aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--doc/patchqueue/pm_filter-priority.mdwn5
1 files changed, 5 insertions, 0 deletions
diff --git a/doc/patchqueue/pm_filter-priority.mdwn b/doc/patchqueue/pm_filter-priority.mdwn
index 16705e33e..f8436f0ae 100644
--- a/doc/patchqueue/pm_filter-priority.mdwn
+++ b/doc/patchqueue/pm_filter-priority.mdwn
@@ -2,6 +2,7 @@ Thanks to [r2677](http://ikiwiki.info/cgi-bin/viewvc.cgi/trunk/pm_filter?revisio
The main problem is that `pm_filter` runs after `./ikiwiki.in`, so the build still fails. If it runs before, we're fine.
+<pre>
===================================================================
--- Makefile.PL (revision 2682)
+++ Makefile.PL (working copy)
@@ -36,6 +37,8 @@ The main problem is that `pm_filter` runs after `./ikiwiki.in`, so the build sti
> your environment a no-op. Unless it sets `PERL5LIB=.:$PERL5LIB` or
> something. --[[Joey]]
+>> Heh, I guess it's obvious that I don't normally do packaging/build stuff -- probably best if I stop with these kind of patches. :-) OK, why not try the latter then? I think this is a worthy goal, because if we can make it easy to install on shared hosts, it will reach far more people, IMO.
+
(Is there any reason to create `ikiwiki.out`? Couldn't we just do an in-place replace on `ikiwiki.in`?)
> Yes, it's there because "ikwiki" == "Ikiwiki" on Mac OS X. Also because
@@ -59,3 +62,5 @@ The last line is necessary for the CGI, because, as with most shared hosts, the
> for you? Maybe that `LIB=~/lib` is defeating it? Not familiar with that
> variable.
> --[[Joey]]
+
+>> You're completely right; without the LIB argument the `Makefile.PL` heuristics work perfectly. I think I was juggling too many patches at once. ;-) \ No newline at end of file