aox:aox.git
8 years agoDon't check lockSanity at startup v3.2.0
Abhijit Menon-Sen [Mon, 10 Mar 2014 13:44:59 +0000 (19:14 +0530)]
Don't check lockSanity at startup

We composed the query before the server_version was discovered, so the
test to decide whether to generate procpid or pid was always generating
procpid.

The point of the failing query was to warn users about locks at startup,
but the utility of the nice log message is outweighed by the price of
having to generate version-dependent queries.

(LockSpotter stays, because it runs much later.)

8 years agoGenerate 'pid' for Postgres >= 9.3 and 'procpid' otherwise
Abhijit Menon-Sen [Mon, 10 Mar 2014 10:24:34 +0000 (15:54 +0530)]
Generate 'pid' for Postgres >= 9.3 and 'procpid' otherwise

8 years agoMerge pull request #2 from felderado/patch-1
Abhijit Menon-Sen [Mon, 10 Mar 2014 04:20:02 +0000 (09:50 +0530)]
Merge pull request #2 from felderado/patch-1

fix schema.pg for new installs

8 years agoUpdate schema.pg
felderado [Mon, 10 Mar 2014 00:01:22 +0000 (19:01 -0500)]
Update schema.pg

schema.pg was upgraded to match what "aox upgrade schema" produces, but doesn't work on a clean install as aox will still believe it is at version 95 and it will bail if the citext extension already exists

8 years agodon't just-send-8 when the smarthost doesn't support EAI
Arnt Gulbrandsen [Fri, 21 Feb 2014 09:49:16 +0000 (10:49 +0100)]
don't just-send-8 when the smarthost doesn't support EAI

this would cause some display-names to be incorrectly squashed by the
smarthost

8 years agouse the new tsearch2 index on the subject field
Arnt Gulbrandsen [Wed, 12 Feb 2014 13:59:55 +0000 (14:59 +0100)]
use the new tsearch2 index on the subject field

8 years agospel correctly, even
Arnt Gulbrandsen [Wed, 12 Feb 2014 11:38:02 +0000 (12:38 +0100)]
spel correctly, even

8 years agospeed up searches on subject
Arnt Gulbrandsen [Wed, 12 Feb 2014 11:22:00 +0000 (12:22 +0100)]
speed up searches on subject

subject searches are fairly common, so this adds an index if the database
mode is mostly-reading or advanced-reading

8 years agolog out idle clients faster.
Arnt Gulbrandsen [Mon, 10 Feb 2014 10:36:41 +0000 (11:36 +0100)]
log out idle clients faster.

broken NAT (even at ISPs) is growing increasingly common, and aox can have
problems with very large numbers of idle/gone connections from mobile
clients. the client gets a new IP address from the network and reconnects,
but when aox sends a few bytes to the old connection(s), the broken NAT
gateway doesn't tell aox that the connection is gone. and so the number of
connections increases, and increases, and increases. eventually aox runs
into resource problems.

this will clean up the idle connections after 30/60 minutes. imap clients
are required to be active at least every 29 minutes, so this is tolerably
conservative.

8 years agothread better/correctly
Arnt Gulbrandsen [Mon, 10 Feb 2014 10:34:40 +0000 (11:34 +0100)]
thread better/correctly

8 years agostop fetching messages if a connection dies.
Arnt Gulbrandsen [Thu, 6 Feb 2014 09:01:20 +0000 (10:01 +0100)]
stop fetching messages if a connection dies.

previously, a client could initiate a big, big fetch and then go away,
and aox would do all of the work. not any more.

8 years agosimpler, more robust injection-time threading
Arnt Gulbrandsen [Thu, 6 Feb 2014 09:00:05 +0000 (10:00 +0100)]
simpler, more robust injection-time threading

this makes fewer threads than the old code, which didn't gather
all threads correctly. I hope this gathers all threads. we'll see.

