Reformatting breaking links

November 7, 2001 10:57 AM

links in comments are getting busted.
posted by moz to Bugs at 10:57 AM (21 comments total)

yeah I noticed that. I found the same problem a couple of weeks ago but only on a couple of links. If it can happen to me, dear god...it could happen to anyone!

Kindly begin panicking now!
posted by Kafkaesque at 11:31 AM on November 7, 2001

Yup, it's happening in the HP thread.
posted by msacheson at 11:32 AM on November 7, 2001

And here, evidently. D'oh!
posted by msacheson at 11:33 AM on November 7, 2001

It's everywhere. The site seems to be reformatting hrefs in a way that breaks them (on some browsers, anyway; everything still works fine for me on Mac IE 5.1).
posted by jjg at 11:35 AM on November 7, 2001

Yeah, jjg, sorry if it seemed like I jumped on you back there on MeFi. I won't try to post any links until it's fixed so as to not add to the load of what would have to fixed.
posted by msacheson at 11:38 AM on November 7, 2001

the thing is that the tag reformatting scheme is using the & quot; character instead of double quotes for the href tags. i'm not sure what the W3C specs say about this situation (they may leave the behavior undefined), but it definitely is breaking some browsers.
posted by moz at 11:41 AM on November 7, 2001

everything still works fine for me on Mac IE 5.1

me too, but for one thing -- jjg, are the '#' symbols in the hrefs (e.g., in moz' links (like this one) at the top here) being changed to '%' symbols for you? And is that happening to everyone?
posted by mattpfeff at 11:42 AM on November 7, 2001

scott's got an answer: use single quotes instead of double quotes for your links in the time being.
posted by moz at 11:43 AM on November 7, 2001

yep. working for my mac, too (IE 5.0)
posted by ColdChef at 11:44 AM on November 7, 2001

>>everything still works fine for me on Mac IE 5.1

I can't get my work done yet on OSX, but I guess that's the whole point of MetaFilter... : P


posted by machaus at 11:44 AM on November 7, 2001

mattpfeff: Nope. #s are still #s.

msacheson: Make it up to me by posting some sort of comment to bring my thread back on topic.
posted by jjg at 11:47 AM on November 7, 2001

jjg: how's this?
posted by msacheson at 11:59 AM on November 7, 2001

In addition, some paragraph breaks are being set up strangely, such as in this case. If you look at the source you can see that there's a mix of BRs and P's.

With all due respect to Leonard, I think maybe the new parsing code should be rolled back out until it's been more carefully tested.
posted by Steven Den Beste at 12:30 PM on November 7, 2001

I just rolled back the code to a previous release. Things should be good now.
posted by mathowie (staff) at 12:36 PM on November 7, 2001

Things should be good now

Are they? Because I'm still seeing it on this page at least.
posted by daveadams at 1:19 PM on November 7, 2001

Oh you mean in the future...

*slaps forehead, falls backwards into mud puddle, cries*
posted by daveadams at 1:20 PM on November 7, 2001

heh, it was late last night, obviously there was a reason i wasn't doing entity filtering on the attribute before (added a line to fix a bug nothing spotted). i have a fixed version now that i just sent to matt. (converts "e; back to " after filtering)

steven - the br's and p's isn't my code. that's all matt's translations of crlf's i believe. the only thing that my code does is make sure any open p's are all getting closed at the end of the text.
posted by lhl at 1:55 PM on November 7, 2001

*slaps forehead, falls backwards into mud puddle, cries*

Now see what you've done to yourself?
posted by iceberg273 at 4:52 PM on November 7, 2001

Isn't "a bug nothing spotted" a line from Shakespeare?

(also: thanks, msacheson)
posted by jjg at 11:36 PM on November 7, 2001

Wiill somebody cry already?
posted by stavrosthewonderchicken at 4:59 AM on November 9, 2001

Oops. Cancel that. Dave cried. All is well. Recall the bombers.



posted by stavrosthewonderchicken at 5:02 AM on November 9, 2001