Closed
Bug 229879
Opened 22 years ago
Closed 22 years ago
Thunderbird does not properly display attachments sent from a mac
Categories
(Thunderbird :: Mail Window Front End, defect)
Tracking
(Not tracked)
RESOLVED
FIXED
Thunderbird0.5
People
(Reporter: nb, Assigned: mscott)
References
Details
Attachments
(1 file)
766 bytes,
patch
|
Bienvenu
:
superreview+
|
Details | Diff | Splinter Review |
Dave sent me a file and all it shows in attachments is one file called %
He said that it should show that, along with another file, like it does when I
view the same message in Outlook Express.
Comment 1•22 years ago
|
||
The file I sent nb was encapsulated in AppleDouble format, which is explained
here: http://www.ietf.org/rfc/rfc1740.txt
Most non-Mac user agents which are not AppleDouble-aware will display both parts
as separate attached files, of which you just discard the one with the % at the
beginning of the filename. Non-Mac user agents which *are* AppleDouble-aware
will usually display only the data fork (i.e. the actual file) and automatically
discard the metadata/resource fork half for you. From what nb described to me
about what he got, I suspect Thunderbird is trying to be AppleDouble-aware and
discarding the wrong half.
Comment 2•22 years ago
|
||
Actually, I take that back... he did say it identified the mime type as
multipart/appledouble, which means it didn't take the pieces apart.
Comment 3•22 years ago
|
||
Problem in creation of multipart/appledouble when mail.file_attach_binary=true
seems to be resolved by Thunderbird 2003-12-10, but problem in disply does not
seem to be resolved yet.
This was reported in a Mozillazine forum.
See http://forums.mozillazine.org/viewtopic.php?t=39285&highlight=appledouble
Comment 4•22 years ago
|
||
Same as Mozilla Mail&News Bug 229855 ?
>check filename extension before Appledouble-encoding attachment having
insignificant resource fork
Comment 5•22 years ago
|
||
Bug 229855 was different problem - problem of AppleDouble attachment creation
when not necessary.
Sorry for spam.
Comment 6•22 years ago
|
||
Mozilla Mail issue is Bug 208196.
Assignee | ||
Comment 7•22 years ago
|
||
anyone have a message they can send me that demonstrates this problem?
i.e. I can try to read the message on my windows debug build and see us pick the
wrong fork.
If someone can get me a test case today, there is a chance a quick fix could
make the thunderbird 0.5 train.
Status: NEW → ASSIGNED
Target Milestone: --- → Thunderbird0.5
Comment 8•22 years ago
|
||
Sent.
Assignee | ||
Comment 9•22 years ago
|
||
for apple/double files, always extract the file name by looking at the second
part and if it does not have a file name, then look in the first part.
The second part is the data fork and the first part is the resource fork
Assignee | ||
Updated•22 years ago
|
Attachment #140759 -
Flags: superreview?(bienvenu)
Assignee | ||
Comment 10•22 years ago
|
||
one weird thing that is unrelated to this fix. When you try to open one of these
attachments (by double clicking it), the helper app dialog lists the content
type as apple double and does not know what to do with it. I thought we used to
extract the data fork automatically.
Updated•22 years ago
|
Attachment #140759 -
Flags: superreview?(bienvenu) → superreview+
Assignee | ||
Comment 11•22 years ago
|
||
fix checked into the M4 branch and the trunk.
Status: ASSIGNED → RESOLVED
Closed: 22 years ago
Resolution: --- → FIXED
Comment 12•22 years ago
|
||
Since resource folk(application/applefile part) is usually placed as first part
in appledouble, your logic is expected to work well.
However, MIME specification for appledouble deos not define part sequence of
resorce falk part and data folk part.
Therefore, I belive logic should be "part other than application/applefile is
data folk".
Scott, what do you think?
Assignee | ||
Comment 13•22 years ago
|
||
WADA, you are probably right, we should check the content type of each part and
use the part that is not the applefile part. Let's fix it that way on the trunk.
I'd rather not risk 0.5 for such a change anymore.
You need to log in
before you can comment on or make changes to this bug.
Description
•