| 00:21.30 | *** join/#asterisk-dev droemel (~droemel@p4FCACC5E.dip.t-dialin.net) |
| 00:54.42 | *** join/#asterisk-dev mjordan (~mjordan@user-69-1-6-49.knology.net) |
| 00:54.55 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 02:00.09 | *** join/#asterisk-dev elguero (~miguel323@2001:470:1f06:12c4::2) |
| 02:27.58 | Corydon76-home | pabelanger: ping |
| 02:31.47 | *** join/#asterisk-dev philipp64|laptop (~chatzilla@50-39-153-143.bvtn.or.frontiernet.net) |
| 04:11.47 | Kobaz | module load order is killer |
| 04:12.04 | Kobaz | anyone know of anyone working on a module dependency tree? |
| 04:12.54 | *** join/#asterisk-dev Fritz09 (~Adium@pop1-3802.catv.wtnet.de) |
| 06:02.27 | *** join/#asterisk-dev philipp64|laptop (~chatzilla@50-39-153-143.bvtn.or.frontiernet.net) |
| 06:40.31 | *** join/#asterisk-dev jkroon (~jkroon@dsl-241-232-76.telkomadsl.co.za) |
| 06:41.19 | jkroon | hi guys, i'd like to get some help on tracking https://issues.asterisk.org/jira/browse/ASTERISK-18453 |
| 06:56.03 | *** join/#asterisk-dev oej (~olle@static-213-115-251-100.sme.bredbandsbolaget.se) |
| 06:56.04 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 06:57.43 | oej | MOrning folks |
| 07:15.58 | jkroon | morning oej |
| 07:17.33 | mvanbaak | mornin oej |
| 07:18.20 | oej | So friends, what's the topic for this morning's discussions? |
| 07:18.57 | mvanbaak | zfs performance |
| 07:19.23 | mvanbaak | we now have 990MB/s write performance on our machine |
| 07:19.45 | mvanbaak | we hope the batterybacked SSD with SDRAM will speed it up |
| 07:19.53 | mvanbaak | we are going to use that for write cache |
| 07:21.30 | oej | Write cache for what? |
| 07:21.35 | mvanbaak | zfs pool |
| 07:21.56 | oej | Can you configure file system partitions as write cache for the overall file system |
| 07:22.24 | mvanbaak | no, you can add log and arc devices to your zpool for read and write cache for that specific pool |
| 07:23.31 | jkroon | ok, in an attempt to try and pinpoint the cause for https://issues.asterisk.org/jira/browse/ASTERISK-18453 I've now run things via valgrind --tool=drd to try and check for race conditions, and quite a number of lock issues cropped up, including in manager.c |
| 07:24.42 | jkroon | and a few in chan_sip.c (do_reload), which I suspect might be the cause of the original issue where chan_sip will just throw away a bunch of it's peers. |
| 07:26.16 | oej | Interesting - which version? |
| 07:27.46 | mvanbaak | 1200 MB/s \o/ |
| 07:27.51 | mvanbaak | 9K iops |
| 07:30.00 | oej | That's cool. What's the application for this, mvanbaak? |
| 07:30.56 | mvanbaak | storage |
| 07:30.59 | mvanbaak | for vps |
| 07:34.27 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 07:34.28 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 07:38.08 | oej | Hello Tzafrir! |
| 07:38.31 | tzafrir | hi |
| 07:45.16 | jkroon | oej, 1.8.5.0 |
| 07:46.13 | snuff-work | what's ur backing disks mvanbaak? |
| 07:46.34 | oej | 1.8 is out of scope for me? But I'm grateful that you work to fix the issues it has so it can become in-scope. |
| 07:46.39 | mvanbaak | snuff-work: 2 TB Seagate SAS disks |
| 07:47.09 | snuff-work | how many in what config? 5 in raidz? |
| 07:47.40 | mvanbaak | 3 raidz2 vdevs with 6 SAS disks each |
| 07:48.07 | mvanbaak | LSI sas controller |
| 07:48.18 | snuff-work | nice |
| 07:49.24 | snuff-work | i just have 2 thumpers at work (X4500 n X4540) they have 48x500 and 48x250 |
