summaryrefslogtreecommitdiff
path: root/doc/bugs/recentchanges_feed_links.mdwn
blob: ef0f9d1c427933e5a78e60ec257fe5b5aadadb07 (plain)

(Moved from [[plugins/recentchanges/discussion]])

I've just upgraded to ikiwiki 2.50 with the recentchanges plugin enabled, and figured out that I have to turn on rss in ikiwiki.setup in order to get the recentchanges feed. Now the feed shows up, but the links in the feed go to the change pages, e.g. recentchanges/change_1700.html. I can see a recentchanges directory created in the working copy, containing files like change_1700._change but for some reason they are not getting htmlized and carried over. I can see in recentchanges.pm that it explicitly registers an htmlize hook for the _change type, but something isn't happening. I also see return if $type=~/^_/; in render() in Render.pm so I guess the upshot is I'm not sure how this is supposed to work; is there a bug here or just something I overlooked that I need to turn on? --Chapman Flack

It's a (minor) bug that recentchanges optimises away generating the change pages, but that the rss/atom feed still links to them. --[[Joey]]

Hmm, ok, what's the intended correct behavior? To really generate the change pages, or to change the links in the feed to point somewhere else that's not missing? If you can easily point me to the right neighborhood in the code I might work on a patch for this. It may be a (minor) bug in the grand scheme of things, but it does seem pretty goofy if you've just clicked an RSS link. :) --Chap (p.s. should this be moved to bugs?)

The latter -- I think making the permalink point to "recentchanges#someid" will probably work. Probably first by addressing the todo about [[todo/ability_to_force_particular_UUIDs_on_blog_posts]], and then by just using that new ability in the page. --[[Joey]]

Ah. The prerequisite todo looks like more than I'd like to take on. In the meantime, would it be very involved to change whatever bug now optimizes away the change pages, or to simply have all the links in the feed point to the recentchanges page itself, with no fragment id? Either would be a bit nicer than having broken links in the feed. --Chap

Does the completion of that todo mean it would be straightforward to get recentchanges working now? Is it just that the recentchanges plugin needs to generate \[[!meta guid=something]] into the internal files, and the inline plugin would then generate working links in feeds? How should the guid be constructed? Something based on the rcs revision number? I guess I'm still not completely clear on your vision for how it ought to work. --Chap

My idea is to use \[[meta guid="http://url/recentchanges#rev"]], with the #rev anchor also included in the change file, and being the rcs's internal revision id. Then the guid is globally unique, and actually links to the change in the recentchanges page. And, when the change has fallen off the page, the link will still go to the recentchanges page.

First, need to check that guids in rss and atom feeds can have anchors in them, and that the anchor is treated as part of the guid. (If the guid is interpreted as just "http://url/recentchanges", then it's not a very good guid.) If using an anchor for a guid is a problem, it could instead generate a random uuid, and use \[[meta guid="urn:uuid:<foo>" permalink="http://url/recentchanges"]]

I had a quick look into this after fixing the "prerequisite", but got bogged down in minor details. Anyway, I'd be happy to help. I think the guid stuff is actually fairly irrelevant, you just need \[[!meta permalink]] (and in fact you're using guid incorrectly, by expecting it to be treated as a link).

My advice would be: first, fix the bug as reported, by using \[[!meta permalink="http://blah/blah/blah#change-$rev"]] (starting anchor names with a number isn't syntactically valid, if I remember correctly, so do have a prefix like "change-" or "rev-" or something).

Then, optionally, force the guid too (although it defaults to the permalink anyway, so this shouldn't actually be necessary).

Some more explanation of how guids work: it's actually easier to think about them in Atom terms than in RSS terms, since Atom has a clearer conceptual model.

The \[[!meta permalink]] becomes the <link> element in Atom, which contains a link that users can follow; if it's not explicitly given, ikiwiki uses its idea of the page's URL.

The \[[!meta guid]] becomes the <id> element in Atom, which contains an opaque, not-necessarily-resolvable identifier; if it's not explicitly given, ikiwiki uses the same URL as the <link>.

In RSS the semantics aren't so clear-cut (which is part of why Atom exists!), but the way ikiwiki interprets them is:

  • <link> is the same as in Atom
  • if \[[!meta guid]] is explicitly given, put it in <guid permalink="no"> (the assumption in this case is that it's a UUID or something)
  • if \[[!meta guid]] is not explicitly given, copy the <link> into the <guid>

I believe RSS aggregators (are meant to) compare <guid>s as opaque strings, so using an anchor there should be fine. Atom aggregators are certainly required to compare <id>s as opaque strings.

--[[smcv]]

Here's my attempt at a [[patch]] for anchor-based change permalinks: http://pastie.org/295016. --[[JasonBlevins]], 2008-10-17

[[JasonBlevins]] nailed it, [[done]] --[[Joey]]

Thanks for applying the patch (and improving it). There's still one small issue: the old opening div tag still needs to be removed (it's hard to see the removed line with the pastie color scheme). --[[JasonBlevins]], 2008-10-18

Thanks, missed that when I had to hand-apply the patch. --[[Joey]]