objection.

> 
> I will put my externals and abstractions in a directory named "muzaq", I
> think. I am hesitating between a set of externals, or a library. Is
> there some neat points that would make me choose such a format instead
> of several C files ?

note that how you organize your C-files does not necessarily have 
anything to do with whether you produce single externals or big libraries.
e.g. i build zexy as a library even though each object has a separate 
C-file.
i am pretty sure that there in the externals/build system there are 
several externals whose code is in the same C-file (although this is a 
waste of ressources).


basically i would advise you to structure your C-files in a modular and 
reusable way; whether you compile as external, library or whatever is 
then a matter of (other!) preferences.


fmg.adsr
IOhannes