8 years agoHave to use SELECT to call a function
Abhijit Menon-Sen [Thu, 6 Feb 2014 08:40:35 +0000 (14:10 +0530)]
Have to use SELECT to call a function

8 years agosend the citext lookup with fewer indirections.
Arnt Gulbrandsen [Wed, 15 Jan 2014 19:06:18 +0000 (20:06 +0100)]
send the citext lookup with fewer indirections.

8 years agoreject unknown addresses much faster
Arnt Gulbrandsen [Wed, 15 Jan 2014 12:07:56 +0000 (13:07 +0100)]
reject unknown addresses much faster

8 years agoavoid infinite loop. how evil.
Arnt Gulbrandsen [Wed, 15 Jan 2014 11:20:07 +0000 (12:20 +0100)]
avoid infinite loop. how evil.

8 years agoavoid pointing to Connection from Session
Arnt Gulbrandsen [Wed, 15 Jan 2014 11:18:31 +0000 (12:18 +0100)]
avoid pointing to Connection from Session

instead, remove the need. the only reason for that pointer was to
be able to close connections, so do that directly instead.

8 years agoC=D is buggy. leads to memory blowup. disable for now.
Arnt Gulbrandsen [Wed, 15 Jan 2014 11:11:39 +0000 (12:11 +0100)]
C=D is buggy. leads to memory blowup. disable for now.

8 years agolet some short-lived objects be collected
Arnt Gulbrandsen [Wed, 15 Jan 2014 10:57:42 +0000 (11:57 +0100)]
let some short-lived objects be collected

8 years agoavoid pointing from Mailbox/Flag to Session
Arnt Gulbrandsen [Tue, 14 Jan 2014 15:36:32 +0000 (16:36 +0100)]
avoid pointing from Mailbox/Flag to Session

instead, start at the Connection and look for relevant Session objects
whenever the Sessions need to be informed of changes.

8 years agoclose the most resource-hungry connection if we use too much memory
Arnt Gulbrandsen [Tue, 14 Jan 2014 12:03:53 +0000 (13:03 +0100)]
close the most resource-hungry connection if we use too much memory

8 years agoold message-id-less messages aren't in any db thread.
Arnt Gulbrandsen [Sat, 11 Jan 2014 14:26:07 +0000 (15:26 +0100)]
old message-id-less messages aren't in any db thread.

that shouldn't happen to new messages, but the old ones have to stay as
is. the thread_roots table is not friendly to id-less messages. so as of
this change, fetch x-gm-thrid reports 0 for such messages and no warning
is logged.

0 is a legal value according to google's spec (such as it is), and will
not conflict with any actual IDs. that all id-less messages are in the
same "thread" is... well... it could almost be seen as a feature.

8 years agoadd message-id even for inbound messages, so threading sucks less.
Arnt Gulbrandsen [Thu, 9 Jan 2014 21:43:23 +0000 (22:43 +0100)]
add message-id even for inbound messages, so threading sucks less.

in that case, the right-hand-side is one that clearly communicates what
I think about id-less messages (instead of Configuration::hostname()).

8 years agoavoid adding eternal objects for TlsThreads
Arnt Gulbrandsen [Thu, 9 Jan 2014 21:42:29 +0000 (22:42 +0100)]
avoid adding eternal objects for TlsThreads

instead, add a pointer from a Connection. cleaner, really.

8 years agotry harder to issue the citext lookup FIRST.
Arnt Gulbrandsen [Thu, 9 Jan 2014 21:41:18 +0000 (22:41 +0100)]
try harder to issue the citext lookup FIRST.

there were cases in which it might be the second query, and srb found
one in which the first query used citext. oops.

