]> git.vanrenterghem.biz Git - git.ikiwiki.info.git/log
git.ikiwiki.info.git
14 years agocomment
Joey Hess [Tue, 6 Apr 2010 18:05:00 +0000 (14:05 -0400)]
comment

14 years agoidea
Joey Hess [Tue, 6 Apr 2010 17:58:55 +0000 (13:58 -0400)]
idea

14 years agofirst question
Joey Hess [Tue, 6 Apr 2010 17:25:26 +0000 (13:25 -0400)]
first question

14 years agooops, branch returned now
http://smcv.pseudorandom.co.uk/ [Tue, 6 Apr 2010 10:47:28 +0000 (10:47 +0000)]
oops, branch returned now

14 years agomore flexible underlays, please?
http://kerravonsen.dreamwidth.org/ [Tue, 6 Apr 2010 04:41:55 +0000 (04:41 +0000)]
more flexible underlays, please?

14 years agoresponse
http://kerravonsen.dreamwidth.org/ [Tue, 6 Apr 2010 04:00:47 +0000 (04:00 +0000)]
response

14 years ago(no commit message)
http://kerravonsen.dreamwidth.org/ [Tue, 6 Apr 2010 03:59:09 +0000 (03:59 +0000)]

14 years agoif applied, reverting this would be problematic
http://smcv.pseudorandom.co.uk/ [Tue, 6 Apr 2010 00:55:54 +0000 (00:55 +0000)]
if applied, reverting this would be problematic

14 years agotag as patch
http://smcv.pseudorandom.co.uk/ [Tue, 6 Apr 2010 00:51:27 +0000 (00:51 +0000)]
tag as patch

14 years agopages talking about escaping should really be escaped correctly
http://smcv.pseudorandom.co.uk/ [Tue, 6 Apr 2010 00:50:51 +0000 (00:50 +0000)]
pages talking about escaping should really be escaped correctly

14 years agoas seen on IRC
http://smcv.pseudorandom.co.uk/ [Tue, 6 Apr 2010 00:49:38 +0000 (00:49 +0000)]
as seen on IRC

14 years agoHTML-encode meta title, description, guid on output, but not in the pagestate
Simon McVittie [Tue, 6 Apr 2010 00:31:38 +0000 (01:31 +0100)]
HTML-encode meta title, description, guid on output, but not in the pagestate

This makes them consistent with the rest of the meta keys. A wiki rebuild
will be needed on upgrade to this version; until the wiki is rebuilt,
double-escaping will occur in the titles of pages that have not changed.

14 years agoswitch branch for review to use ready/foo convention
http://smcv.pseudorandom.co.uk/ [Tue, 6 Apr 2010 00:04:47 +0000 (00:04 +0000)]
switch branch for review to use ready/foo convention

14 years agoready for review, I think
http://smcv.pseudorandom.co.uk/ [Tue, 6 Apr 2010 00:02:18 +0000 (00:02 +0000)]
ready for review, I think

14 years agoOnly store authorsort in meta pagestate if it differs from author
Simon McVittie [Mon, 5 Apr 2010 23:59:16 +0000 (00:59 +0100)]
Only store authorsort in meta pagestate if it differs from author

14 years agoOnly store titlesort in meta pagestate if it differs from title
Simon McVittie [Mon, 5 Apr 2010 23:58:55 +0000 (00:58 +0100)]
Only store titlesort in meta pagestate if it differs from title

14 years agoUse $a and $b for SortSpec cmp callbacks
Simon McVittie [Mon, 5 Apr 2010 21:50:51 +0000 (22:50 +0100)]
Use $a and $b for SortSpec cmp callbacks

14 years agofix bug that left stray </p> tags
Joey Hess [Mon, 5 Apr 2010 21:18:30 +0000 (17:18 -0400)]
fix bug that left stray </p> tags

Both markdown and tidy add paragraph tags around text, that needs to be
stripped when the text is a short, one line fragment that is being inserted
into a larger page. tidy also adds several newlines to the end, and this
broke removal of the paragraph tags.

14 years agotxt: Add a special case for robots.txt.
Joey Hess [Mon, 5 Apr 2010 21:02:10 +0000 (17:02 -0400)]
txt: Add a special case for robots.txt.

14 years agocomments: Fix missing entity encoding in title.
Joey Hess [Mon, 5 Apr 2010 20:34:49 +0000 (16:34 -0400)]
comments: Fix missing entity encoding in title.

The meta title data set by comments needs to be encoded the same way that
meta encodes it. (NB The security implications of the missing encoding
are small.)

