mailing list archives

meli community discussions

⚠️ if something does not work as intended when interracting with the mailing lists,
reach out Github mirror Gitea repo @epilys:matrix.org

[Imap-protocol] APPEND flags

E-mail headers
From: Bill Shannon <bill.shannon@sun.com>
To: imap-protocol@u.washington.edu
Date: Fri, 08 Jun 2018 12:34:36 -0000
Message-ID: 44049C49.9010607@sun.com permalink / raw / eml / mbox 
When using JavaMail to copy messages from one IMAP server to a different
IMAP server, JavaMail will use the APPEND command and will include the
flags that were present on the message in the source server.  What
should a server do if it gets flags in an APPEND command that it does
not support?  At least one server is returning BAD in such a case.

Assuming that's allowed behavior, how should a client react to such a
response?  Since APPEND can be applied to a mailbox that hasn't been
selected, the client may not know what flags are supported by that
mailbox.  Should the client SELECT the destination mailbox, get the
permanent flags, and perform an intersection between that set and the
set in the source message and use that set for another APPEND command?
Or should the client simply try the APPEND again with only the system
flags, or no flags?

Yes, I know, "it depends".  I'm interested in your advice.

Thanks.
Reply
E-mail headers
From: MRC@CAC.Washington.EDU
To: imap-protocol@localhost
Date: Fri, 08 Jun 2018 12:34:36 -0000
Message-ID: Pine.WNT.4.65.0602281211570.1004@Tomobiki-Cho.CAC.Washington.EDU permalink / raw / eml / mbox 
On Tue, 28 Feb 2006, Bill Shannon wrote:
> What
> should a server do if it gets flags in an APPEND command that it does
> not support?  At least one server is returning BAD in such a case.

A BAD is definitely wrong behavior.

A server should either:
  . create the flag(s) as needed
  . disregard the flag(s) and append anyway (perhaps issuing a warning in
    an untagged NO)
  . reject the command with a NO (not a BAD; there is no syntax error).

It may be reasonable for a STORE command to return a BAD for unknown flags 
if the server does not advertise \*.  But in the case of APPEND, the 
client has never been alerted with a FLAGS response as to what flags the 
mailbox offers (including \*) and thus BAD (which indicates syntax 
problems) is inappropriate.

Retry the command with only the predefined "system" flags.

> Should the client SELECT the destination mailbox, get the
> permanent flags, and perform an intersection between that set and the
> set in the source message and use that set for another APPEND command?

I think that is more work than is called for, all to deal with a broken 
server.

> Or should the client simply try the APPEND again with only the system
> flags, or no flags?

Retrying with just the system flags seems right.

And complain to the server vendor about its "BAD" behavior...

-- Mark --

http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.
Si vis pacem, para bellum.
Reply