Ar an triochadú lá de mí na Nollaig, scríobh Tim Landscheidt:
Fedora's package of XEmacs 20171230hg92757c2b8239
unsuccessfully tries to
set user-mail-address to a fixed string
"mockbuild(a)fedoraproject.org" to generate reproducible
builds of *.elc files (XEmacs puts that information there).
While looking into this, I noticed that:
| make EMACSDEBUGPATHS=yes XEMACS_BATCH='$(DO_XEMACS) $(BATCH) -eval "(setq
user-mail-address \"mockbuild(a)fedoraproject.org\")"'
works, but *.elc files compiled early on by temacs still
bear the default (in mock environments unpredictable)
user-mail-address.
So I tried:
| make EMACSDEBUGPATHS=yes BATCH='-no-packages -batch -eval "(setq
user-mail-address \"mockbuild(a)fedoraproject.org\")"'
but that fails with:
| ./temacs -nd -no-packages -batch -eval "(setq user-mail-address
\"mockbuild(a)fedoraproject.org\")" -l
/builddir/build/BUILD/xemacs-beta/src/../lisp/update-elc.el
| temacs can only be run in -batch mode.
What options are supported by temacs that could be used to
set user-mail-address?
What you’ve done is reasonable, as far as I can see it’s a bug in the temacs
code that it handles -eval badly there.
(My fallback solution is to directly patch
lisp/update-elc{,-2}.el, but I'd prefer to resolve the issue
at make level if possible.)
What’s your concern with patching Lisp? You worry that it’s less
maintainable?
P. S.: I tried to subscribe first to this mailing list via
https://lists.xemacs.org/, but the link in the "Con-
firmation needed" mail redirects to
https://lists.xemacs.org/users/address_activation/[40-hex-string]/
and then gives a "Server Error (500)", with the sub-
scription request lingering in "pending" state.
Partially sorted, sorry about the delay!
--
‘As I sat looking up at the Guinness ad, I could never figure out /
How your man stayed up on the surfboard after forty pints of stout’
(C. Moore)