Note that meta's encoding of title, description, and guid data, and not
other data, is probably a special case that should be removed. Instead,
these values should be encoded when used. I have avoided doing so here
because that would mean forcing a wiki rebuild on upgrade to have the data
consitently encoded.

14 years agoresponse
Joey Hess [Mon, 5 Apr 2010 19:28:54 +0000 (15:28 -0400)]
response

14 years agopotential performance improvements
http://smcv.pseudorandom.co.uk/ [Mon, 5 Apr 2010 19:19:00 +0000 (19:19 +0000)]
potential performance improvements

14 years agospeed
Joey Hess [Mon, 5 Apr 2010 18:59:29 +0000 (14:59 -0400)]
speed

14 years agofix test suite
Joey Hess [Sun, 4 Apr 2010 21:06:28 +0000 (17:06 -0400)]
fix test suite

14 years agoclose bug, comment
Joey Hess [Sun, 4 Apr 2010 16:28:12 +0000 (12:28 -0400)]
close bug, comment

14 years agomore idiomatic perl
Joey Hess [Sun, 4 Apr 2010 16:23:12 +0000 (12:23 -0400)]
more idiomatic perl

14 years agorename hash and avoid unnecessary hash ref
Joey Hess [Sun, 4 Apr 2010 16:20:26 +0000 (12:20 -0400)]
rename hash and avoid unnecessary hash ref

Variable renamed to be a bit more self-explanatory.
Probably more idiomatic perl to not use a hash ref when a hash can be used.

14 years agochangelog
Joey Hess [Sun, 4 Apr 2010 16:17:36 +0000 (12:17 -0400)]
changelog

14 years agofix indentation & whitespace
Joey Hess [Sun, 4 Apr 2010 16:17:03 +0000 (12:17 -0400)]
fix indentation & whitespace

14 years agoFix bzr plugin to work with bzr 2.0.
Lars Wirzenius [Sun, 4 Apr 2010 09:05:07 +0000 (21:05 +1200)]
Fix bzr plugin to work with bzr 2.0.

The output of "bzr log" seems to have changed a bit, so we change the
parsing accordingly. This has not been tested with earlier versions of
bzr.

Several problems seemed to occur, all in the bzr_log subroutine:

1. The @infos list would contain an empty hash, which would confuse the
   rest of the program.
2. This was because bzr_log would push an empty anonymous hash to the
   list whenever it thought a new record would start.
3. However, a new record marker (now?) also happens at th end of bzr log
   output.
4. Now we collect the record to a hash that gets pushed to the list only
   if it is not empty.
5. Also, sometimes bzr log outputs "revno: 1234 [merge]", so we catch only
   the revision number.
6. Finally, there may be non-headers at the of the output, so we ignore
   those.

14 years agoadvertise some bugfixes
http://smcv.pseudorandom.co.uk/ [Sun, 4 Apr 2010 14:05:42 +0000 (14:05 +0000)]
advertise some bugfixes

14 years agobug report + patch: unnecessary YAML::Any dependency
http://smcv.pseudorandom.co.uk/ [Sun, 4 Apr 2010 14:03:51 +0000 (14:03 +0000)]
bug report + patch: unnecessary YAML::Any dependency

14 years ago(no commit message)
http://hendry.iki.fi/ [Sun, 4 Apr 2010 13:00:52 +0000 (13:00 +0000)]

14 years agoTypo.
PaulePanter [Sun, 4 Apr 2010 12:03:41 +0000 (12:03 +0000)]
Typo.

14 years ago(no commit message)
http://liw.fi/ [Sun, 4 Apr 2010 09:23:05 +0000 (09:23 +0000)]

14 years ago(no commit message)
http://liw.fi/ [Sun, 4 Apr 2010 09:13:01 +0000 (09:13 +0000)]

14 years ago(no commit message)
http://liw.fi/ [Sun, 4 Apr 2010 09:12:23 +0000 (09:12 +0000)]

14 years agoAdd the Banu website
https://mukund.startssl.com/ [Sun, 4 Apr 2010 07:16:41 +0000 (07:16 +0000)]
Add the Banu website

14 years agoRemove my homepage
https://mukund.startssl.com/ [Sun, 4 Apr 2010 07:13:51 +0000 (07:13 +0000)]
Remove my homepage

14 years agoRename sort parameter to meta title/author to sortas=bar
Simon McVittie [Sat, 3 Apr 2010 13:19:18 +0000 (14:19 +0100)]
Rename sort parameter to meta title/author to sortas=bar

Joey pointed out that sort=x usually takes a sort order.

