Jakub Narebski wrote:
quoted text
> Andreas Fuchs <asf@boinkor.net> wrote:
> 
>> * Wrap the commit message in <pre>
> We use <div class="pre"> in "commit" view if I remember correctly.

That's ok for rendered HTML output, but in my experience, the way feed
readers interpret that ranges from "badly" to "not at all"; it's better
to stick to explicit structure hints only in feeds. /-:

So, this is the only thing I haven't fixed in the attached patch (:

quoted text
>> * Make file names into an unordered list
> Good idea.
> 
>> * Add links (diff, conditional blame, history) to the file list.
> I'd rather keep RSS output as simple as possible, no frills.

I can see that, but it would be very useful on aggregation sites like
http://planet.sbcl.org/. You mentioned on IRC that you'd prefer to
forward-port the current RSS generation to a more modern feed format
like Atom 1.0.

I took the liberty to back out that change from the RSS generator, and
implement Atom 1.0 output that is more fully featured. For testing, I
left both in. Both feeds validate at feedvalidator.org for me, the
choice is yours (:

quoted text
> esc_html does to_utf8, so to_utf8 is unnecessary (and spurious).
> But it is a good catch: esc_html is certainly needed.

The attached patch doesn't use to_utf8 on already-escaped strings anymore.

quoted text
> We have introduced esc_path for escaping pathnames. Use it!

I changed that, too.

quoted text
> Two unnecessary calls to git command. Use 
>      my %difftree = parse_difftree_raw_line($line)
> instead. The conditions would probably be 
>      next if (!$difftree{'from_id'});
> (or equivalent).

Thanks for the hint; I included that in the Atom output and re-worked
the RSS generator to not use the hideous regexp anymore.

quoted text
> esc_url, not esc_html here. Or use the href() subroutine with -full=>1
> option (after applying the patch I send which added this to href()).

Okay; I changed all occurrences of esc_html where URLs are escaped to
use esc_url.

In addition to the above points, the attached patch emits a
Last-Changed: HTTP response header field, and doesn't compute the feed
body if the HTTP request type was HEAD. This helps keep the web server
load down for well-behaved feed readers that check if the feed needs
updating.

Hope you like it,
-- 
Andreas Fuchs, (http://|im:asf@|mailto:asf@)boinkor.net, antifuchs