Re: Naming issues, solutions

On Aug 25, 2014 1:14 PM, "Dmitry Stogov" <dmitry@zend.com> wrote: > > > > > On Mon, Aug 25, 2014 at 2:59 PM, Pierre Joye <pierre.php@gmail.com> wrote: >> >> >> On Aug 25, 2014 12:56 PM, "Dmitry Stogov" <dmitry@zend.com> wrote: >> > >> > >> > >> > >> > On Mon, Aug 25, 2014 at 2:45 PM, Pierre Joye <pierre.php@gmail.com> wrote: >> >> >> >> >> >> On Aug 25, 2014 9:22 AM, "Dmitry Stogov" <dmitry@zend.com> wrote: >> >> > >> >> > Hi Pierre, >> >> > >> >> > I'm glad, you agree to rename IS_INT back to IS_LONG. >> >> > >> >> > zend_long and size_t usage looks fine. >> >> > >> >> > I see no problems changing zend_string related API and related macros introduced in NG. They are new for everyone. >> >> > I hope, the changes won't make API less clear or useful (so it would be great to review them). >> >> > >> >> > I don't see a big reason to rename "zend_uint" into >> >> > "uint32_t", but it's just my own opinion. I would prefer keep it as is, or at worse case rename into "zend_uint32" or "uint32". Anyway, I'll agree with majority. >> >> > >> >> >> >> uint32_t :) >> > >> > 3 people are not the majority (or may be I missed discussion on IRC). >> > It's better to think before changing something in many places without a real reason. >> >> Why do we need it when the standard type exists and does exactly what we want? Same for size_t. It is mostly contained in the engine, no other impact and avoid yet another type definition for existing standard type (we have php_stdint.h to make them always available). > > We never had zend_size_t before and used size_t. > but we had zend_uint for ages and changing it to new name in thousand places won't make a lot of sense. > If you like to make it always be a 32-bit number, just change the definition. > If you are going to rename zend_uint, should we expect renaming of all the similar zend_* types from zend_types.h? > > typedef unsigned char zend_bool; > typedef unsigned char zend_uchar; > typedef unsigned int zend_uint; > typedef unsigned long zend_ulong; > typedef unsigned short zend_ushort; That would be a very good thing IMHO. These types exist because of the lack of stdint. Now that we have it it could be easier to use them when we can. Nikita?