Merge branch 'master' of ssh://git.ikiwiki.info/srv/git/ikiwiki.info

Conflicts:
	doc/post-commit/discussion.mdwn

5 files changed, 74 insertions, 10 deletions

diff --git a/doc/bugs/html5_support.mdwn b/doc/bugs/html5_support.mdwn
index 783f5e47c..239474275 100644
--- a/doc/bugs/html5_support.mdwn
+++ b/doc/bugs/html5_support.mdwn
@@ -37,10 +37,6 @@ This element is poorly supported by browsers. As a workaround, `style.css` needs
 
 Internet Explorer will display it as a block, though you can't seem to be able to further control the style.
 
-# Validator complains about no h1-h6 in header
-
-* [#509](http://bugzilla.validator.nu/show_bug.cgi?id=509)
-
 ## Time element
 
 The [time element](http://www.whatwg.org/specs/web-apps/current-work/multipage/text-level-semantics.html#the-time-element) ideally needs the datatime= attribute set by a template variable with what [HTML5 defines as a valid datetime string](http://www.whatwg.org/specs/web-apps/current-work/multipage/infrastructure.html#valid-global-date-and-time-string).
diff --git a/doc/plugins/contrib/album.mdwn b/doc/plugins/contrib/album.mdwn
index f550ca64c..395c99bce 100644
--- a/doc/plugins/contrib/album.mdwn
+++ b/doc/plugins/contrib/album.mdwn
@@ -65,6 +65,13 @@ The next/previous links are themselves implemented by
 photo, using a special template (by default `albumnext.tmpl`
 or `albumprev.tmpl`), in `archive`/`quick` mode.
 
+> With hindsight, using an inline here is wrong - I should just
+> run hooks and fill in the template within the album plugin.
+> inline has some specialized functionality that's overkill
+> here, and its delayed HTML substitution breaks the ability
+> to have previous/up/next links both above and below the
+> photo, for instance. --[[smcv]]
+
 ## Writing the album
 
 The album contains one `\[[!album]]` directive. It may also
diff --git a/doc/plugins/contrib/cvs.mdwn b/doc/plugins/contrib/cvs.mdwn
new file mode 100644
index 000000000..1ff71d274
--- /dev/null
+++ b/doc/plugins/contrib/cvs.mdwn
@@ -0,0 +1,31 @@
+[[!template id=plugin name=cvs core=0 author="[[schmonz]]"]]
+
+This plugin allows ikiwiki to use [[!wikipedia desc="CVS" Concurrent Versions System]] as an [[rcs]].
+
+* Diffs are against [[3.14159|news/version_3.14159]]. `cvs.pm` started life as a copy of `svn.pm`.
+* `IkiWiki.pm:wiki_file_prune_regexps` avoids copying CVS metadata into `$DESTDIR`.
+* [[ikiwiki-makerepo]]:
+ * creates a repository,
+ * imports `$SRCDIR` into top-level module `ikiwiki` (vendor tag IKIWIKI, release tag PRE_CVS),
+ * creates a small post-commit wrapper to prevent `cvs add <directory>` from being seen by ikiwiki's [[post-commit]] hook,
+ * configures the wrapper itself as a post-commit hook in `CVSROOT/loginfo`.
+* [`cvsps`](http://www.cobite.com/cvsps/) is required (`rcs_recentchanges()` and `rcs_diff()` need it to work).
+* CVS multi-directory commits happen separately; the post-commit hook sees only the first directory's changes in time for [[recentchanges|plugins/recentchanges]]. The next run of `ikiwiki --setup` will correctly re-render such a recentchanges entry. It might be possible to solve this problem with scripts like `commit_prep` and `log_accum` from CVS contrib.
+* Due to the name of CVS's metadata directories, it's impossible to create `.../CVS/foo.mdwn`. On case-insensitive filesystems it's also impossible to create `.../cvs/foo.mdwn`.
+* No testing or special-casing has been done with [[attachments|plugins/attachment]], but they'll probably need `cvs add -kb`.
+
+Having a `$HOME/.cvsrc` isn't necessary. Sure does make using CVS more livable, though. Here's a good general-purpose one:
+
+    cvs -q
+    checkout -P
+    update -dP
+    diff -u
+    rdiff -u
+
+Not knowing how the tests get set up, I blindly attempted to add subversion-like tests to `t/file_pruned.t`. They fail. But the plugin definitely works. :-)
+
+### Code
+* [`cvs.pm`](http://www.netbsd.org/~schmonz/ikiwiki-cvs/cvs.pm)
+* [`cvs-IkiWiki.pm.diff`](http://www.netbsd.org/~schmonz/ikiwiki-cvs/cvs-IkiWiki.pm.diff)
+* [`cvs-ikiwiki-makerepo.diff`](http://www.netbsd.org/~schmonz/ikiwiki-cvs/cvs-ikiwiki-makerepo.diff)
+* [`cvs-t-file_pruned.t.diff`](http://www.netbsd.org/~schmonz/ikiwiki-cvs/cvs-t-file_pruned.t.diff)
diff --git a/doc/plugins/contrib/trail.mdwn b/doc/plugins/contrib/trail.mdwn
index 52dea52d6..eb233e095 100644
--- a/doc/plugins/contrib/trail.mdwn
+++ b/doc/plugins/contrib/trail.mdwn
@@ -13,6 +13,17 @@ This plugin can benefit from
 [[another_of_my_branches|todo/inline_plugin:_specifying_ordered_page_names]]
 but does not require it.
 
