T.A.Meyer at massey.ac.nz
Mon Jun 23 16:11:56 EDT 2003
> > This seems to fit the behaviour. I guess the correct
> > behaviour is to check for a PERMANENTFLAGS response before
> > appending the flags, and removing any flags that are in both
> > sets. The way to check this would
> > be to see if 'NotJunk' is indeed missing from a PERMANENTFLAGS
> > response (the -i4 switch is good for this).
> Doubtful that I'll get to it soon.
For the moment, I've checked in a change that means that imapfilter will
try to append without flags if they are causing a problem (a warning is
printed out). This does mean that all flags (seen, draft, etc) will be
lost. It would be interesting to know if this solves the problem.
The trouble with the PERMANENTFLAGS response checking is that it is
optional, not required, and usually sent with the folder information,
not along with the flag information, so it involves much more
In addition, if this flag is created/used by Mail.app, then I'm no
longer convinced that I have the source of the problem correctly
identified. I presume that Mail.app acts only as a client, and not a
server, which means that a client *should* be able to set this flag.
If anyone has both time and access to a server that has messages with
this 'NotJunk' flag, then they could play around and figure out what the
rules about it are.
More information about the Spambayes