8 years agothread correctly even when a message-id exists several times in the db
Arnt Gulbrandsen [Wed, 8 Jan 2014 22:23:48 +0000 (23:23 +0100)]
thread correctly even when a message-id exists several times in the db

8 years agoadd support for gmail-like message and thread IDs.
Arnt Gulbrandsen [Wed, 8 Jan 2014 12:56:13 +0000 (13:56 +0100)]
add support for gmail-like message and thread IDs.

this uses a custom x-aox capability, and I may change it. the fetch/search
items are named as on gmail. the parser is a bit hacky... hm...

8 years agoA few syntax fixes
Abhijit Menon-Sen [Tue, 10 Dec 2013 08:22:57 +0000 (13:52 +0530)]
A few syntax fixes

8 years agoremove a few tabs and add braces around an else.
Arnt Gulbrandsen [Mon, 9 Dec 2013 13:49:07 +0000 (14:49 +0100)]
remove a few tabs and add braces around an else.

in aox either both parts of an if/else have braces, or neither.

8 years agobe compatible with what 'aox upgrade schema' makes.
Arnt Gulbrandsen [Mon, 9 Dec 2013 13:46:25 +0000 (14:46 +0100)]
be compatible with what 'aox upgrade schema' makes.

We can change ldn to dln, but it has to be done as a new schema,
so upgraded and newly installed databases agree.

8 years agoMerge srb's various patches
Arnt Gulbrandsen [Mon, 9 Dec 2013 13:45:40 +0000 (14:45 +0100)]
Merge srb's various patches

Conflicts:
aox/queue.cpp

8 years agoSuperfluous utf8 conversion omitted.
Stephen R. van den Berg [Fri, 6 Dec 2013 11:06:18 +0000 (12:06 +0100)]
Superfluous utf8 conversion omitted.

8 years agoCast citext to test in 'select ... from'
Arnt Gulbrandsen [Mon, 9 Dec 2013 11:37:49 +0000 (12:37 +0100)]
Cast citext to test in 'select ... from'

Detecting the OID of citext "at startup" wasn't always quite early enough,
so the second option is to make sure no citexts are returned.

This reverts commit a0ddbe03a28868fdbe3f4f447af60c94e383896a.

8 years agoRevert md5 anti-warning change, shut up gcc instead.
Stephen R. van den Berg [Sun, 1 Dec 2013 14:31:36 +0000 (15:31 +0100)]
Revert md5 anti-warning change, shut up gcc instead.

8 years agoFix multiple sieve responses.
Stephen R. van den Berg [Thu, 30 Jun 2011 10:51:48 +0000 (12:51 +0200)]
Fix multiple sieve responses.

8 years agoDelivery to multiple mailboxes now actually works.
Stephen R. van den Berg [Thu, 23 Jun 2011 19:11:06 +0000 (21:11 +0200)]
Delivery to multiple mailboxes now actually works.

8 years agoFix endless parsing loop
Stephen R. van den Berg [Tue, 7 Jun 2011 07:24:51 +0000 (09:24 +0200)]
Fix endless parsing loop

8 years agoAbort misbehaving overlapping POP3 sessions, detected by apparently empty messages.
Stephen R. van den Berg [Sun, 22 May 2011 19:58:16 +0000 (21:58 +0200)]
Abort misbehaving overlapping POP3 sessions, detected by apparently empty messages.

8 years agoEarly delete of deleted messages in POP3 session. Bugfix for duplicate and empty...
Stephen R. van den Berg [Sun, 22 May 2011 19:56:47 +0000 (21:56 +0200)]
Early delete of deleted messages in POP3 session.  Bugfix for duplicate and empty POP3 message retrieval.

8 years agoImproved POP3 logging.
Stephen R. van den Berg [Sun, 22 May 2011 19:55:03 +0000 (21:55 +0200)]
Improved POP3 logging.

8 years agoUse the DNS cache as a backup, in case DNS lookup (temporarily) fails.
Stephen R. van den Berg [Sun, 1 May 2011 23:04:24 +0000 (01:04 +0200)]
Use the DNS cache as a backup, in case DNS lookup (temporarily) fails.

