Documentation Channel for #raku | This channel is logged | Roadmap:
Set by [Coke] on 23 May 2022.
01:03 cfa joined
cfa [Coke]: i'm seeing a few failures like this, 01:07
# Failed test 'Pod::FormattingCode<2463305740736>.type<C<Cool>> should be C<Cool> - bad parent FormattingCode - self reference'
(the self-reference failure is correct)
there are also a number of false positives like this: 01:08
'L<C<Bool>|/routine/Bool> should be L<C<Bool>|/type/Bool> - bad link'
i.e. methods that match known types
01:24 cfa left
Geth doc/main: 8200e07f8e | cfa++ | doc/Language/5to6-perlvar.rakudoc
Fix link target
03:04 rf left 09:04 raiph joined 09:05 raiph left
Geth doc/type-links: 6ac220d0b0 | (Will Coleda)++ | xt/rakudoc-types.rakutest
Allow links to /routine pages

[Coke] cfa: that allows anything with a link to /routine/* to pass
sounded like the first item with the formattingcode was that the test was OK, but the output was badly rendered. 13:35
Geth doc/type-links: 814fd3a4e1 | (Will Coleda)++ | xt/rakudoc-types.rakutest
Fix interpolation

Avoid interpreting the literal < as an index of the variable.
[Coke] fixed that also
Getting closer...
14:22 cfa joined
cfa [Coke]: 👍 14:22
should i rebase normalise-type-links
onto main?
(or is having it as a child of your test branch convenient for now?) 14:23
we're down to 62 failures btw
Geth doc/type-links: b986c47e24 | (Will Coleda)++ | xt/rakudoc-types.rakutest

14:48 cfa left
[Coke] I was going to rebase and squash commits but if you're downstream I didn't want to break it. 14:54
can you rebase when I'm done and be OK?
(and force push, there'd be a force push in there) 14:55
... oops, he left. 15:04
.ask cfa - is it ok if I rebase test-links branch and squash commits and force push?
15:12 cfa joined
cfa [Coke]: sure, please go ahead 15:12
i'll force push a revised branch (rebased onto main) within the next hour or so
remaining automated rewrites are possible but probably more error prone; given the number of failures we have, curated edits seem to make more sense? 15:13
perhaps we can automate stuff like B<Str> => L<C<Str>|/type/Str> if it's recurrent 15:14
Geth doc/type-links: 5 commits pushed by (Will Coleda)++ 15:23
[Coke] .. why is that different commits?
I think the branch is rebased on main and is now one commit
15:45 rf joined
cfa why is what different commits? 15:46
is Geth misreporting something? 15:47
[Coke] I just squashed everything and pushed a single commit, so I'm not sure why geth is reporting the old individual commits there... but it was a force push, so not normal anyway 15:48
cfa ah okay
i'll refresh my branch in a sec
so you can take another look
done: 15:56
18:30 cfa left 20:27 raiph joined
raiph o/ 20:28
First, thanks ugexe++ for fixing up some doc links in an SO answer I wrote 20:29
Second, is there some group initiative related to such updating 20:31
Third, is there a plan to bring missing content back online? 20:34
I ask the last because I'm thinking how best to go through my 420 answers 20:40
And I've yet to find the content I linked to in some cases 20:41
So I'm thinking it's gone for the time being (and maybe for good)
More generally there are issues like that 20:42
Content moving, sections being renamed, etc 20:43
And I'm wondering if it's best to wait a year or so for that to settle
For example, most of the links ugexe updated don't arrive at their intended target 20:44
Some go to the right page but not the section
Others go to the wrong page though the right (new) one is in place
For others still the original content doesn't seem to currently be in place anywhere in the site 20:45
For reference, here's the sample SO that ugexe updated: 20:57
And here is his revision work: 20:59
Ah, I've got an idea about how best to go forward. 21:04
Anyhow, any related info / comments / links appreciated 21:05
21:14 raiph left
[Coke] Yes, we are trying to fix all missing content and #ids. - if you see a class of items missing that isn't on the doc/website issue list, please ping us. 23:35