Otavio Salvador <[EMAIL PROTECTED]> wrote: > Jim Meyering <[EMAIL PROTECTED]> writes: > >> Otavio Salvador <[EMAIL PROTECTED]> wrote: >>> Jim Meyering <[EMAIL PROTECTED]> writes: >>> Your proposed changes looks right but I do think we should provide a >>> method to make it easier? >> >> It's a one time conversion, so making the transition "easier" >> is less important than making the resulting code readable. >> Also, with the standard functions, you won't have to publicize >> the new interface if/when you want to use the same technique >> in another .c file. > > Yes. It'll need to be done once but we'll also continue to code and > then we'll use it every now and then.
But *everyone* knows about fputs, and no one knows about a just-invented printp wrapper, be it function or macro. > By 'publicize' you need the need of #include <foo> on the file header? If you use a wrapper function, then it has to be declared "extern" (not static) with a prototype in parted.h. > Well, while it's true we can also put it together with parted.h since > it's very specific. IMHO, it's better to stick to standard mechanisms when possible, just because they are well known. Of course, if a use requires a non-trivial %s format, %d, etc., then you'd use printf or fprintf. My point is that when just printing a literal string, there's no point in using the *printf functions. _______________________________________________ parted-devel mailing list [email protected] http://lists.alioth.debian.org/mailman/listinfo/parted-devel
