[Dovecot] questions/problems virtual plugin pop3 leave copies on the server for x days then delete

[Robert Schetterer]

Am 16.03.2011 18:18, schrieb Timo Sirainen:
> On Sun, 2011-03-13 at 12:40 +0100, Robert Schetterer wrote:
> 
>> RealMails/
>> RealMails/*
>> -RealMails/Trash
>> -RealMails/Trash/*
>> -RealMails/Sent
>> -RealMails/Sent/*
>> -RealMails/Templates
>> -RealMails/Templates/*
>> -RealMails/Drafts
>> -RealMails/Drafts/*
>> -RealMails/Archives
>> -RealMails/Archives/*
>>  unseen
> ..
>> that the mails stay in "new" folder of the maildir
>> after i downloaded via pop3
>> ( i am not sure is this expected ? , i would thought they get to cur   )
> 
> They get moved to cur/ when the \Recent flag gets cleared.
> 
>> thunderbird does not download this mails again ( good ), but it doesnt
>> delete them from maildir new ever( bad ) , so the whole leave copy and
>> delete after x days function pop3 is not working
> 
> What if that mailbox was an IMAP mailbox called "All mails" and user
> opened it, should that clear the \Recent flags from all the mailboxes?
> I'm not entirely sure what the right answer should be, but I'm thinking
> probably no.. Especially since such mailbox is used by some webmails to
> implement "search in all mailboxes".

nobody reported imap problems
if users use both pop3 ( with stay mails on the server ) and imap
mails get to cur wtih seen flag ( as far i know by small tests ) if they
connect via imap
the problem users mostly use outlook 2010/2007 no webmail etc
and my test client is thunderbird ( current stable )

> 
> So when should the \Recent flags be cleared with virtual mailboxes, and
> when not? Is it simply "clear with POP3, don't clear with others"? Then
> again others have wanted to create a virtual INBOX that shows full
> threads, i.e. messages also from Sent mailbox. That basically acts like
> an INBOX, so it should clear \Recent flags from INBOX.
> 
> Maybe this should be configurable. For example:
> 
> +a
> b
> +b/*
> 
> would keep \Recent flags in mailbox "b", but clear them from "a" and
> "b/*". Hmmh. Yeah, I guess that's the right solution. See if this works:
> 


> http://hg.dovecot.org/dovecot-2.0/rev/be0f0de75e34

i will give this a try and report



> 
> 


-- 
Best Regards

MfG Robert Schetterer

Germany/Munich/Bavaria