8 years agoSimplify code and make sure it works with resolving Endpoint.
Stephen R. van den Berg [Sun, 1 May 2011 22:26:09 +0000 (00:26 +0200)]
Simplify code and make sure it works with resolving Endpoint.

8 years agoMake Endpoint use the Resolver when evaluating configuration items.
Stephen R. van den Berg [Sun, 1 May 2011 22:23:50 +0000 (00:23 +0200)]
Make Endpoint use the Resolver when evaluating configuration items.

8 years agoEliminate superfluous lower() calls.
Stephen R. van den Berg [Sun, 20 Mar 2011 17:21:20 +0000 (18:21 +0100)]
Eliminate superfluous lower() calls.

8 years agolet us we ignore the result of setreuid() if we want.
Arnt Gulbrandsen [Fri, 29 Nov 2013 09:09:15 +0000 (10:09 +0100)]
let us we ignore the result of setreuid() if we want.

in some places we check it, in some places we wouldn't do don't, because
we wouldn't anything differently based on the result.

8 years agobetter phrasing for the sql trigger to make aox act like gmail
Arnt Gulbrandsen [Wed, 27 Nov 2013 12:53:37 +0000 (13:53 +0100)]
better phrasing for the sql trigger to make aox act like gmail

8 years agoUpdate man page with correct version and date
Arnt Gulbrandsen [Wed, 27 Nov 2013 09:31:12 +0000 (10:31 +0100)]
Update man page with correct version and date

8 years agolearn the OID of citext at startup.
Arnt Gulbrandsen [Wed, 27 Nov 2013 09:11:15 +0000 (10:11 +0100)]
learn the OID of citext at startup.

until now, we used to assume the often-allocated value of 96087.

8 years agoWork around text/citext prepare limitation
Arnt Gulbrandsen [Tue, 26 Nov 2013 14:42:10 +0000 (15:42 +0100)]
Work around text/citext prepare limitation

Axel Rau had a message which 3.2.0 could not inject. I was able to
reproduce it, and together with Abhijit traced it down to a select which
succeeds on its own but not via prepare/execute.

For some reason, issuing "select * from addresses where domain='foo'"
is case-insensitive directly, but case-sensitive via prepare/execute.

This change forces insenstivity.

8 years agoit's 3.2.0. wow.
Arnt Gulbrandsen [Tue, 26 Nov 2013 10:03:33 +0000 (11:03 +0100)]
it's 3.2.0. wow.

8 years agolet us we ignore the result of setreuid() if we want.
Arnt Gulbrandsen [Fri, 29 Nov 2013 09:09:15 +0000 (10:09 +0100)]
let us we ignore the result of setreuid() if we want.

in some places we check it, in some places we wouldn't do don't, because
we wouldn't anything differently based on the result.

8 years agobetter phrasing for the sql trigger to make aox act like gmail
Arnt Gulbrandsen [Wed, 27 Nov 2013 12:53:37 +0000 (13:53 +0100)]
better phrasing for the sql trigger to make aox act like gmail

8 years agoUpdate man page with correct version and date
Arnt Gulbrandsen [Wed, 27 Nov 2013 09:31:12 +0000 (10:31 +0100)]
Update man page with correct version and date

8 years agolearn the OID of citext at startup.
Arnt Gulbrandsen [Wed, 27 Nov 2013 09:11:15 +0000 (10:11 +0100)]
learn the OID of citext at startup.

until now, we used to assume the often-allocated value of 96087.

8 years agoWork around text/citext prepare limitation
Arnt Gulbrandsen [Tue, 26 Nov 2013 14:42:10 +0000 (15:42 +0100)]
Work around text/citext prepare limitation

Axel Rau had a message which 3.2.0 could not inject. I was able to
reproduce it, and together with Abhijit traced it down to a select which
succeeds on its own but not via prepare/execute.

