Closed Bug 117361 Opened 23 years ago Closed 23 years ago

Can't send mail through Yahoo mail

Categories

(Core :: Networking, defect, P1)

Product:
Core
Component:
Networking
Platform:
All
Windows 2000
Type:
defect

Tracking

()

Status:
RESOLVED DUPLICATE of bug 117363
Milestone:
mozilla0.9.9

People

(Reporter: bugzilla, Assigned: darin.moz)

References

(
URL
)

Details

(Keywords: hang, helpwanted, qawanted)

Attachments

(4 files)

Log file
175.01 KB, text/plain
Details
caputered using ethereal, saved in tcpdump format
3.84 KB, application/octet-stream
Details
yahoo send msg (IE traffic), caputered using ethereal and saved in tcpdump format
38.90 KB, application/octet-stream
Details
successful send traffic from older build (tcpdump format )
39.27 KB, application/octet-stream
Details 
Build ID: 2001122803 comm.

A number of 6.2 users have reported an inability to send mail through Yahoo 
mail. I can confirm that this problem still exists in new nightlies.  I found 
other Yahoo mail-related issues reported, but couldn't find one about sending.

The browser just hangs (throbber/status messages going) when trying to send.

Test account ID: mozdebug
Password: mozilla
Nominating for machv since this is a very high profile site.
Keywords: nsbeta1
This isn't bug 116210, I assume (since you're seeing it in a current non-branch 
nightly)?
Yup. Trunk nightly.
darin
Assignee: neeti → darin
will try to investigate this one soon.
Status: NEW → ASSIGNED
Priority: -- → P1
Target Milestone: --- → mozilla0.9.8
Keywords: hang
WFM on W2K
WFM too
can anyone reproduce this problem using a recent nightly build?
Keywords: qawanted
Still doesn't work for me in 1/9 comm build, win2k...Ben, can you give it a shot?
This worked for someone else on irc. I don't know why it's still not working for
me. It apparently was broken at some point for 6.2.1 users, so I'm hesitant to
toss it outside too abruptly. Let me see if I can get some more testing from people.
perhaps it has something to do with the commercial build?!?
i tested sending several different messages (some w/ attachments, some w/o) from
the mozdebug account on mail.yahoo.com... all were successful.

marking WORKSFORME (win32 commercial build 2002-01-14-09 under winnt4sp5)
Status: ASSIGNED → RESOLVED
Closed: 23 years ago
Resolution: --- → WORKSFORME
I gotta reopen this because of bugs 122058 and 122262.
Status: RESOLVED → REOPENED
Resolution: WORKSFORME → ---
I am seeing this bug in the latest build I downloaded. I am working on Windows NT.
I'm seeing this as well in build 2002012203.
WFM 2001013003 W2K
if you can reproduce this problem with a nightly build, then please please
please set the following environment variables to generate a HTTP log.  w/o
this, it's next to impossible to have any hope of fixing this.

here's what to do:

on windows:

1) open up a DOS prompt
2) type: set NSPR_LOG_MODULES=nsHttp:5
3) type: set NSPR_LOG_FILE=c:\http.log
4) type: cd \path\to\mozilla
5) type: .\mozilla
6) reproduce the problem
7) attach c:\http.log to this bug report.

thx in advance to anyone able to generate this log file!

leaving this untargeted for now
Status: REOPENED → ASSIGNED
Keywords: helpwanted
Priority: P1 → --
Target Milestone: mozilla0.9.8 → ---
christian, jeeva: please see comment #17 regarding this bug.  thx!
Ahh, WORKSFORME on build 2002020103, Win2K. Thanks for doing whatever it was
that fixed it.
I still see this bug on a latest build.

Mozilla 0.9.8+
Mozilla/5.0 (Windows; U; WinNT4.0; en-US; rv:0.9.8+) Gecko/20020128


jeeva: can you try a february build?  and, if you can repro the problem can you
please try generating a log file following the instructions in comment #17.  thx!
Attached file Log file — 
I killed mozill because the brower hangs, when I click send button.. However it
lets me to go back.
jeeva: thanks for supplying the http log.  was this captured using the build
from comment #20?
Nope..

