Date: Mon, 15 May 2000 14:34:01 +0300 (IDT) From: Eli Zaretskii X-Sender: eliz AT is To: djgpp AT delorie DOT com Subject: Re: OT: RHIDE and G77 (FORTRAN 77) In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Reply-To: djgpp AT delorie DOT com Errors-To: nobody AT delorie DOT com X-Mailing-List: djgpp AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk On Mon, 15 May 2000, Andris Pavenis wrote: > > The same problem was fixed in previous versions of GCC and RHIDE by > > defining RHIDE_TYPED_LIBS_DJGPP.f. That's why this variable was added > > to DJGPP.ENV. If the reason for these new problems is that RHIDE was > > changed, I think this is an unfortunate change. I wish people would > > refrain from making such changes, to avoid breaking something that > > took a long time to make work. > > > > > To summarize, the correct fix is to put the line > > > RHIDE_TYPED_LIBS.f=g2c m > > > in either RHIDE.ENV, or the [rhide] section of DJGPP.ENV. > > > > I would argue that, rather than change DJGPP.ENV, RHIDE should be > > changed to use RHIDE_TYPED_LIBS_DJGPP.f, as it did previously. > > > > Well. Looked through source archives I had on HD and didn't find > RHIDE_TYPED_LIBS_DJGPP.f anywhere (only RHIDE_TYPED_LIBS.f). I picked that name from your message. See the message "Re: Rhide & g77-2.81" posted by you to this news group on August 19, 1998. I didn't realize that the general name RHIDE_TYPED_LIBS does not imply the DJGPP-specific name. When DJGPP v2.03 was in the works, I suggested to add RHIDE_TYPED_LIBS_DJGPP.f to DJGPP.ENV (see my messages to djgpp-workers around August 25, 1999 "Re: RHIDE section in DJGPP.ENV"). Neither you nor Robert Hoehne objected. Needless to say, since I don't use RHIDE, I have to rely on yours and Robert's judgement. > Also I don't see any special reason why to introduce such as there are > no related file naming conflicts even for plain DOS: Perhaps the fact that this variable is in DJGPP.ENV is a reason good enough to add it. In fact, it would make sense that the two variables RHIDE_TYPED_LIBS.x and RHIDE_TYPED_LIBS_DJGPP.x would both resolve to the same string when the OS is DJGPP, if the second variable was not explicitly redefined.