For some reason, issuing "select * from addresses where domain='foo'"
is case-insensitive directly, but case-sensitive via prepare/execute.

This change forces insenstivity.

8 years agoMerge branch 'master' of ssh://strange/home/git/aox
Abhijit Menon-Sen [Tue, 26 Nov 2013 11:27:53 +0000 (16:57 +0530)]
Merge branch 'master' of ssh://strange/home/git/aox

8 years agoCreate the citext extension in installer.cpp
Abhijit Menon-Sen [Tue, 26 Nov 2013 11:27:21 +0000 (16:57 +0530)]
Create the citext extension in installer.cpp

8 years agoit's 3.2.0. wow.
Arnt Gulbrandsen [Tue, 26 Nov 2013 10:03:33 +0000 (11:03 +0100)]
it's 3.2.0. wow.

8 years agoFix typo
Abhijit Menon-Sen [Mon, 25 Nov 2013 09:55:21 +0000 (15:25 +0530)]
Fix typo

8 years agoupgade schema was broken for 96->97. fix.
Arnt Gulbrandsen [Fri, 22 Nov 2013 15:58:34 +0000 (16:58 +0100)]
upgade schema was broken for 96->97. fix.

8 years agodon't compile views.cpp, it's gone
Arnt Gulbrandsen [Fri, 22 Nov 2013 15:46:28 +0000 (16:46 +0100)]
don't compile views.cpp, it's gone

8 years agoRemove the views table and all use of it
Arnt Gulbrandsen [Thu, 21 Nov 2013 11:30:28 +0000 (12:30 +0100)]
Remove the views table and all use of it

The views table and gmail-like views are incompatible. Views has to go
before gmail-like views can be written.

8 years agoremoved some notes that are never going to get done
Arnt Gulbrandsen [Wed, 20 Nov 2013 14:45:32 +0000 (15:45 +0100)]
removed some notes that are never going to get done

8 years agomore on gmail compatibility
Arnt Gulbrandsen [Wed, 20 Nov 2013 14:37:12 +0000 (15:37 +0100)]
more on gmail compatibility

8 years agoavoid failure if citext is present already
Arnt Gulbrandsen [Fri, 11 Oct 2013 08:40:15 +0000 (10:40 +0200)]
avoid failure if citext is present already

8 years agoTreat addresses properly case-insensitively when inserting many
Arnt Gulbrandsen [Sun, 6 Oct 2013 13:22:34 +0000 (13:22 +0000)]
Treat addresses properly case-insensitively when inserting many

8 years agoTreat localparts case-insensitively on injection
Arnt Gulbrandsen [Sun, 6 Oct 2013 12:47:52 +0000 (14:47 +0200)]
Treat localparts case-insensitively on injection

Now that we support unicode localparts, we have to do this. No longer can
we distinguish between Alex.Beta@example.com and alex.beta@example.com.

This change also removes an unnecessary call to lower().

8 years agoavoid extra forking of serverProcesses is one
Arnt Gulbrandsen [Wed, 2 Oct 2013 09:45:25 +0000 (11:45 +0200)]
avoid extra forking of serverProcesses is one

this helps debugging, important if your name happens to be arnt

8 years agoUse the schema with unicode-capable addresses
Arnt Gulbrandsen [Wed, 2 Oct 2013 09:41:49 +0000 (11:41 +0200)]
Use the schema with unicode-capable addresses

8 years agomake MimeField::rfc822 be an override again.
Arnt Gulbrandsen [Sun, 29 Sep 2013 09:31:05 +0000 (11:31 +0200)]
make MimeField::rfc822 be an override again.

when HeaderField::rfc822 change signature, I forgot to handle the
mimefield reimplementation.