14 years agometa: generalize meta_title into meta(title); support author, updated, date
Simon McVittie [Sat, 3 Apr 2010 13:17:48 +0000 (14:17 +0100)]
meta: generalize meta_title into meta(title); support author, updated, date

I've left meta_title in, undocumented, as a possible replacement for
sort=title in IkiWiki 4.0 or something.

14 years agoupdated branch
http://smcv.pseudorandom.co.uk/ [Sun, 4 Apr 2010 00:34:36 +0000 (00:34 +0000)]
updated branch

14 years agoupdate fieldsort plugin to be compatible with the latest version of my branch
http://smcv.pseudorandom.co.uk/ [Sun, 4 Apr 2010 00:27:20 +0000 (00:27 +0000)]
update fieldsort plugin to be compatible with the latest version of my branch

14 years agoAdd a unit test for changed-link calculation
Simon McVittie [Sun, 4 Apr 2010 00:20:02 +0000 (01:20 +0100)]
Add a unit test for changed-link calculation

14 years agoRemove the typedlink(tag foo) pagespec feature, which is less friendly than tagged()
Simon McVittie [Sat, 3 Apr 2010 23:26:03 +0000 (00:26 +0100)]
Remove the typedlink(tag foo) pagespec feature, which is less friendly than tagged()

Plugins that introduce a link type should also introduce pagespec syntax
for it.

14 years agoRemove tagged_is_strict option, and just behave as though it was enabled
Simon McVittie [Sat, 3 Apr 2010 23:24:27 +0000 (00:24 +0100)]
Remove tagged_is_strict option, and just behave as though it was enabled

Jon was worried about the backwards-compatibility break involved in making
tagged() not match non-tag links, but Joey seems less concerned about it.

14 years agoFix some typos that would break Render during changed-link calculation
Simon McVittie [Sat, 3 Apr 2010 23:21:01 +0000 (00:21 +0100)]
Fix some typos that would break Render during changed-link calculation

14 years agoimplement typed links; add tagged_is_strict config option
Simon McVittie [Thu, 1 Apr 2010 23:28:02 +0000 (00:28 +0100)]
implement typed links; add tagged_is_strict config option

14 years agoanswer about autofiles for tags
David Riebenbauer [Sat, 3 Apr 2010 21:38:05 +0000 (23:38 +0200)]
answer about autofiles for tags

14 years agolink to commits
David Riebenbauer [Fri, 5 Feb 2010 11:54:07 +0000 (12:54 +0100)]
link to commits

14 years agoMake sure deleted tag pages don't get recreated.
David Riebenbauer [Sat, 3 Apr 2010 19:10:16 +0000 (21:10 +0200)]
Make sure deleted tag pages don't get recreated.

The reason to do this is basically a user interaction design decision.

It is achieved by adding an entry, associated to the creating plugin, to
%pagestate. To find out if files were deleted a new global hash %del_hash is
%introduced.

14 years agoRevert "Revert the effects of find_del_files() for (re)autoadded files."
David Riebenbauer [Sat, 3 Apr 2010 19:17:20 +0000 (21:17 +0200)]
Revert "Revert the effects of find_del_files() for (re)autoadded files."

This reverts commit 31680111f0062f07727d14fcf291c98978ad5a2f.

14 years agoCheck for existence off srcfile in add_autofile
David Riebenbauer [Wed, 3 Feb 2010 03:29:10 +0000 (04:29 +0100)]
Check for existence off srcfile in add_autofile

add_autofile has to have checks, whether to create the file, anyway, so this
will make things more consistent.

Correcter check for the result of verify_src_file().

Cosmetic rename of a variable $addfile to $autofile.

14 years agoslightly more consistent capitalisation of ikiwiki
Joey Hess [Sat, 3 Apr 2010 20:07:32 +0000 (16:07 -0400)]
slightly more consistent capitalisation of ikiwiki

For a while, I was avoiding capitalizing ikiwiki at the beginning of a
sentence. I now think that's a bad idea (unless explicitly referring to
the `ikiwiki` command). Still, I don't go all the way and always cap it,
as a proper noun. That would make the logo look bad. ;)

I also tend to avoid capping it as IkiWiki, except when referring to the
perl internals, which do use that capitalization. (Too late to change
that.) However, it's also reasonable to do so in a WikiLink, as a nod to
historical camelcase wikis.

