Re: Any objections against xemacs/*/xemacs.mak files in 21.2?

APA> Hello Reviewers, APA> while trying to improve makefile[1] support for Windows NT native I am APA> faced with this problem: APA> ALL make rules for building XEmacs Windows NT native[2] are in APA> nt\xemacs.mak. APA> However, ＠setfilename in man\*\*.texi are obviously relative to man\*, APA> not nt. man/Makefile uses makeinfo with the -o option, which explicitly specifies the output filename. So the ＠setfilenames are ignored. There seems to be no way to explicitly specify an output filename with texinfmt's batch-texinfo-format, and so the ＠setfilename must be used. But maybe that's a bug in texinfo. An alternative is not to create files in a new directory at all. Generate the .info files in the current directory, and then use mv to move them to the target directory. Besides, why do we even bother with a separate info directory? We could just create all the .info files in man/ (where they can be used), and copied to some info directory at `make install' time. We could also take a look at what GNU Emacs 20.4 does. APA> Having nt\xemacs.mak to call out to man\xemacs.mak to handle the info APA> target (just like on Unix) would be must superior. There is no danger APA> in clashing with Unix makefiles since they are always called Makefile. or GNUmakefile. APA> I would be willing to send in patches for 21.2 to make xemacs.mak APA> build recursively if any objections/problems can be resolved. APA> Any objections? It seems reasonable. Check the FSF implementation first. If we can be compatible with them easily, that would also be a win. Martin