It's nice to see reports that some large SMS services run on Gammu. I know there are more, but mostly such information is hidden and nobody except people who wrote that know about used software.
Make such reports more often, it makes people happy. Usually mailing list is just full of complains that something is not working and few good words can't make anything bad :-).
Well I expected this would be more complicated than defining DNS
blacklist servers list. Hower thanks to great defaults, all you need is
to put list into user defines (separate file which is included early like
/etc/exim4/conf.d/main/000_localdefs should be best place):
CHECK_RCPT_DOMAIN_DNSBLS = dnsbl.njabl.org : \ bl.spamcop.net : \ blackholes.five-ten-sg.com : \ cbl.abuseat.org : \ smtp.dnsbl.sorbs.net : \ spam.dnsbl.sorbs.net : \ zombie.dnsbl.sorbs.net : \ web.dnsbl.sorbs.net : \ rhsbl.sorbs.net : \ sbl-xbl.spamhaus.org : \ psbl.surriel.com : \ list.dsbl.org
phpMyAdmin started to be hosted on SourceForge long time ago (AFAIR it was in 2001). It has served mostly good for it's purpose, but these days, it's getting like nightmare.
First, at the end of March, CVS servers crashed and since that time anonymous CVS is disabled. Well at least that does not stop developing. Last week developer CVS server crashed. It has been disabled since then and we're now waiting for new CVS infrastructure, which should be ready several months ago (at least they promised so). This stopped phpMyAdmin development and also delayed deployment of some security fixes.
Okay, problems happen, but why the hell they didn't announce this on status page until some personal intervencies happened?
And I'm a bit confused about their "priority support" which is provided for subscribers. More than two days old report and no answer? I'm probably not going to pay next year...
I told this for many times before, so I just repeat: Such project needs version control system. No matter which one, but having at least CVS would make things lot of better than now. However Marcin seems unfortunately not to be interested in such "innovation". So I will still use my fixes branch, some people will want to use snapshots from that and we will sync with Marcin by sending patches and randomly forgetting about some parts (good luck that bazaar won't forget).
Why the hell do wakeup devices for ACPI need that cryptic names? I'd like to enable wakeup on keyboard, however I have no idea whether my notebook supports is or not. And the names don't help much:
$ cat /proc/acpi/wakeup Device Sleep state Status USB1 3 disabled USB2 3 disabled USB4 3 disabled LAN 4 disabled MPC0 3 disabled MPC1 3 disabled VIY0 3 disabled AMDM 3 disabled LID 4 * enabled PWRB 4 * enabled
It looks like XSS attacks against phpMyAdmin will be there forever. There is floating another in the air (if SourceForge would have working CVS, it would be already released). Maybe it's time to make complete rewrite of phpMyAdmin codebase? But that's too long term project and nobody seems to be interested in that. Maybe some security audit would help resolving at least most painful issues, but this is again same problem. Any volunteer to do this hard work?
However they have some ideas which we might incorporate. First think I
noticed in documentation is that they have settings for broken setups
of MySQL which lead to wrong displaying of data. Just put
names="latin1" and it should work. Nice, but I still think that
people should rather learn how to use encodings in MySQL properly.
Second usefull thing is reload link in left frame. Well they aren't able to reload it automatically, so they need it right now, but that simple link is requested by our users anyway.
Most of users noticed that there is not much progress in Wammu development. This is mostly caused by problems which wxPython already caused me and drops my motivatio to create something new using it to zero. However I want Wammu to have future, so it looks obvious, that it has to leave wxPython.
For now I'm decided to go either with PyGTK or using some high level framework as Kiwi. The latter gives advantage of better abstraction, but pure PyGTK could attract more developers. I'm not yet decided about this choice and I'm creating proposal for highly modular core, which should allow (somewhat, at least for beginning) idependent plugins to handle different data in phone (and later maybe even different backends).
If you would like to join me in developing wammu-ng (the name will probably change), you're welcome to drop me a mail on firstname.lastname@example.org.