Closed Bug 6193 Opened 26 years ago Closed 25 years ago

NECKO: M5 Mozilla will not load Yahoo! News Stories links on Yahoo main page

Categories

(Core :: Layout, defect, P3)

x86
Linux
defect

Tracking

()

VERIFIED FIXED

People

(Reporter: thully, Assigned: gagan)

References

()

Details

Kernel 2.2.6, Mozilla M5, KDE 1.1, Red Hat 5.2 Yahoo! News stories Links on the main Yahoo! page will not load when clicked in Mozilla M5.
Assignee: rickg → joki
Well it looks like the problem is with the url being given which is something like '/homer/?http://...'. Looking further to see if I'm munging the URL or if this is a netlib problem.
Assignee: joki → gagan
The problem seems to be in NS_MakeAbsoluteURL. Reassigning to netlib.
With the latest Linux Build (1999063008), I can't reproduce the problem described. Clicking on the New stories link goes to correct page.
Pushed past necko landing...
Summary: M5 Mozilla will not load Yahoo! News Stories links on Yahoo main page → NECKO: M5 Mozilla will not load Yahoo! News Stories links on Yahoo main page
Pl. verify with Necko.
Windows NT 4 build 7/25/08 My Netscape seems to have the same problem. Clicking on any of the links in the Netcenter Apps box yields "Error Loading URL http://" I looked at the source from the page and the markup looks like this: <A HREF="z.t?cp=myna/http://home.netscape.com/netcenter/mail/index.html">
Known problem with parsing. Marking for M10.
Target Milestone: M9 → M10
Assignee: gagan → kipp
Status: ASSIGNED → NEW
Kipp wants to help. Kipp: from what I see we are parsing for :// and since this is a relative link that portion throws off our logic. So we first need to check for ? to detect any random characters after it as query. Let me know if you need more help.
Status: NEW → ASSIGNED
Target Milestone: M10 → M11
Can you explain what all the '?' cases are for nsStdURL::Parse? I don't recall ? being allowed in so many places...thanks
Once more, with gagan on the cc list: Can you explain what all the '?' cases are for nsStdURL::Parse? I don't recall ? being allowed in so many places...thanks
Assignee: kipp → gagan
Status: ASSIGNED → NEW
Ok, I've done *some* work to make the url parsing work better. I updated urltest to test the cases implied by this bug and they seem to work (note: there are a few test cases in urltest that don't pass -- try "urltest -abs foo" to see them). Basically I made it so that "abc;param?query#ref" will always parse with "#" taking precedence over "?" which takes precedence over ";". Therefore, if there are funny characters in refs or queries, they will retain them. Which makes this sort of thing: "?http://one-two-three;hello" do the "right" thing. It's not clear if the ref handling is exactly right, however.
*** Bug 14697 has been marked as a duplicate of this bug. ***
can this be marked fixed
Status: NEW → RESOLVED
Closed: 25 years ago
Resolution: --- → FIXED
links to news stories off www.yahoo.com work for me on 11/5 linux and win32 builds. marking fixed.
Status: RESOLVED → VERIFIED
With the Nov 15th Linux build, this problem is fixed.
You need to log in before you can comment on or make changes to this bug.