8 years agoEAI support now compiles
Arnt Gulbrandsen [Mon, 16 Sep 2013 20:15:08 +0000 (22:15 +0200)]
EAI support now compiles

8 years agotesting against a->base is an off-by-one (sometimes)
Arnt Gulbrandsen [Tue, 10 Sep 2013 21:27:42 +0000 (23:27 +0200)]
testing against a->base is an off-by-one (sometimes)

8 years agofix address deduplication
Arnt Gulbrandsen [Mon, 9 Sep 2013 20:04:21 +0000 (22:04 +0200)]
fix address deduplication

this was broken:
  To: A <a@example.com>, a@example.com

8 years agovalid() is invariant in this loop and need not be called
Arnt Gulbrandsen [Mon, 9 Sep 2013 19:17:59 +0000 (21:17 +0200)]
valid() is invariant in this loop and need not be called

8 years agohandle addresses ins " æ <example@example.org>
Arnt Gulbrandsen [Mon, 9 Sep 2013 19:17:08 +0000 (21:17 +0200)]
handle addresses ins " æ <example@example.org>

those aren't terribly common, except in unit tests. eh.

8 years agospeling
Arnt Gulbrandsen [Mon, 9 Sep 2013 19:16:39 +0000 (21:16 +0200)]
speling

8 years agoChange to better/more modern cipher suite preferences.
Arnt Gulbrandsen [Fri, 30 Aug 2013 10:00:06 +0000 (12:00 +0200)]
Change to better/more modern cipher suite preferences.

The new list eliminates the medium-security cipher suites, since there's
wide support for high-security ciphers today.

The new list also prefers cipher suites with perfect forward security (ie.
ones that cannot be snooped) over other high-security suites. IMO Edward
Snowden has revealed that anything without PFS isn't really high security.

8 years agoreenable THREAD=*, since the bug that broke them if fixed
Arnt Gulbrandsen [Thu, 29 Aug 2013 19:56:40 +0000 (21:56 +0200)]
reenable THREAD=*, since the bug that broke them if fixed

8 years agosend fewer autoreplies by default
Arnt Gulbrandsen [Mon, 26 Aug 2013 21:47:12 +0000 (23:47 +0200)]
send fewer autoreplies by default

before, two messages inbound with different Subject would both receive an
autoreply, since the default handle depended on the subject. now, the
default handle is empty, so all messages from the same sender have the
same handle, and the number of messages to that recipient will be limited
more strictly by default.

8 years agodo not send autoreplies to *noreply* or mailer-daemon
Arnt Gulbrandsen [Mon, 26 Aug 2013 21:46:32 +0000 (23:46 +0200)]
do not send autoreplies to *noreply* or mailer-daemon

8 years agomake Address::localpart() and domain() return UString
Arnt Gulbrandsen [Mon, 29 Jul 2013 17:09:29 +0000 (19:09 +0200)]
make Address::localpart() and domain() return UString

update callers. This is certainly broken, since it depends on many
database columns to be of type citext, and it's a big enough change
that I expect it's broken for other reasons as well.

8 years agoadd UString::section()
Arnt Gulbrandsen [Mon, 29 Jul 2013 17:09:20 +0000 (19:09 +0200)]
add UString::section()

8 years agoadvertise EAI support for incoming mail
Arnt Gulbrandsen [Mon, 17 Jun 2013 20:19:11 +0000 (22:19 +0200)]
advertise EAI support for incoming mail

8 years agoAdd support for RFC6531
Arnt Gulbrandsen [Mon, 17 Jun 2013 20:07:16 +0000 (22:07 +0200)]
Add support for RFC6531

This is dead code so far, since there's other code that doesn't
support 6531.

8 years agocomplete 6855 input support
Arnt Gulbrandsen [Sun, 16 Jun 2013 09:58:39 +0000 (11:58 +0200)]
complete 6855 input support

still needs to downgrade on output

8 years agoTHREAD does not take a charset.
Arnt Gulbrandsen [Sun, 6 Oct 2013 08:40:40 +0000 (10:40 +0200)]
THREAD does not take a charset.

Pure bugfix.

