Closed
Bug 82385
Opened 24 years ago
Closed 24 years ago
Check in workaround to nsStdURL
Categories
(Core :: Networking, defect)
Core
Networking
Tracking
()
VERIFIED
WONTFIX
mozilla0.9.1
People
(Reporter: dougt, Assigned: jud)
Details
(Keywords: crash, Whiteboard: 0.9.1 workaround ready and will check into branch)
Based on bug 73845, less risky workaround will be checked into the 0.9.1
branch. This will prevent a crash caused by some wrong usages of necko.
Please read 73485 and my netlib newsgroup positing for background.
Reporter | ||
Updated•24 years ago
|
Keywords: crash,
mozilla0.9
Whiteboard: workaround ready and will check into branch
Target Milestone: --- → mozilla0.9.1
Updated•24 years ago
|
Whiteboard: workaround ready and will check into branch → 0.9.1 workaround ready and will check into branch
![]() |
||
Comment 1•24 years ago
|
||
Jud, we've branched now, so can this go in?
![]() |
Assignee | |
Comment 2•24 years ago
|
||
so, turns out no-one ever ran w/ my patch until I just did. I can't even load a
url from the url bar w/ the patch.
Radha, session history is being handed strings directly from the url bar which
seems a little odd to me. So, if I type "www.foo.bar" (no scheme), session
history tries to create a URI out of it here
http://lxr.mozilla.org/seamonkey/source/xpfe/browser/resources/content/sessionHistoryUI.js#143
and with the patch, that JS fails. Shouldn't session history be building it's
URI list from things *after* they've gone through the URI fixup stuff in the
docloader (ie. after the load has started?)?
I suspect gtkEmbed will run just fine (will verify after this posting), and that
this session history issue is a highly visible edge case (just like the other
crashers that seemingly appear at random) of bug 73845.
![]() |
Assignee | |
Comment 3•24 years ago
|
||
yea, gtkEmbed works fine for general browsing. I guess I'm just as uneasy about
this patch as I was from the start, and would prefer to have 73845 fixed instead
(I'm sure we all would, it's just that it's a hard fix :-/). there are a couple
dozen places where we create raw URL's, and at least half of those have the
*potential* for getting something schemeless in there which would produce
unpredictable behavior.
![]() |
Assignee | |
Comment 4•24 years ago
|
||
I'm snuffing this, AFAIK this isn't a frequent crasher, and we'll wait for the
real deal in 0.9.2.
Status: NEW → RESOLVED
Closed: 24 years ago
Resolution: --- → WONTFIX
![]() |
Assignee | |
Comment 5•24 years ago
|
||
cc'ing radha re: the session history questions above.
![]() |
||
Comment 6•24 years ago
|
||
That code in sessionHistoryUI.js is for urlbar history. That segment of code
compares the url in the urlbar to all the entries in RDF datastore and decides
if the typed-in value should be added to ub-history. Session History adds urls
to its history in docshell after it has been resolved by docloader.
You need to log in
before you can comment on or make changes to this bug.
Description
•