-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
Hi,
I have added a pull request "Get channel binding binding information
for TLS connection" #44
https://github.com/pyca/pyopenssl/pull/44
If it's get accepted in pyOpenSSL it will be possible to implement
SCRAM-SHA-1-PLUS authentication mechanisms
https://python-nbxmpp.gajim.org/ticket/16
I have already a working SCRAM-SHA-1-PLUS implementation in
python-nbxmpp, but the code needs more cleanup. I have tested
authentication mechanisms using Prosody trunk version, which has an
implementation of SCRAM-SHA-1-PLUS on server side.
Fedor
-----BEGIN PGP SIGNATURE-----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=8J1H
-----END PGP SIGNATURE-----
Hello,
There is a problem with Gajim which, I suspect, is related to SQLite
file cache.db.
"Gajim isn't exactly solid when it comes to Window objects writing.
It takes it 10 minutes to load a roster with 250 entries while with
Swift it takes maybe 5 seconds." --written by an XMPP server admin.
It was reported to me that Disk I/O rate is at 200 K/s while occurs.
Liferea, a GTK+ Feed Reader, used to have a problem which manifested in
freezing UI while scanning feeds.
Since Liferea 1.8.3 (released on March 23rd 2012), the feed scanning
process has has improved performance by 60 times, due to lower Disk I/O
rate resulted by utilising sqlite3 3.7+ WAL journaling[0]. However, due
to the use of WAL journaling, a new problem[1] emerged which caused to
database file liferea.db-wal to grow to a ridiculous size, and later
fixed[2] by Sven Hartge in Liferea 1.8.6[3].
* I think that the file liferea.db-wal got radically big, especially,
in NFS file-systems, but that problem was fixed for any file-system.
[0]: http://sourceforge.net/mailarchive/message.php?msg_id=29029130
[1]: http://sourceforge.net/mailarchive/message.php?msg_id=29183881
[2]: http://sourceforge.net/mailarchive/message.php?msg_id=29412284
[3]: http://sourceforge.net/mailarchive/message.php?msg_id=29412285
Regards,
--GK.
--
Defending the constitution is expensive!
Help Lavabit.com by donating to the Lavabit Legal Defense Fund
http://lavabit.com/ <support(a)lavabit.com>