Message-ID: <3977FBFD.EF6DA2D7@softhome.net> Date: Fri, 21 Jul 2000 09:30:05 +0200 From: Laurynas Biveinis X-Mailer: Mozilla 4.73 [en] (Win98; U) X-Accept-Language: lt,en MIME-Version: 1.0 To: djgpp-workers AT delorie DOT com Subject: Re: matters References: <39774EA5 DOT 7AA0AE0F AT softhome DOT net> <200007201919 DOT PAA31783 AT envy DOT delorie DOT com> <39775851 DOT A2449BF8 AT softhome DOT net> <200007201950 DOT PAA32032 AT envy DOT delorie DOT com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Reply-To: djgpp-workers AT delorie DOT com DJ Delorie wrote: > I.e. don't #error just because we *should* have gcc's headers. Only > #error if we *must* have gcc's headers. Exactly. It depends on if we are going to have more GCC headers. If we will have only *arg[s].h, then I won't put this in. In other case, if we rely on other GCC headers to do something important we should put an error, well not there - but in those crucial headers. On the second note, if we will use GCC's stdarg.h, IMHO there is no point mirroring all its definitions anyway. It's error prone and we will have to track it in future. What if stdarg.h from GCC changes again? Laurynas