On 22/09/15 17:32, Cedric BAIL wrote: > Le 22 sept. 2015 02:30, "Daniel Kolesa" <[email protected]> a écrit : >> >> On Mon, Sep 21, 2015 at 11:24 PM, Shilpa Singh <[email protected]> > wrote: >>> cedric pushed a commit to branch master. >>> >>> > http://git.enlightenment.org/core/efl.git/commit/?id=abaf29cb768375957c9ee0b64d36034c21c618ea >>> >>> commit abaf29cb768375957c9ee0b64d36034c21c618ea >>> Author: Shilpa Singh <[email protected]> >>> Date: Mon Sep 21 23:48:16 2015 +0200 >>> >>> eina_tmpstr: add eina_tmpstr_strftime >> >> This API seems awfully arbitrary and I don't really see the point. >> Might as well be any other function that prints to strings - are you >> gonna add these too? Sounds horrible to me. > > Is it better to have our code cluttered by static buffer and no overflow > check ? Obviously not. Yes,I expect we refactor all our buffer print code > as it is pretty bad right now and may even lead to security issue in some > case.
Just to chime in, I think this doomsday scenario is a bit exaggerated. I think having code like: char buf[SOME_LEN]; strftime(...buf...); return tmpstr_add(buf); Is a much cleaner solution than adding all of the string functions of the world to tmpstr, strbuf, stringshare and etc. I really don't like code duplication, and I think that's where Daniel is coming from. What do you think? -- Tom. ------------------------------------------------------------------------------ _______________________________________________ enlightenment-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
