[go: nahoru, domu]

Open Bug 122674 Opened 23 years ago Updated 2 years ago

Problem with forwarding No-Mime message when Autodetect on

Categories

(MailNews Core :: Internationalization, defect)

x86
Windows 2000
defect

Tracking

(Not tracked)

People

(Reporter: marina, Unassigned)

Details

(Keywords: intl, Whiteboard: intl)

this problem is different than the one discribed in bug 122465. Bug # 122465
deals with nomime messages when the folder charset forces the correct display.
This one has no folder charset assigned.
To reproduce:
- set the Autodetect to ALL;
- in the newsgroup select a messge that has nomime info in the body;
- the messages with short body won't give enough info for Autodetect to sniff
the encoding ( you can see that the bullet will point to the Western);
- when such message is forwarded its body will be garbled and header will be blank;
- when the message body is long enough to sniff its encoding the forward won't
have any problem. By the way how we define the lenght of the message body being
enough to detect the encoding?
Summary: Forward of No-Mime message with Autodetect on → Problem with forwarding No-Mime message when Autodetect on
Currently, the auto-detect result is not used for forward, I think the same for
reply quote.
Status: NEW → ASSIGNED
Do we have any plans to use it? and what is the lenght of the string that would
be satisfactory for Autodetect to recognize the body encoding?
That is not planned. We might be able to set an override flag when we detect the
body charset, but that kind of change has to be done very carefully.
I don't know about the size but the body auto-detection is done. The problem is
that the result is not used for forward inline, so the input string size is not
the reason here.
Naoki, is that conceivable to make a feature request..?
I see the result by auto-detection is not check marked correctly in the menu.
For example, a no MIME JA messages is shown fine but the check mark is set to
ISO-8859-1.
Marina, do you see that? In your report,

>- the messages with short body won't give enough info for Autodetect to sniff
>the encoding ( you can see that the bullet will point to the Western);
The check mark is set to Western but was the message shown correctly before
forwarding?
yes, that the behavior i was seeing, though not in every case, in case of a long
enough message the bullet was pointing to the right encoding, that's why i asked
about the lenght of the body to be saficient to autodetect...
But if the detection was not done then we wouldn't be able to see them correctly.
though the body is displayng correctly and i am trying to find a common case
when the bullet is pointing to the wrong encoding ( i thoght it might be the
body length but it is not consistent)
QA Contact: ji → marina
Target Milestone: --- → mozilla1.2
Keywords: intl
Target Milestone: mozilla1.2alpha → ---
Product: MailNews → Core
Product: Core → MailNews Core
QA Contact: marina → i18n
Assignee: nhottanscp → nobody
Status: ASSIGNED → NEW
Whiteboard: intl
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.