The main issue I see with using embryo (or small, or Java) or any other byte-code/VM based machine is that it seems to make it much more difficult for the end-user to fix problems on their end. With the existing libtool they can hunt and peck through the scripts and implement local fixes, which may eventually become useful patches sent back to libtool. With a scheme which distributes "compiled" bytecode files, there is no possiblity of the end-user fixing the problem unless he installs the full libtool distribution package to he can modify the bytecode.
That would be no less true of a compiled ltmain.c.
Even with a compiled ltmain.c, I expect that the platform specific rules would be encapsulated in data files rather than being compiled into ltmain.c. That would allow end-users to adjust the rules. Ltmain would simply be a rules processor.
It is possible to also distribute the byte-code compiler so the end-user can modify source code, but the byte-code compiler is much larger and more complex than the byte-code execution processor so the size advantage is lost.
Bob ====================================== Bob Friesenhahn [EMAIL PROTECTED] http://www.simplesystems.org/users/bfriesen
_______________________________________________ Libtool mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/libtool
