Description of problem: Does rpm-spec-mode work for you? I'm using current emacs devel spec file, but with emacs cvs source. Using \c-c \c-b a (build all) gives nasty error. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Some commands including 'C-c C-b i' and 'C-c C-b p' works fine by applying patch of bugzilla #227418. But 'C-e C-e' does not work with following message. Invalid function: ignore-errors
That is odd since, AFAICT, ignore-errors is autoloaded from cl-macs. (And, if not, there's a compatibility defmacro in rpm-spec-mode.el.) Maybe try (require 'cl-macs).
Created attachment 191641 [details] rpm-spec-mode as a real compilation mode I've been working on changing rpm-spec-mode to use a real compilation mode. I think this is a better way to fix the problem than to patch up the existing, rather kludgey version. Compare, for example, grep.el. Chip
I just noticed this too after upgrading the latest emacs package for F7. :-(
Re-bytecompiling rpm-spec-mode.el by hand seems to fix the problem. Not sure why it is broken or if rpm-spec-mode.el should require cl-macs at compile time.
Ah I see there is bug 306841 for ignore-errors.
(In reply to comment #3) > Created an attachment (id=191641) [edit] > rpm-spec-mode as a real compilation mode Great! Note: I had to add "(require 'compile)" at the beginning of the file. Without it, `define-compilation-mode was not known by Emacs. I am look forward to see that file in Fedora...
*** This bug has been marked as a duplicate of 227418 ***
Created attachment 271411 [details] rpm-spec-mode using a real compilation mode Here's a first cut at an rpm-spec-mode that uses a real compilation mode. Give it a try and report any bugs. One thing that I know work is signing rpms automagically after build. Chip
Tested once.(In reply to comment #9) > Created an attachment (id=271411) [edit] > rpm-spec-mode using a real compilation mode > > Here's a first cut at an rpm-spec-mode that uses a real compilation mode. Give > it a try and report any bugs. Tested once. Works fine. I have added it to my load path, and will test regularly. RPM signing does *not* work.
(In reply to comment #10) > > RPM signing does *not* work. Yes, that was what I meant to say in my previous, mangled comment. Once we've worked out all the bugs and have all the functionality back, I want to push this change upstream. It seems logical to me that rpm building should be done in a compilation buffer. Chip