+Bugs:
+
+* \[[!inline pages="..." trail=yes]] currently tries to work out
+  what pages are in the trail, and their order, at scan time. That
+  won't work, because matching a pagespec at scan time is
+  unreliable - pages we want might not have been scanned yet! I
+  haven't worked out a solution for this. I think
+  \[[!inline pagenames="..." trail=yes]]
+  ([[patch_required|todo/inline_plugin:_specifying_ordered_page_names]])
+  would be safe, though.
+
 ----
 
 [[!template id=plugin name=trail author="[[Simon_McVittie|smcv]]"]]
diff --git a/doc/post-commit/discussion.mdwn b/doc/post-commit/discussion.mdwn
index 85853c945..6ae0d9bcb 100644
--- a/doc/post-commit/discussion.mdwn
+++ b/doc/post-commit/discussion.mdwn
@@ -73,12 +73,30 @@ proper order. (cvs add isn't recursive, sadly.)
 
 Can you offer an educated guess what's going wrong here? --[[Schmonz]]
 
-> Got `rcs_recentchanges` working, believe it or not, thanks to
-> [cvsps](http://www.cobite.com/cvsps/). If I can figure out this
-> interaction between the post-commit hook and `cvs add` on directories,
-> the CVS plugin is mostly done. Could it be a locking issue? Where should
-> I be looking? Any suggestions appreciated. --[[Schmonz]]
-
+> Got `rcs_recentchanges` working, believe it or not, thanks to [cvsps](http://www.cobite.com/cvsps/). If I can figure out this interaction between the post-commit hook and `cvs add` on directories, the CVS plugin is mostly done. Could it be a locking issue? Where should I be looking? Any suggestions appreciated. --[[Schmonz]]
+
+>> Okay, it is definitely a locking issue. First, on the conjecture that
+>> `cvs add <directory>` was triggering the post-commit hook and confusing
+>> ikiwiki, I wrapped the ikiwiki post-commit binary with a shell script
+>> that exited 0 if the triggering file was a directory. The first half of
+>> the conjecture was correct -- my wrapper got triggered -- but the web
+>> add of `one/two/three.mdwn` (where `one` and `two` weren't existing
+>> CVS-controlled dirs) remained hung as before. There were two ikiwiki
+>> processes running. On a whim, I killed the one with the higher PID; `cvs
+>> add one` immediately completed successfully, then back to a hang and two
+>> ikiwiki processes. I killed the newer one again and then `cvs add
+>> one/two` and `cvs add one/two/three.mdwn` completed and the web add was
+>> successful. --[[Schmonz]]
+
+>>> Aaaaaand I was wrong about the second half of the conjecture being
+>>> wrong. The wrapper script wasn't correctly identifying directories;
+>>> with that fixed, everything works. I've created a
+>>> [[plugins/contrib/cvs]] plugin page. Thanks for listening. :-)
+>>> --[[Schmonz]]
+
+>> Here is a comment I committed to my laptop from Madrid Airport before
+>> your most recent updates, in case it's still useful:
+>>
 >> Locking certianly seems likely to be a problem. ikiwiki calls `rcs_add`
 >> *before* disabling the post-commit plugin, since all over VCS allow
 >> adding something in a staged manner. You can see this in, for example,
@@ -97,3 +115,4 @@ Can you offer an educated guess what's going wrong here? --[[Schmonz]]
 >> especially since ikiwiki always commits after an add, in the same
 >> process, so you could just use a temporary list of things to add.
 >> --[[Joey]] 
+