Mailing List Archive

It is used to track changes, like a mailbox version. The new imap features use those. If you modify the state of a mailbox, you need to bump the version.

Andrea Brancatelli <abrancatelli@schema31.it> schreef op 6 februari 2015 09:24:48 CET:
>Can anyone spend a few words for me on the new "seq" column?
>
>What's it's main purpose?
>
>Is it functionally relevant or it something like a cache that can be
>reset, modified or whatever with no big issues?
>
>I'm used to move messages between mailboxes with queries like UPDATE
>dbmail_messages SET mailbox_idnr = 89 WHERE mailbox_idnr = 47981; but
>the seq column would be no more monotonically increasing in such case.
>Is it a problem?
>
>Futhermore in a dual master scenario the column doesn't seem really
>replication-aware, so I was curious if having duplicates in it may be
>problematic.
>
>Thanks.
>
>--
>*Andrea Brancatelli
>
>Schema31 S.p.a.
>Responsabile IT*
>
>ROMA - FIRENZE - PALERMO
>ITALY
>Tel: +39.06.98358472
>Cell: +39.331.2488468
>Fax: +39.055.71880466
>Società del gruppo SC31 ITALIA
>
>
>------------------------------------------------------------------------
>
>_______________________________________________
>DBmail mailing list
>DBmail@dbmail.org
>http://mailman.fastxs.nl/cgi-bin/mailman/listinfo/dbmail

--
Verzonden van mijn Android telefoon met K-9 Mail.

Am 06.02.2015 um 21:31 schrieb Paul J Stevens:
> It is used to track changes, like a mailbox version. The new imap
> features use those. If you modify the state of a mailbox, you need to
> bump the version

hm - if i get that correctly that means bump that number for every
move/delete action like cleanup trash folders with a' update table set
seq=seq+1 where mailboxid=y'?

good to know - we have webui actions to move mails between folders and
even users like 'move all messages from user1.inbox to user2.archive'
which needs in that case a update for dbmail 3.2

well, and the trash cleanup below too

$dbmail->log('Trash: Cleanup started');
$trash_folders = array
(
'TRASH',
'DELETED MESSAGES',
'DELETED ITEMS',
'DELETED MAILS',
'REMOVED MESSAGES',
'REMOVED ITEMS',
'REMOVED MAILS',
'PAPIERKORB',
'GEL&APY-SCHTE ELEMENTE',
);
$total_size = 0;
foreach($trash_folders as $cur_folder)
{
$result = $dbmail->db->query_fetch_all('select' . SQL_BIG_NO_CACHE .
' id,messagesize,internal_date from `dbmail_physmessage` where id in
(select physmessage_id from `dbmail_messages` where mailbox_idnr in
(select mailbox_idnr from `dbmail_mailboxes` where `name` like \'' .
$dbmail->db->addslashes($cur_folder) . '%\')) and internal_date <
date_sub(now(), interval 1 month);');
if(count($result) > 5)
{
foreach($result as $row)
{
$total_size += $row['messagesize'];
$dbmail->db->query('delete from dbmail_messages where
physmessage_id=' . (int)$row['id'], 1, 1);
}
echo utf8_encode(strtoupper(mb_convert_encoding($cur_folder,
'ISO-8859-1', 'UTF7-IMAP'))) . ': ' . count($result) . MY_LE;
}
}
if($total_size > 0)
{
echo MY_LE . 'SIZE: ' . GetSizeString('', $total_size) . MY_LE;
}

> Andrea Brancatelli <abrancatelli@schema31.it> schreef op 6 februari 2015
> 09:24:48 CET:
>
> Can anyone spend a few words for me on the new "seq" column?
>
> What's it's main purpose?
>
> Is it functionally relevant or it something like a cache that can be
> reset, modified or whatever with no big issues?
>
> I'm used to move messages between mailboxes with queries like UPDATE
> dbmail_messages SET mailbox_idnr = 89 WHERE mailbox_idnr = 47981;
> but the seq column would be no more monotonically increasing in such
> case. Is it a problem?
>
> Futhermore in a dual master scenario the column doesn't seem really
> replication-aware, so I was curious if having duplicates in it may
> be problematic