Incorrect OPML source for comments

31 views
Skip to first unread message

Zach Thomae

unread,
Sep 23, 2012, 12:38:57 PM9/23/12
to fronti...@googlegroups.com
When I try to download the OPML source of any comment (for demonstration purposes, this one) with the XML button at the bottom of the page, I don't get the complete source--it's cut off after the opening body tag. In other words, this is an example of what I get:

<?xml encoding="ISO-8859-1" version="1.0"?>
<!-- OPML generated by OPML Editor v0.75 on 9/23/2012; 12:09:52 PM -->
<opml version="2.0">
<head>
<title>Dody Gunawinata</title>
<ownerName>Dave Winer</ownerName>
<ownerEmail>dave....@gmail.com</ownerEmail>
<windowTop>200</windowTop>
<windowLeft>200</windowLeft>
<windowBottom>600</windowBottom>
<windowRight>800</windowRight>
</head>
<body>

But it's not just incomplete--the ownerName and ownerEmail tags have Dave's information in them. Knowing the date and the id of that comment, I can find it manually at http://static.scripting.com/pensacola/comments/2012/09/23/00119.opml. The head tree is noticeably different, and the body is complete now:

<?xml encoding="ISO-8859-1" version="1.0"?>
<!-- OPML generated by OPML Editor v10.1a8 on Sun, 23 Sep 2012 15:57:25 GMT -->
<opml version="2.0">
	<head>
		<title>Comment from Dody Gunawinata</title>
		<dateCreated>Sun, 23 Sep 2012 15:57:25 GMT</dateCreated>
		<dateModified>Sun, 23 Sep 2012 15:57:25 GMT</dateModified>
		<ownerName></ownerName>
		<ownerEmail></ownerEmail>
		<expansionState></expansionState>
		<vertScrollState>1</vertScrollState>
		<windowTop>169</windowTop>
		<windowLeft>332</windowLeft>
		<windowBottom>314</windowBottom>
		<windowRight>657</windowRight>
		</head>
	<body>
		<outline text="We have seen already what happened when a complex system gets too big to understand four years ago with the catastrophic meltdown of the global financial system. There are people literally thrown out to the street in Greece and Spain because of this financial calamity. At some point these 'clouds' will become too big to fail and governments will have to bail them out because everything digital that we have would be stored there."></outline>
		<outline text=""></outline>
		</body>
	</opml>

It looks like the code generating the OPML source with the ?format=opml param is misbehaving.

Zach

Dave Winer

unread,
Sep 23, 2012, 1:00:58 PM9/23/12
to fronti...@googlegroups.com
It's an edge case that's never come up before. A chance to fix a bug. To get closer to perfection. :-)

Ivan Vučica

unread,
Sep 23, 2012, 1:10:48 PM9/23/12
to fronti...@googlegroups.com
Hi,

On 23. 9. 2012., at 18:38, Zach Thomae <zth...@gmail.com> wrote:

<?xml encoding="ISO-8859-1" version="1.0"?>

I didn't look at the generated OPML lately, but in light of recent discussions and problems people had with encodings, this line has screamed at my eyes like voices of a thousand banshees.

Is there a particular reason why US-and-Western-Europe-centric ISO-8859-1 is used in place of UTF-8?


Dave Winer

unread,
Sep 23, 2012, 1:13:26 PM9/23/12
to fronti...@googlegroups.com
This is a support list. 

Write a blog post about banshees screaming at you.

Dave

Dave Winer

unread,
Sep 23, 2012, 4:02:13 PM9/23/12
to fronti...@googlegroups.com
It should work now.


Dave

Zach Thomae

unread,
Sep 23, 2012, 4:11:01 PM9/23/12
to fronti...@googlegroups.com
I tested it on a few comments, and yes, it works on them now.

However, it doesn't work on the comments outline page--same problem.

Zach

Dave Winer

unread,
Sep 23, 2012, 4:15:10 PM9/23/12
to fronti...@googlegroups.com
Try it again. :-)

Zach Thomae

unread,
Sep 23, 2012, 4:20:33 PM9/23/12
to fronti...@googlegroups.com
Yup, that works now too.

Thanks,
Zach
Reply all
Reply to author
Forward
0 new messages