8 years agocitext has type id 96087; accept that type ID properly
Arnt Gulbrandsen [Fri, 4 Oct 2013 20:29:02 +0000 (22:29 +0200)]
citext has type id 96087; accept that type ID properly

8 years agobegin to add support for using TLS as a client.
Arnt Gulbrandsen [Fri, 4 Oct 2013 19:38:58 +0000 (21:38 +0200)]
begin to add support for using TLS as a client.

aoximport-from-gmail needs that (but don't hold your breath).

8 years agoteach aox to compile correctly with clang (3.2)
Arnt Gulbrandsen [Sat, 14 Sep 2013 12:39:56 +0000 (14:39 +0200)]
teach aox to compile correctly with clang (3.2)

8 years agoexpression order bugfix.
Arnt Gulbrandsen [Fri, 13 Sep 2013 21:21:00 +0000 (23:21 +0200)]
expression order bugfix.

not serious; this would only lead to the wrong error messages being
returned for (some) syntax errors.

8 years agoClean up singlepart multipart mail submitted by kaiten mail
Arnt Gulbrandsen [Thu, 12 Sep 2013 13:30:11 +0000 (15:30 +0200)]
Clean up singlepart multipart mail submitted by kaiten mail

If kaiten now sends a message whose toplevel content-type is multipart,
but which contains only a single bodypart, the message will now be
fixified before being passed on.

8 years agoAlso handle downgrading from schema 96 to 95
Arnt Gulbrandsen [Mon, 5 Aug 2013 17:12:31 +0000 (19:12 +0200)]
Also handle downgrading from schema 96 to 95

8 years agoMerge branch 'master' of ssh://strange/home/git/aox
Abhijit Menon-Sen [Fri, 2 Aug 2013 08:44:00 +0000 (14:14 +0530)]
Merge branch 'master' of ssh://strange/home/git/aox

8 years agoRevision 96: convert addresses.{localpart,domain} to citext
Abhijit Menon-Sen [Fri, 2 Aug 2013 08:43:17 +0000 (14:13 +0530)]
Revision 96: convert addresses.{localpart,domain} to citext

(Only partly tested.)

8 years agoEliminate now-dead code (special cases for pg 8.1)
Arnt Gulbrandsen [Tue, 30 Jul 2013 08:34:55 +0000 (10:34 +0200)]
Eliminate now-dead code (special cases for pg 8.1)

8 years agorequire 9.1 at installation time too
Arnt Gulbrandsen [Tue, 30 Jul 2013 08:31:54 +0000 (10:31 +0200)]
require 9.1 at installation time too

8 years agoRequire PostgreSQL 9.1.x at least
Abhijit Menon-Sen [Tue, 30 Jul 2013 08:12:58 +0000 (13:42 +0530)]
Require PostgreSQL 9.1.x at least

We've avoided doing this for a long time, but the need to "CREATE
EXTENSION citext" finally tipped the scales. Anyway, by now 9.1.x
is everywhere, and even 9.2.x is widely used/packaged.

We can benefit greatly from 9.1 performance improvements in some
cases, and we can also drop some 8.x-specific things we do.

People still running 8.4 can use pg_upgrade to switch to 9.x; those
running older versions will need to pg_dump/restore their database.

8 years agoI like stable sort
Arnt Gulbrandsen [Tue, 16 Jul 2013 07:26:41 +0000 (09:26 +0200)]
I like stable sort

8 years agoMerge branch 'master' of ssh://strange.aox.org/home/git/aox
Arnt Gulbrandsen [Tue, 16 Jul 2013 07:19:13 +0000 (09:19 +0200)]
Merge branch 'master' of ssh://strange.aox.org/home/git/aox

8 years agoUpdate from internic (see jamfile)
Arnt Gulbrandsen [Tue, 16 Jul 2013 07:16:40 +0000 (09:16 +0200)]
Update from internic (see jamfile)