Closed Bug 4230 Opened 26 years ago Closed 25 years ago

M3 doesn't work with AutoConnect's "Used Cars for Sale" feature

Categories

(Core :: Networking: Cookies, defect, P1)

x86
Windows NT
defect

Tracking

()

VERIFIED WONTFIX

People

(Reporter: simon.arthur, Assigned: morse)

References

()

Details

When attempting to use "Used Cars for Sale", I get an error message after filling in the second form. This page works OK under Mozilla 4, 4.5 and IE.
Priority: P1
Simon, Can you please provide a auto make and zip code that will actually find a car so that I can get to the 2nd form. I tried several makes with "any" body style and the zip codes I know about, but it did not find any cars so I didn't see the 2nd form that is causing the problem.
Volkswagen and 95112 (downtown San José) work using 4.51, but not using seamonkey. Presumably there's a problem in the form submission itself... or with JavaScript? Copying and pasting from the DOS window, this is what I found using that search: url=http://www.autoconnect.com/findacar/findacar_form2.jtmpl?make=VOLKS&ca e=Any&address=95112&borschtid=85462902330121807920&search_type=FindSale&sor e=sort_by_price&form_status=FindACar_form&x=18&y=8 data=?make=VOLKS&car_type=Any&address=95112&borschtid=85462902330121807920 ch_type=FindSale&sort_type=sort_by_price&form_status=FindACar_form&x=18&y=8
Assignee: karnaze → morse
Component: Form Submission → Cookies
I think this is a cookie problem and so I'm reassigning to Steve. In Nav4.5 I went to the 1st url (call it url 1) and entered (volkswagon,any,95112) and saved the url of the 2nd form (call it url 2). url 2 worked in viewer but viewer could not generate a url like url 2 that would work. Also, after some time or after visiting url 1 over and over again (I could not duplicate it), even Nav4.5 generated the "sorry ..." message on url 2 above. So, since the comments in url 1 mention cookies, I think they are the culprit. Also, Nav4.5 generated 2 occurrences of borschtid=.. in url 2 but removing 1 or both of them from url 2 had no bad effects.
Status: NEW → ASSIGNED
Assignee: morse → karnaze
Status: ASSIGNED → NEW
I'm not convinced this is a cookie problem. But, unfortunately I can test this out because of a regression that prevents me from getting to the second form. Get a crash with the stack trace listed below. Looks like layout to me so I'm reassigned to karnaze. If we still get an error message after this regression is fixed, assign it back to me and I'll look further. BTW, I tried to run a second time to see if the crash was reproducible but unfortantely the site must have gone down -- I now keep getting an error message "Unspecified error has occurred" from the site when I go the the indicated url. And this is not seamonkey specific; the same error message occurs in 4.5. Was this the error message that the original bug report was about? If so, it could be a problem with the site and not with seamonkey. IL_InterruptContext(_IL_GroupContext * 0x01203fc0) line 1981 + 3 bytes ImageGroupImpl::Interrupt() line 286 + 12 bytes nsPresContext::~nsPresContext() line 88 GalleyContext::~GalleyContext() line 41 GalleyContext::`scalar deleting destructor'(unsigned int 1) + 15 bytes nsPresContext::Release(nsPresContext * const 0x0249c4b0) line 121 + 34 bytes PresShell::~PresShell() line 545 + 27 bytes PresShell::`scalar deleting destructor'(unsigned int 1) + 15 bytes PresShell::Release(PresShell * const 0x011fd830) line 483 + 34 bytes DocumentViewerImpl::~DocumentViewerImpl() line 216 + 18 bytes DocumentViewerImpl::`scalar deleting destructor'(unsigned int 1) + 15 bytes DocumentViewerImpl::Release(DocumentViewerImpl * const 0x02560850) line 153 + 99 bytes nsWebShell::Embed(nsWebShell * const 0x011d9120, nsIContentViewer * 0x025d88c0, char * 0x025e0530, nsISupports * 0x00000000) line 711 + 27 bytes nsDocumentBindInfo::OnStartBinding(nsDocumentBindInfo * const 0x025e04e0, nsIURL * 0x025e07e0, char * 0x025d69b0) line 1898 + 36 bytes OnStartBindingProxyEvent::HandleEvent(OnStartBindingProxyEvent * const 0x025d6960) line 506 StreamListenerProxyEvent::HandlePLEvent(PLEvent * 0x025d6964) line 471 + 12 bytes PL_HandleEvent(PLEvent * 0x025d6964) line 476 + 10 bytes PL_ProcessPendingEvents(PLEventQueue * 0x00ee1c00) line 437 + 9 bytes _md_EventReceiverProc(void * 0x01080868, unsigned int 49314, unsigned int 0, long 15604736) line 799 + 9 bytes USER3
Status: NEW → ASSIGNED
Target Milestone: M6
Moving to M6.
Assignee: karnaze → morse
Status: ASSIGNED → NEW
Steve, the crash is gone now on my 5/6 debug WinNT. Reassigning to you.
Status: NEW → ASSIGNED
Status: ASSIGNED → RESOLVED
Closed: 25 years ago
Resolution: --- → WONTFIX
Much as I hate to do this, I am closing out this bug as "won't fix". Here's why: First off, the symptoms are now different. Specifically, when I submit the first form I get the error message saying "There are no Volkswagens available at this time" (I never even get to the second form). And with a 4.51 browser I am able to get the VW info. As far as I can determine, cookies are being set up correctly and are being transmitted to the site. So my hunch is that the site's cgi script is for some reason is rejecting requests from a browser it doesn't recognize and it apparently doesn't recognize seamonkey. In an attempt to get more information, I tried to contact the site. There is no phone contact information given anywhere on the site. There is a form to fill out to send them e-mail questions. So I tried that (using a 4.51 browser so there wouldn't be any problems) and got back a message saying that there was an error encountered in sending the e-mail. So, as a last resort, I sent the following e-mail to postmaster@autoconnect.com. If I get a reply back from them, I will investigate further to see if we can fix the problem. Otherwise it has to remain a "won't fix" since there is nothing further we can do at our end. Here's the e-mail I just sent to postmaster@autoconnect.com: ************* We at Netscape are experiencing a problem using autoconnect with our 5.0 browser which is under development. In short, we get the message "sorry but there are no ... available at this time." With the same input we are able to get correct results with a 4.51 browser. As far as we can determine, we are correctly passing the appropriate cookies to your website. Please contact me (415-775-0810 day or evening) so we can find the problem and be able to support your site with our 5.0 browser. Thank you. -- Steve Morse P.S. I tried to send this message via the contact information on your webpage but kept getting an error message when I pressed submit. And that was with a 4.51 browser!
Waiting to verify; will do so either in 24 hours or if simon.arthur agrees that there's problem on autoconnect.com's end.
Following message was just received from Simon Arthur at autoconnect: > Is there anything different in the way cookies are transmitted in 4.51 > vs. Seamonkey? Our site works with most browsers: WebTV, Lynx, etc. To which I replied: > As I indicated in the bug report and in the message that I sent to your > postmaster, there is nothing further that we at Netscape can do by > ourselves on this bug. If autoconnect is interested in having this > fixed, then we are going to need the help of a technical person at your > site. Basically we need to run the test and see what is happening at > the server side as to why the server is not giving us the information. If > you are that person, then please give me a call (415-775-0810). If it's > not you, then please find out who the correct person is and have him or > her call me. This is the only way we can procede on this bug.
Status: RESOLVED → VERIFIED
You need to log in before you can comment on or make changes to this bug.