>>>> "sb" == SL Baur <steve(a)xemacs.org>
writes:
sb> Karl M Hegbloom <karlheg(a)cathcart.sysc.pdx.edu> writes in
sb> xemacs-beta(a)xemacs.org:
> Shouldn't the configurables at the top of XEmacs.rules be
> pulled out and put into `Local.rules' as well?
sb> XEmacs.rules isn't pulled into the toplevel Makefile or the
sb> first level Makefiles. There is a path problem as
sb> XEmacs.rules assumes it will only be included two directory
sb> levels down from the top.
Alright... so I guess there needs to be a `toplevel' variable, and then
rather than doing:
include ../.../Local.rules
You could do:
include ${toplevel}/Local.rules
Or...
MAKEFLAGS += -I ${toplevel}
include Local.rules
`toplevel' ought to resolve into an absolute path, but be formed at
run-time. I'll try and look into doing this stuff soonish... I will
be starting on the xemacs-packages for Debian in a week or so (eta).
I'll do it then if ever.
sb> Yes, I agree that it should be possible to have a non-CVSed
sb> local file provide those defaults.
> I've made a few modifications to `XEmacs.rules' to make
it
> possible to run in place from a CVS checkout.
sb> It is only an accident that this works. Count on it being
sb> broken by future updates to the installed packaging structure.
Ok, so you plan to force me to learn more `make' tricks. :-)
Why don't I get what I've done squared away and looking pretty, then
send it in for yous to look over. Perhaps it can become part of
whatever changes might break whatever? I'll show it to yous and let
yous decide. (later today... I'm in a silly "teraterm" on an NT box
at the library right now... It's difficult to work from.)
Is it Ok to assume that GNU make will be used? Can we require that?
(I hope so. It's the only one I know at all.)