Yeah, I think the complaint is mainly about possible loss of precision so (I may be wrong), the float to int cast might be fine.
On Dec 2, 2017, at 11:57 PM, Miller Puckette msp@ucsd.edu wrote:
Or (aha) - I could make up new function names for the "int" versions and change the pd sources to use them, and declare the existing ones obsolete...?
In fact, is there any reason one can't just globally replace every call to atom_getint and atom_getintarg with the atom_getfloat equivalent - let externs blithely call atom_getint and get a t_int back all they want.
That would touch a lot of files so if I do it perhaps I should make sure to do all the PR-merging I possibly can beforehand.
AND: there's no reason I can't assign a float to an int without a cast, is there? As I understand it the only clang complaint is int-to-smaller-int conversions. So int x = atom_getfloat(&atom) is still kosher, correct?
thanks Miller On Sat, Dec 02, 2017 at 02:50:57PM -0800, Miller Puckette wrote:
I think the use of "t_int" in m_pd.h is incorrect - it should have been int. But it's a mistake I think is now ironed in and we're stuck with it.
M
On Sat, Dec 02, 2017 at 10:25:07PM +0100, Dan Wilcox wrote:
I was following IOhannes' prompt about t_int: "rule of thumb: never use it for anything but passing data to perform-routines."
On Dec 2, 2017, at 10:22 PM, Miller Puckette msp@ucsd.edu wrote:
I'm pretty confused about this. I believe it was "t_int" in 0.48-0, and I see that your PR changesit from "t_int" to "int" - and I believe it has to be "t_int" for back compatibility...
cheers M
On Sat, Dec 02, 2017 at 10:16:44PM +0100, Dan Wilcox wrote:
I think I had already fixed this: https://github.com/pure-data/pure-data/pull/223 https://github.com/pure-data/pure-data/pull/223 (?) Or am I missing something?
On Dec 2, 2017, at 8:40 PM, Miller Puckette msp@ucsd.edu wrote:
I had one small ouch: I don't think I can compatibly change t_int to int in m_pd.h (this is mentioned on another thread somewhere). I don't know how to make clang pipe down about this short of casting almost every call to atom_getint*() in the whole tree. Yuck... Maybe it's better just to tell clang to be more permissive (if that's possible)?
Dan Wilcox @danomatika http://twitter.com/danomatika danomatika.com http://danomatika.com/ robotcowboy.com http://robotcowboy.com/
Dan Wilcox @danomatika http://twitter.com/danomatika danomatika.com http://danomatika.com/ robotcowboy.com http://robotcowboy.com/
Pd-dev mailing list Pd-dev@lists.iem.at https://lists.puredata.info/listinfo/pd-dev
Pd-dev mailing list Pd-dev@lists.iem.at https://lists.puredata.info/listinfo/pd-dev
-------- Dan Wilcox @danomatika http://twitter.com/danomatika danomatika.com http://danomatika.com/ robotcowboy.com http://robotcowboy.com/