488592 - Planet is stripping headers in meeting minutes

Status: RESOLVED FIXED

(Websites :: planet.mozilla.org, defect)

Description

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

In this week's meeting minutes, I noticed it looked like some "headers" were missing (big chunks of white space that looked like they should have informative text in them).

When checking the originating meeting-notes blog, I noticed the headers were present; I also verified their presence in the feed, so it seems planet.m.o itself is responsible for stripping.  I don't remember this happening last week; did something change in the past week?

STR:

1) View the Thunderbird (or Platform) meeting notes post on planet.m.o
2) Observe blank spaces where it seems like there should be headers
3) View-source and observe only blank space inside the <h*> tag:

<p><a name="Agenda" id="Agenda"></a><br />
</p><h4> </h4>
<ul>

<li> Who’s taking minutes? –&gt; <b>dmose</b><p></p>

4) Open http://blog.mozilla.com/meeting-notes/archives/162 and note headers ('Agenda') are present on the page:

<p><a id="Agenda" name="Agenda"></a><br />
<h2> <span class="mw-headline"> Agenda </span></h2>
<ul>
<li> Who&#8217;s taking minutes? &#8211;&gt; <b>dmose</b></p>

5) Open/view-source http://blog.mozilla.com/meeting-notes/feed and note headers ('Agenda') are present in the feed:

<p><a id="Agenda" name="Agenda"></a><br />
<h2> <span class="mw-headline"> Agenda </span></h2>
<ul>
<li> Who&#8217;s taking minutes? &#8211;&gt; <b>dmose</b></p>

ER: non-blank <h*> tags on planet.m.o
AR: blank <h*> tags on planet.m.o

Comment 1

[Reed Loden [:reed]]

No changes have been made to planet in the last week besides maybe adding somebody new, so I doubt it's anything new.

Comment 2

[Robert Accettura [:raccettura]]

Doesn't look like any changes made to meeting-poster lately that would do this...
http://hg.mozilla.org/users/bsmedberg_mozilla.com/meeting-poster/

cc: bsmedberg as he was the original author.

Comment 3

[Benjamin Smedberg]

Well, since the meeting-notes feed appears to be well-formed, I'm gonna call this a planet bug, not that that helps much.

Was there a change to wikimo? The HTML in meeting-poster is basically just wikimo stripped of the "outside" content. And I did have to make a local change to the meeting-poster because formerly missing pages didn't have a 404 HTTP code, and in the past few weeks they started returning a 404 response code.

Comment 5

[Geoff Lankow (:darktrojan)]

Unless I'm much mistaken, http://blog.mozilla.com/meeting-notes/ hasn't picked up new meeting notes since January, and therefore neither has planet. What gives?

Comment 6

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

Looking back at the HTML snippets I posted in comment 0, the missing header text is all inside <span></span> tags (inside the H*).  

Hypothesis: Planet's sanitizer is being overly zealous and is stripping the content of the <span></span> tags as well as the tags themselves?  Or perhaps that plus the extra spaces (<h2> <span …> Agenda </span></h2> are triggering some sort of edge-case bug in the sanitizer where it sees the space and thinks that it all of the content of the headers?

Comment 7

[Benjamin Smedberg]

The machine which was running meeting-notes was reinstalled and I never resurrected it anywhere.

Comment 8

[Geoff Lankow (:darktrojan)]

This has been fixed upstream... we just need someone to update Planet Mozilla to the latest version.

Comment 9

[Geoff Lankow (:darktrojan)]

.... or at least there appears to have been. (The previous comment may have been based on faulty information, and/or me jumping to conclusions.)

In any case I added meeting notes to my own experimental Venus install and it's working fine.

Comment 10

[Reed Loden [:reed]]

(In reply to comment #8)
> This has been fixed upstream... we just need someone to update Planet Mozilla
> to the latest version.

We're already running the latest version of Venus, so I don't think that's the problem.

Comment 11

[Geoff Lankow (:darktrojan)]

(In reply to comment #10)
> We're already running the latest version of Venus, so I don't think that's the
> problem.

You're right, it isn't. To atone for my sins I've found the bug in the minhead.py filter and just posted it to the Venus mailing list.

Comment 12

[Geoff Lankow (:darktrojan)]

I see headings (including the mysterious h7 element)! Thanks for getting that fix over the finish line, Reed.

Comment 13

[Reed Loden [:reed]]

(In reply to comment #12)
> I see headings (including the mysterious h7 element)! Thanks for getting that
> fix over the finish line, Reed.

:)