[Dovecot] zlib_save per-user or per-mailbox?
We'd like to be able to activate zlib_save per-user or per-mailbox, but it seems to be global, all or nothing. Search of this list revealed a comment from Timo in 2012:
http://www.dovecot.org/list/dovecot/2012-March/064909.html
where he was thinking that compression per-namespace would be a worthy feature. Was that done?
I'm in the process of replacing a 2.0 system with 2.2 EE. The old system had zlib_save activated, but seems to deliver to maildirs without compression. Does a userdb_mail entry override global zlib_save? (I'm about to test that, but duh, I am asking first.)
Thanks.
http://rob0.nodns4.us/ -- system administration and consulting Offlist GMX mail is seen only if "/dev/rob0" is in the Subject:
On Thu, Sep 19, 2013 at 06:14:32PM -0500, /dev/rob0 wrote:
We'd like to be able to activate zlib_save per-user or per-mailbox, but it seems to be global, all or nothing. Search of this list revealed a comment from Timo in 2012:
http://www.dovecot.org/list/dovecot/2012-March/064909.html
where he was thinking that compression per-namespace would be a worthy feature. Was that done?
I'm in the process of replacing a 2.0 system with 2.2 EE. The old system had zlib_save activated, but seems to deliver to maildirs without compression. Does a userdb_mail entry override global zlib_save? (I'm about to test that, but duh, I am asking first.)
The test user having a nonstandard userdb_mail "maildir:~/mail", new mail was delivered in compressed format.
http://rob0.nodns4.us/ -- system administration and consulting Offlist GMX mail is seen only if "/dev/rob0" is in the Subject:
On 20.9.2013, at 2.14, /dev/rob0 rob0@gmx.co.uk wrote:
We'd like to be able to activate zlib_save per-user or per-mailbox, but it seems to be global, all or nothing. Search of this list revealed a comment from Timo in 2012:
http://www.dovecot.org/list/dovecot/2012-March/064909.html
where he was thinking that compression per-namespace would be a worthy feature. Was that done?
No. It's a bit difficult to implement since it requires some major settings code changes, and the code is already horribly complex. Maybe some day when the whole thing is redesigned.
participants (2)
-
/dev/rob0
-
Timo Sirainen