[Fail2ban-users] please test current master

[Yaroslav H. <li...@on...>]

Motivated by Tom's feedback/contribution on introducing <matches> I have
pushed it into the main fail2ban repository's master.

Since 0.8.5 there were few changes but as Tom points out it might be worth of a
new release but we need testing (I will upload current master into debian
unstable some time as well).  Besides some cruel typos and 'matches' handling
it should be as stable (or even much better in case of multiple jails) than
0.8.4 -- so get it while it is hot: https://github.com/fail2ban/fail2ban --
clone (somewhat easier and better to contribute back) or download.

So, brief summary of changes since 0.8.5: 

* major one: 3a58d0e  Lock server's executeCmd to prevent racing among iptables
  callIs

 many kudos go to Michael Saavedra for elegantly addressing long standing
 issue ;)

* few fixups in testing framework,
* added <matches> into actions
* modelines all around for emacs/vim to ease contribution work

outside of codebase:

* Yehuda helped a lot to bring sanity to wiki.

detailed log:

$> git lg 0.8.5..
* 22b7007 - (HEAD, origin/master, origin/_tent/matches_info, gh-yarikoptic/master, gh-yarikoptic/_tent/matches_info, master, _tent/matches_info) format output of matches data. (4 minutes ago) [Tom Hendrikx]
* 344effb - ENH: minor unittest to see if tickets carry correct 'matches' (4 weeks ago) [Yaroslav Halchenko]
* de8786d - ENH: introduced usa of Ticket.__matches throughout (4 weeks ago) [Yaroslav Halchenko]
* b52d420 - ENH: added 'matches' to the Ticket(s) and deprecated "custom" constructors for derived *Tickets (4 weeks ago) [Yaroslav Halchenko]
* ed6daa7 - ENH: modelines for emacs and vim to assure consistent indentation scheme (tabs) (4 weeks ago) [Yaroslav Halchenko]
* 19c9ac4 - ENH: failmanager -- additional debug message about # of known failures (4 weeks ago) [Yaroslav Halchenko]
* 1674b7b - ENH: rudimentary __str__ for the ticket (4 weeks ago) [Yaroslav Halchenko]
* 94aa94e - ENH: more human-accessible printout of the dates if any comparison fails (4 weeks ago) [Yaroslav Halchenko]
* 6641b1c - ENH: few debug messages and use MyTime.localtime instead of straight time.time (4 weeks ago) [Yaroslav Halchenko]
* b6d50c1 - ENH: Added localtime() to MyTime (4 weeks ago) [Yaroslav Halchenko]
* 5a2d518 - BF: set TZ to CEST while unittesting so dates matching would work (4 weeks ago) [Yaroslav Halchenko]
* 08fced9 - ENH: added a .pylintrc to help with consistent appearance and catch obvious problems (5 weeks ago) [Yaroslav Halchenko]
* ec4fda8 - Removed Subversion keyword (5 weeks ago) [Cyril Jaquier]
* 877de0d - adjusted the version in README to match the most recent "release" 0.8.5 (5 weeks ago) [Yaroslav Halchenko]
* 90167a1 - DOC: moved THANKS into a THANKS file for better visibility, concise README (5 weeks ago) [Yaroslav Halchenko]
* c321593 - Thank Yehuda for all his work on bringing sanity to wiki (5 weeks ago) [Yaroslav Halchenko]
* 3152afb - Recognise time-stamped kernel messages (6 weeks ago) [Adam Spiers]
* 3a58d0e - BF: Lock server's executeCmd to prevent racing among iptables calls (Closes: #554162) (6 weeks ago) [Yaroslav Halchenko]
* 3eb5e3b - BF: Allow for trailing spaces in sasl logs (3 months ago) [Yaroslav Halchenko]

Cheers!

On Tue, 08 Nov 2011, Yaroslav Halchenko wrote:

> Hi Tom,

> Thanks for giving it a shout!!! not sure why it didn't apply nicely on
> top of 0.8.4 but indeed it might have been a while ;)   for the patch --
> looks good to me ... I will either 'git am' it or feel free to send a
> pull request.... I am bit "under the rain" at the moment and will be at
> a conference till end of next week

> as for release -- 

> since there were no objections or further recommendations on having
> <matches> -- let's merge this branch (with our patch) in, let
> people drive it and if everything is splendid -- release it to be ;-)

> On Mon, 07 Nov 2011, Tom Hendrikx wrote:
> > > I just deployed this on a host, and it looks fine. The single nitpick I
> > > have is that the output is an unformatted python list:

> > > ['log line 1 including newline
> > > ', 'log line 2 including newline
> > > ', 'log line 3 including newline
> > > ']

> > > This is a bit ugly, so you might add some reformatting at the point
> > > where the final output is created: slap it into a single string, with
> > > single newlines between the entries.

> > update: attached two-liner works for me :)
-- 
=------------------------------------------------------------------=
Keep in touch                                     www.onerussian.com
Yaroslav Halchenko                 www.ohloh.net/accounts/yarikoptic