Michael Shapiro wrote: > > CC:'ing gatekeepers since it's AFAIK their decision to accept or > > reject this (and I am praying to whatever Discworld god currently > > cares about programmers (Bel-Shamharoth ?) that this proposal gets > > a "thumbs up" since parts of the ksh93-integration prototype004 > > Makefiles heavily depend on SHELL=/bin/ksh (which is thought as > > intermediate step to the use of SHELL=/bin/ksh93, e.g. this > > RFE)). > > > > My basic proposal is to allow the usage of either SHELL=/bin/sh > > or SHELL=/bin/ksh93 in OS/Net Makefiles as _choice_ for all OS/Net > > projects. > > I think the choice option is awful: having people have to remember that > in subdirectory A I can use this shell syntax and in subdirectory B
Erm, ksh93 is (mainly) backwards-compatible to the old bourne shell (see http://mail.opensolaris.org/pipermail/opensolaris-code/2007-March/004621.html) which means that existing Makefiles will continue to work without problems (unless you really craft something like a "Bourne shell vs. POSIX shell"-test into an OS/Net Makefile (which is AFAIK unlikely... :-) )). > I can't > just seems way too confusing and opens the door for not just two variants Why ? Exactly _two_ choices (e.g. { SHELL=/usr/bin/sh | SHELL=/usr/bin/ksh93 }) would be allowed by gatekeepers - AFAIK that doesn't open any doors for more variants, right (I guess gk will have their teeth deeply in the throat of anyone who would try to use SHELL=/bin/csh ... =:-) ) ? > but many more which I think is a slipperly slope to chaos. > > Switching to ksh93 (or bash, or ksh88, or whatever Erm, I hoped that we can reduce the usage of "ksh88" down to zero over time (if possible) which reduces the choices down to "ksh93", "bash" and "/usr/xpg4/bin/sh" (that are AFAIK the only (more or less) POSIX-like shells in Solaris). > ) would be eminently > sensible: just do the putback to change *everything* to one thing, Erm... I see two problems: 1. I am not a huge fan of forcing everyone - I'd like to provide the _choice_ between two shells and not forcibly "convert" everyone 2. We need to start somehow. First I'd like to "proove" that this works within OS/Net (well, it worked for me and everyone else who compiled the ksh93-integration prototypes) and _then_ think about further steps... > make it > the same, fix any extant broken-ness, and confirm that the world builds > and boots on SPARC and x86 and that you haven't regressed build time (and > judging by your existing data, tell us how much faster it goes) Currently ksh93 in OS/Net won't be much faster because many of the builtin commands have been disabled (that was done mainly to keep PSARC 2006/550 simple) which reduces the current performance benefits of swapping SHELL=/usr/bin/sh with SHELL=/usr/bin/ksh93 to a smaller subset (however this will improve a lot with further ARC cases when we re-enable more builtins and further resource reduction work during the upcoming ksh93 busybox project). Much more can be gained via using ksh93 features explicitly but that requires to crawl over the tree manually and touch some Makefiles. ---- Bye, Roland -- __ . . __ (o.\ \/ /.o) [EMAIL PROTECTED] \__\/\/__/ MPEG specialist, C&&JAVA&&Sun&&Unix programmer /O /==\ O\ TEL +49 641 7950090 (;O/ \/ \O;) _______________________________________________ opensolaris-code mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/opensolaris-code