Mozilla 0.9.8+
Mozilla/5.0 (Windows; U; WinNT4.0; en-US; rv:0.9.8+) Gecko/20020202

Is this the one you want?

BTW, I have not configure mozilla as my default web brower, does that have 
something to do with this bug? 



jeeva: yeah, that's the build info i wanted.. thx!  and no it shouldn't matter
that mozilla is not setup as your default browser.

investigating for mozilla 0.9.9
Priority: -- → P1
Target Milestone: --- → mozilla0.9.9
from the log file it appears that it is a long time before the server responds.
 however, when we read from the server, we reach end-of-file immediately.  this
results in the POST request being automatically resent.  however, it runs into
the same problem... the response from the server is delayed.  i'm not sure what
is causing this problem yet.

can someone better describe the hang they are seeing?  does the browser's user
interface freeze up?  or is it that the message is simply never sent successfully?

does it matter how large the message is?  is this bug 100% reproducible with the
2002-02-02 build?

for a while i thought that i was able to reproduce this bug, but now i realize
that i was seeing something different.
when I was able to reproduce it, the progressmeter, statusbar etc. were just
stuck and nothing was happening.  The chrome didn't actually freeze.
darin>> can someone better describe the hang they are seeing?

Simply the message is never sent. I am able to stop or go back on the same 
window, so all the other events are working except sending the messages.

I tried with small and relatively big message size all of them have the same 
problem.

It cannot be a problem with the server  because it works with IE.

I did more investigation by sending an email from hotmail. It too has the same 
problem.. In the status bar I could see "Transfering Data from ....." message, 
but the email was never sent out. However for yahoo mail I see "Document: Done 
(...secs)" in the status bar. My wild guess is "the send logic falls into 
infinite loop, however it is able to handle other events". (I have no idea 
about the code)


jeeva: thanks for the info... it does help a lot!  of course, if you are seeing
the problem 100% of the time, then you'd think i would be able to reproduce it.

can you try this: can you try creating a new profile?  (run mozilla with the
-ProfileManager command line argument.)  perhaps something has corrupted your
profile?!?  wishful thinking probably, but it would be a good thing to test.

thx!
I created a new profile and tried but of no use...

If you want I can caputer the network traffic of both mozilla and IE and send it
to you.
jeeva: yes please!!
*** Bug 117363 has been marked as a duplicate of this bug. ***

*** This bug has been marked as a duplicate of 117363 ***
Status: ASSIGNED → RESOLVED
Closed: 23 years ago → 23 years ago
Resolution: --- → DUPLICATE
Attached is the network traffic between yahoo and mozilla.. After waiting for
mins (around 4-5 mins) I pressed "ESC" and that also generated some traffic.

Check the IE caputure in the next bug report...
Pls. compare Attachement 68318 (traffic generated thru. mozilla)

It looks like the second http msg ("continuation") from the mozilla client is
delayed a little bit and the server (yahoo) ACK before that msg. This "may be"
a culprit for the bug or this could because of the load in my system. :)
jeeva: thank you for the packet traces... it shows that mozilla is incorrectly
reporting a content-length 64 bytes greater than the actual data size it sends.
this is why the server is getting confused.
Darin >> how fast is your network connection

T1 line shared 15 ppl. 

How does the speed matters in the TCP connection? 

My guess is that the server is waiting for something (may be for the extra data
(+64) reported by the content-length). My colleague has mozilla (older build)
on both Linux and XP box, he never seen  any problem with it.. 

I captured his traffic and the content-length seem to match with the data
size... I am attaching it for your investigation
> How does the speed matters in the TCP connection? 

it effects packet sizes as well as the timing on mozilla's end.  mozilla is
multithreaded, and network speed could mask certain bugs in mozilla related to
passing data between threads, buffering network data, and polling for certain
events, etc.

jeeva: what is the date of the "older" build that doesn't exhibit this bug?  can
you help narror down the specific date at which this problem first occured? 
then i could look at all the code changes that happened during that interval.  thx!
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: