Dear Mozilla, My first official bug report. Bear with me if it sucks. The page in question is http://www.panic.com/index.html . Turn off JavaScript in Netscape 4.5 and take a peek at it. There's a somewhat simple table, surrounded by <CENTER> tags to center the table on the page. This page, when viewed in Mozilla's viewer (or apprunner), doesn't seem to notice the <CENTER> tags, i.e. the table in question is left-justified. It's done this for quite a few versions, I should've reported it a while ago, I know. ;) In the latest Mac build (M4), pulling up this page in viewer or apprunner also causes a reproducable crash on my Macintosh G3/300 running 8.5.1. I'm not sure if it affects other platforms. Hope this helps! Regards, Cabel Panic

Pierre, can you find out why this is crashing on MAC. It didn't crash on my WinNT debug build. After that, please reassign to me to address the table centering problem.

It looks like it is another problem with the Mac timers. Reassigned to <jfrancis> because <beard> is already doomed for M5. Or maybe <gordon> should get it because it's related to NetLib too... Anyhow, don't forget to reassign to <karnaze> when you're done with the crash.

accepting bugs

I need some serious netlib help with this. So far this does not appear to be a mac timer bug. Here's the scoop: net_ProcessHTTP() is getting called recursively. In the nested call, it is detecting and error and freeing the connection data. Then when it pops back up to the top net_ProcessHTTP() call, the code is not savvy to the fact that the connection data has disappeared out from under it. The call chain between the two net_ProcessHTTP() calls is long and winding, starting in netlib (of course, and going through the parser, layout, dom, and javascript, plus a bunch more stuff, before getting back into netlib. To see the call chain, set a breakpoint on line 3594 of network/protocol/http/mkhttp.c. The crash then happens later on line 3684. I do not want to touch this stuff - I don't understand it. Gagan? Someone? Will someone find the right person for this bug? cc'ing gagan, gordon, and hofmann.

adding jevering to the cc list to help make the call. we are in a tough place right now with libnet. short term work on the old code just adds to the schedule for getting the new necko code landed... We may have to live with this until M6 or M7.

assigning to gagan with his permission. i believe he is going to set this to M6 or later.

Marking till Necko lands...

Per DP's suggestion marking these till M8. Though Necko lands with M7, we will be able to verify it for M8.

When the crash is verified fixed, please reassign to me to fix centering problems.

I'm moving this to target M9, Necko will be enabled somewhere during late M8 or early M9. We will need to get on this and it cannot be postponed past the M9 milestone.

Pl. verify with Necko

Modified summary since Apprunner no longer crashs when loading this page on the Mac. However, the table at the above url is not centered and appears left aligned on Mac 8.6, Win NT, Win 98, and Linux Aug 3rd Builds.

I'm sure this is a dup of another bug. Reassigning to Kipp.

There are a couple of problems in the HTML code here: * The CENTER tag is centering a table of WIDTH=100%. Whether it's centred or not is irrelevant * The ALIGN attribute of the outer table cell is using "MIDDLE" instead of "CENTER", although most browsers seem to interpret this mistake the way the author intended (see http://www.w3.org/TR/REC-html40/sgml/dtd.html#cellhalign for details) Finally, bug 7112 prevents a table from being centred inside another table. Marking this one a duplicate of bug 7112. *** This bug has been marked as a duplicate of 7112 ***

Agreed. This bug is a duplicate of 7112.

Agreed. This bug is a duplicate of 7112.