Charles Lane wrote:
[warning: line breaks may be broken in this email response] > I don't know about the rest of you, but when I'm building Perl over > and over I get rather tired of having to do the whole CONFIGURE.COM > dialog manually. > > The "-des" option is a help, but only if you want a "standard" version, > and I almost always want non-standard, at least to the extent of > a VMS-debug enabled version. > > The following patch makes many of the CONFIGURE.COM options > selectable via command-line -D or -U options. For example, my favorite > choices would be: > @CONFIGURE "-Dusedevel" "-Dusevmsdebug" "-des" > > So that the "use..." options now available now include: > usevmsdebug > useperldebug > usemultiplicity > usethreads > useithreads > usecasesensitve > useieee > usesecurelog > usedefaulttypes > > in addition to the previously available: > usedevel > use64bitint > use64bitall > usemymalloc > useperlio These lists ought to go into README.vms since global symbols like: $ usevmsdebug == "define" will affect the build. (Sebastian Bazley will be upset otherwise). It would be necessary to SEARCH configure.com f$type to route most out (that will turn up some false positives though). I think that usedevel is supposed to remain undocumented though... not sure (didn't I see a p5p patch that added it to the main INSTALL file?). > The patch doesn't change any of the defaults. > > diff -uBb configure.com-orig configure.com > --- configure.com-orig Thu Oct 25 22:13:29 2001 > +++ configure.com Thu Oct 25 22:13:21 2001 > @@ -2181,6 +2181,7 @@ > $ echo "perl itself. This can be a useful feature if you are " > $ echo "embedding perl in a program." > $ dflt = "n" > +$ if f$type(usevmsdebug) .nes. "" .and. usevmsdebug .eqs. "define" then dflt="y" Great - I am glad to see use being made of these things in configure.com. [snip] > +$ if f$type(usemultiplicity) .nes. "" .and. usemultiplicity .eqs. "define" then >dflt = "y" > $ rp = "Build Perl for multiplicity? [''dflt'] " > $ GOSUB myread > $ IF ans.eqs."" then ans = dflt > -$ IF ans > +$ IF F$EXTRACT(0, 1, F$EDIT(ans,"COLLAPSE,UPCASE")) .eqs. "Y" > $ THEN > $ usemultiplicity="define" Why the change from C<IF ans> to C<IF F$EXTRACT(0, 1, F$EDIT(ans,"COLLAPSE,UPCASE")) .eqs. "Y"> ? I had in fact wanted to go through configure.com and change all, or at least most such to the C<IF ans> form. I thought that allowing folks to answer "0" or "1" might help them (e.g. if they cannot answer "OUI"|"NON" or "SI"|"NO" - at least these don't work on North American English DCL). It will take a bit of work but perhaps we ought to stick the C<$ IF ans.eqs."" then ans = dflt> statement into the GOSUB myread target. The equivalent myread shell script on Unix does have that set in place. A few of the answer handlers would need to be rewritten in configure.com in order to affect such a change though. Peter Prymmer
