Closed
Bug 39219
Opened 25 years ago
Closed 25 years ago
Mozilla gets "NNTP-error No Such Article" when trying to open certain articles.
Categories
(MailNews Core :: Networking: NNTP, defect, P3)
Tracking
(Not tracked)
M16
People
(Reporter: carljohan, Assigned: sspitzer)
Details
(Keywords: regression)
From Bugzilla Helper:
User-Agent: Mozilla/4.7 [en] (X11; U; Linux 2.2.14 i686)
BuildID: 2000041811
When using Mozilla to read news, one persons posts were unreadable.
The nntp-server responded "430 No Such Article\r\n" even though
the article is on the server. Netscape 4.7 could read it so I
used ethereal to find out what Mozilla was trying to fetch.
Netscape 4.7 fetches articles by the article number but
Mozilla fetches them by the Message-ID header in the
post.
When using Netscape to fetch one of the persons post the
Message-ID of it was:
<1ead2vv.m0dewzx1mdbN%d98dan+news@dtek.chalmers.se>
When Mozilla tries to fecth the same article it tries
to fetch the article with the Message-ID:
<1ead2vv.m0bdewzx1mdbN\3318dan+news@dtek.chalmers.se>
Notice how the "%d9" has changed to the octal character
\331 (Hex 0xd9). So Mozilla tries to get the wrong article,
hence the NNTP-error
The Message-Id is correctly stored in the .msf for the
group.
It is just this guys posts, and all his posts, that can't be
read.
Reproducible: Always
Steps to Reproduce:
1. Log on to news.dd.chalmers.se (only people from Chalmers University
can do this)
2. Open the group cth.dtek.general .
3. Try to open a post from Dan Karlsson d98dan+news@dtek.chalmers.se
Actual Results: Mozilla responds: "NNTP error No Such Article"
Expected Results: Mozilla should show the article.
Assignee | ||
Comment 1•25 years ago
|
||
crap. this has happened a couple of times in the past.
I wonder if the url parsing code has changed again?
I'll investigate.
Assignee | ||
Comment 2•25 years ago
|
||
adding andreas to the cc list, this might be url parsing related.
Comment 3•25 years ago
|
||
Url parsing code has not changed recently, however some charset encoding before
urlparsing has changed. Move to UTF8, using unicode, ... I would start the
search there.
Assignee | ||
Updated•25 years ago
|
Assignee | ||
Comment 4•25 years ago
|
||
thanks for the info, andreas. I'll debug and see what's going on.
adding regression to the keyword, since message id's like that used to work.
carljohan@kjellander.com, which version of mozilla are you using?
I'm using Mozilla M15 BuildID 2000041811 (the one from mozilla.org) with
PSM 1.1.
Assignee | ||
Comment 6•25 years ago
|
||
this is fixed in m16. please test with m16 and let me know if you still have a
problem.
(FYI: official m16 builds have not been made yet.)
*** This bug has been marked as a duplicate of 35739 ***
Status: ASSIGNED → RESOLVED
Closed: 25 years ago
Resolution: --- → DUPLICATE
I've tried the latest M16 build and now it fetches
post by the article number instead of the Message-ID
so the problem is gone.
Hope the translation of "%d9" to 0xd9 hex doesn't
cause trouble elsewere.
Assignee | ||
Comment 9•25 years ago
|
||
try this, in a recent build:
open a browser window
type in that article url (news://host/message-id) in the location bar.
does it display the article in the browser window?
Reporter | ||
Comment 10•25 years ago
|
||
Nope! I checked with build 2000052420 and it still tries
to fetch the wrong article according to ethereal.
The %d9 gets translated to the hex 0xd9.
If you change the % to %25 (ascii '%') it tries to
get the correct article. But what is the syntax
for selecting the correct group?
Updated•20 years ago
|
Product: MailNews → Core
Updated•16 years ago
|
Product: Core → MailNews Core
You need to log in
before you can comment on or make changes to this bug.
Description
•