14 years agoadd news item for ikiwiki 3.20100403
Joey Hess [Sat, 3 Apr 2010 19:15:25 +0000 (15:15 -0400)]
add news item for ikiwiki 3.20100403

14 years agoreleasing version 3.20100403 3.20100403
Joey Hess [Sat, 3 Apr 2010 19:15:04 +0000 (15:15 -0400)]
releasing version 3.20100403

14 years agofix buggy program installation code
Joey Hess [Sat, 3 Apr 2010 19:09:04 +0000 (15:09 -0400)]
fix buggy program installation code

14 years agoupdated branch
http://smcv.pseudorandom.co.uk/ [Sat, 3 Apr 2010 13:39:44 +0000 (13:39 +0000)]
updated branch

14 years agoMove sort hooks to the IkiWiki::SortSpec namespace
Simon McVittie [Sat, 3 Apr 2010 12:57:38 +0000 (13:57 +0100)]
Move sort hooks to the IkiWiki::SortSpec namespace

Also rename cmpspec_translate (internal function) to sortspec_translate
for consistency.

14 years agoRemove support for check_cmp_foo (pre-sort checks)
Simon McVittie [Sat, 3 Apr 2010 12:49:20 +0000 (13:49 +0100)]
Remove support for check_cmp_foo (pre-sort checks)

14 years agoSplit out sortnaturally into a plugin
Simon McVittie [Sat, 3 Apr 2010 12:48:30 +0000 (13:48 +0100)]
Split out sortnaturally into a plugin

14 years agoresponse
Joey Hess [Sat, 3 Apr 2010 02:46:31 +0000 (22:46 -0400)]
response

14 years agoresponse
Joey Hess [Sat, 3 Apr 2010 02:23:36 +0000 (22:23 -0400)]
response

14 years agocross-reference
http://smcv.pseudorandom.co.uk/ [Sat, 3 Apr 2010 01:15:28 +0000 (01:15 +0000)]
cross-reference

14 years agovague musings about wikilinks
http://smcv.pseudorandom.co.uk/ [Sat, 3 Apr 2010 01:13:08 +0000 (01:13 +0000)]
vague musings about wikilinks

14 years agoperhaps the typedlink(tag foo) pagespec isn't so useful
http://smcv.pseudorandom.co.uk/ [Sat, 3 Apr 2010 00:53:59 +0000 (00:53 +0000)]
perhaps the typedlink(tag foo) pagespec isn't so useful

14 years agoreduced difference
http://smcv.pseudorandom.co.uk/ [Sat, 3 Apr 2010 00:45:16 +0000 (00:45 +0000)]
reduced difference

14 years agosort-order could usefully be overridden for meta author, too
http://smcv.pseudorandom.co.uk/ [Sat, 3 Apr 2010 00:41:31 +0000 (00:41 +0000)]
sort-order could usefully be overridden for meta author, too

14 years agoactually I can see a second use for "nonlinear" syntax - but I don't think it's worth it
http://smcv.pseudorandom.co.uk/ [Sat, 3 Apr 2010 00:25:28 +0000 (00:25 +0000)]
actually I can see a second use for "nonlinear" syntax - but I don't think it's worth it

14 years agomake questions to Joey more explicit
http://smcv.pseudorandom.co.uk/ [Sat, 3 Apr 2010 00:12:59 +0000 (00:12 +0000)]
make questions to Joey more explicit

14 years agorespond at some length
http://smcv.pseudorandom.co.uk/ [Sat, 3 Apr 2010 00:00:53 +0000 (00:00 +0000)]
respond at some length

14 years agorespond (also unindent Joey's review to avoid very deep indentation)
http://smcv.pseudorandom.co.uk/ [Fri, 2 Apr 2010 22:25:16 +0000 (22:25 +0000)]
respond (also unindent Joey's review to avoid very deep indentation)

14 years agoidea
Joey Hess [Fri, 2 Apr 2010 21:37:38 +0000 (17:37 -0400)]
idea

14 years agocomments
Joey Hess [Fri, 2 Apr 2010 21:26:32 +0000 (17:26 -0400)]
comments

14 years agominor comment
Joey Hess [Fri, 2 Apr 2010 21:16:12 +0000 (17:16 -0400)]
minor comment

14 years agofix branch name
Joey Hess [Fri, 2 Apr 2010 21:03:33 +0000 (17:03 -0400)]
fix branch name

14 years agoqualify template_file
Joey Hess [Fri, 2 Apr 2010 20:55:29 +0000 (16:55 -0400)]
qualify template_file

14 years agotemplate: Search for templates in the templatedir, if they are not found as pages...
Joey Hess [Fri, 2 Apr 2010 20:54:06 +0000 (16:54 -0400)]
template: Search for templates in the templatedir, if they are not found as pages in the wiki.

14 years agoreview from the woods
Joey Hess [Fri, 2 Apr 2010 20:41:04 +0000 (16:41 -0400)]
review from the woods

14 years agohtmlscrubber: Allow colons in url fragments after '?'
Joey Hess [Fri, 2 Apr 2010 20:05:14 +0000 (16:05 -0400)]
htmlscrubber: Allow colons in url fragments after '?'

Colons are not allowed at the start of urls, because it can be interpreted
as a protocol, and allowing arbitrary protocols can be unsafe
(CVE-2008-0809). However, this check was too restrictive, not allowing
use of eg, "video.ogv?t=0:03:00/0:04:00" to seek to a given place in a
video, or "somecgi?foo=bar:baz" to pass parameters with colons.

It's still not allowed to have a filename with a colon in it (ie
"foo:bar.png") -- to link to such a file, a fully qualified url must be
used.

14 years agoDespam, again. Someone ban adrianna please?
Simon McVittie [Fri, 2 Apr 2010 17:28:34 +0000 (18:28 +0100)]
Despam, again. Someone ban adrianna please?

14 years ago(no commit message)
adrianna [Fri, 2 Apr 2010 12:29:22 +0000 (12:29 +0000)]

14 years ago(no commit message)
adrianna [Fri, 2 Apr 2010 12:26:46 +0000 (12:26 +0000)]

14 years ago(no commit message)
adrianna [Fri, 2 Apr 2010 12:24:14 +0000 (12:24 +0000)]

14 years ago(no commit message)
adrianna [Fri, 2 Apr 2010 12:18:36 +0000 (12:18 +0000)]

14 years ago(no commit message)
adrianna [Fri, 2 Apr 2010 12:07:42 +0000 (12:07 +0000)]

14 years agoDespam
Simon McVittie [Fri, 2 Apr 2010 12:07:22 +0000 (13:07 +0100)]
Despam

14 years ago(no commit message)
adrianna [Fri, 2 Apr 2010 12:05:18 +0000 (12:05 +0000)]

14 years ago(no commit message)
adrianna [Fri, 2 Apr 2010 12:03:58 +0000 (12:03 +0000)]

14 years ago(no commit message)
adrianna [Fri, 2 Apr 2010 12:01:24 +0000 (12:01 +0000)]

14 years ago(no commit message)
adrianna [Fri, 2 Apr 2010 12:00:46 +0000 (12:00 +0000)]

14 years agoimplemented!
http://smcv.pseudorandom.co.uk/ [Thu, 1 Apr 2010 23:40:37 +0000 (23:40 +0000)]
implemented!

14 years agoExplain tab-delimited tables a bit better
Simon McVittie [Thu, 1 Apr 2010 17:13:38 +0000 (18:13 +0100)]
Explain tab-delimited tables a bit better

14 years agoconsidering tags and fields
http://kerravonsen.dreamwidth.org/ [Thu, 1 Apr 2010 07:22:37 +0000 (07:22 +0000)]
considering tags and fields

14 years ago(no commit message)
http://smcv.pseudorandom.co.uk/ [Thu, 1 Apr 2010 01:35:35 +0000 (01:35 +0000)]

14 years agoresponse; thoughts about inline and tags
http://kerravonsen.dreamwidth.org/ [Thu, 1 Apr 2010 00:23:31 +0000 (00:23 +0000)]
response; thoughts about inline and tags

14 years agopage.tmpl: Add Cache-Control must-revalidate to ensure that users (especially of...
Joey Hess [Wed, 31 Mar 2010 21:51:46 +0000 (17:51 -0400)]
page.tmpl: Add Cache-Control must-revalidate to ensure that users (especially of Firefox) see fresh page content.

Since Firefox version 3, it's done aggressive caching of visited pages, and
does not, by default, check if the cached content is still valid when
reloading or revisiting a page. By default, Firefox seems to not re-contact
the web server at all. Compare with eg, Epiphany and Chromium, which appear
to always check, and get back a 304 when the page is unchanged.

This header makes Firefox do the right thing, at least for html files. It
still over-caches if css, javascript, images, etc, are changed.

14 years agoresponse
http://kerravonsen.dreamwidth.org/ [Wed, 31 Mar 2010 03:01:41 +0000 (03:01 +0000)]
response

14 years agothey do work, with the right syntax (C backslash escapes aren't interpreted)
http://smcv.pseudorandom.co.uk/ [Wed, 31 Mar 2010 02:07:52 +0000 (02:07 +0000)]
they do work, with the right syntax (C backslash escapes aren't interpreted)

14 years ago(no commit message)
http://edrex.myopenid.com/ [Wed, 31 Mar 2010 00:24:22 +0000 (00:24 +0000)]