Age | Commit message (Collapse) | Author |
|
* Support building on systems that lack asprintf.
* mercurial getctime is currently broken, apparently by some change in
mercurial version 0.9.4. Turn the failing test case into a TODO test case.
|
|
most web sites serve ikiwiki xhtml files as text/html and mozilla browsers
get confused. So it's best for ikiwiki to follow the compatability
recommendations in appendix C of the XHTML spec. Closes: #432045
|
|
This turns out to have occured if the cgi wrapper was created by an
ikiwiki invocation that included --rebuild. Thanks to Carl Worth for
tracking that down.
|
|
once more.
|
|
w/o confusing ikiwiki.
* Rcs modules are now imported, so other modules that need similar
normalisation of configs etc can do so.
|
|
|
|
|
|
|
|
|
|
newlines.
|
|
|
|
ESCAPE=HTML for titles in the templates for these feeds, and instead
escape the title going in to the template. Previously, the title was
sometimes double-escaped in a feed (if set via meta title), and sometimes
not (if set from the page filename).
* In the meta plugin, when a title is set, encode the html entities in it
numerically. This works better in the current landscape of a rss spec that
doesn't specify encoding, and variously broken feed consumers, according
to <http://www.rssboard.org/rss-profile#data-types-characterdata>.
|
|
|
|
through the img plugin. Also allow the id attribute to be passed through.
|
|
old files.
* Change where the img plugin puts scaled images. It's better to make the
scaled images subpages of the page that embeds them, rather than putting
them alongside the original image, since if two pages scale the same image
the same way, this prevents complications in dealing with two pages
creating the same file. The move will be handled transparently, though you
might want to rebuild your wiki to make it occur in one step.
|
|
|
|
always be tracked correctly.
|
|
|
|
change after the lock is dropped, and before the lock is regained by the
build process. The state will now be reloaded by the build process.
|
|
until the wiki is building and already locked, unless it's aggregating.
When aggregating, it does not wait for the lock if it cannot get it, and
instead exits, to prevent aggregating processes from piling up.
|
|
* Fix links to smilies generated by the smiley plugin for inlined pages. The
old links were often wrong, but often still worked by accident.
|
|
- If the password is empty in preferences, don't clear the existing
password.
- Actually check the confirm password field, even if it's left empty.
|
|
|
|
|
|
change to the backlinks() function.
|
|
and style sheet updates, and unless you're using customised versions,
you'll want to rebuild wikis on upgrade to this version to avoid
inconsistencies.
* Allow WIKINAME to to used in footers, as an example of something to put
there.
|
|
set in the wiki setup, passwordauth will require the password in
order to create an account.
|
|
ikiwiki builds when refreshing the wiki.
|
|
not need to examine every page in the wiki.
* Use this to simplify the goodstuff detection in the plugin template.
|
|
|
|
|
|
parameters, and detecting if a pagespec failed to compile.
|
|
in.
|
|
|
|
|
|
|
|
styled using the stylesheet, rather than by creating signin and prefs
templates.
* Make the openid login form nicely styled.
|
|
|
|
disable itself, this avoids needing a hard dependency on the perl module.
|
|
re-evaluate after 2.0.
|
|
"login with openid" as the input field label
|
|
|
|
passwordauth page to the basewiki describing password
authentication; like openid, it uses conditional to check which
forms of authentication the wiki allows. Add conditional cross-
links between the openid and passwordauth pages, to help the user
understand how they can log in.
|
|
|
|
|
|
form.
|
|
Especially cleans up the ordering of the admin's preferences form.
|
|
using the mercurial backend. Not 100% sure why it failed w/o the full
path, but this still passes the test suite, and indeed, is how the test
suite calls hg add.
|
|
(Get a good message when a PageSpec fails due to a negated success by
creating success objects with a reason string, which morph into failure
objects when negated.)
|
|
scalar context, evaluates to a reason why the match failed.
* Add testpagespec plugin, which might be useful to see why a pagespec isn't
matching something.
|