| 07:49.33 | snuff-work | i just have 2 thumpers at work (X4500 n X4540) they have 48x500 and 48x250 |
| 07:49.44 | snuff-work | stupid pidgin freezin |
| 07:53.25 | snuff-work | they do a nice job.. not as good as i'd like though.. unless i turn off sync writes.. |
| 07:53.59 | snuff-work | they have a intel x25-e for zil but they still get slowed down.. |
| 07:54.11 | oej | knows very little about storage |
| 07:54.21 | oej | I know when it fails though. |
| 07:56.34 | snuff-work | mm we all know that feeling |
| 07:57.15 | mvanbaak | snuff-work: we tested with an x25-e for zil |
| 07:57.42 | mvanbaak | but we now get one SSD with SDRAM on it |
| 07:57.46 | mvanbaak | dont know the specifics |
| 07:57.59 | snuff-work | mm.. i'm thinking of.. ddrdrive |
| 07:58.13 | snuff-work | that's sdram with flash backup |
| 07:58.26 | mvanbaak | yeah, that's we are going to use |
| 07:58.38 | mvanbaak | it also has a battery |
| 07:58.46 | snuff-work | they do very nice :) |
| 07:58.53 | mvanbaak | so if things go down, the sdram content will be written to the ssd |
| 07:58.56 | snuff-work | some of the initial blog posting etc were on that |
| 08:01.08 | mvanbaak | and we use some intel SSD as readcache |
| 08:02.43 | snuff-work | mm.. putting 250gb read cache in wouldn't hurt ;) |
| 08:03.01 | snuff-work | i found that our ram.. 32gb and 24gb.. keep most of the good stuff in l1 |
| 08:03.30 | snuff-work | so because our l2arc isn't big enough (17gb) it barely gets hit |
| 08:05.15 | snuff-work | getting fustrated @ the ast-test suite atm.. refuses to give me good error msgs |
| 08:18.42 | oej | Merging pinefrog into my 1.4 to put in production on over 500 servers... |
| 08:18.50 | mvanbaak | nice |
| 08:18.52 | oej | Time for some real action after over five million calls |
| 08:18.59 | oej | RTCP is really nice. |
| 08:19.12 | oej | Nice to have to monitor trunks and devices. |
| 08:19.12 | mvanbaak | good luck ! |
| 08:19.25 | oej | The only thing that can happen is crashes :-) |
| 08:19.31 | mvanbaak | you still run 1.4 in production ? |
| 08:21.00 | oej | That's the only version that works properly... |
| 08:21.10 | oej | 1.8 keeps crashing, locking, doing funny stuff |
| 08:21.12 | wdoekes2 | I don't blame him, I restart my 1.6.2 nightly, my 1.4's keep on running\ |
| 08:21.34 | oej | 1.6 was a nightmare and we should be ashamed. |
| 08:21.49 | oej | Asterisk lacks product management. |
| 08:22.21 | wdoekes2 | (oej, did you see my query?) |
| 08:22.35 | oej | Hmm. Where? |
| 08:22.38 | wdoekes2 | irc |
| 08:29.11 | *** join/#asterisk-dev oron_laptop (~oron_lapt@local.xorcom.com) |
| 08:45.32 | *** join/#asterisk-dev anny__ (~chatzilla@77.42.241.114) |
| 09:05.31 | *** join/#asterisk-dev droemel (~droemel@p4FCAD3EE.dip.t-dialin.net) |
| 09:05.53 | *** join/#asterisk-dev apten (~apten@carbon.gonicus.de) |
| 09:21.07 | wdoekes2 | according to "Reviewboard and commit policy change", trivial fixes don't need reviewing. I have two that need a simple commit: ASTERISK-18354 and ASTERISK-18355 |
| 09:32.28 | oej | wdoekes2: You got feedback on mailing list. Selecting transport is not easy. |
| 09:34.57 | wdoekes2 | drat. you are right in that it will break if someone has tcp as preferred setting |
| 09:35.48 | oej | Yep. |
| 09:36.04 | oej | I don't particularly agree with the current settings. They're awkward. |
| 09:40.52 | wdoekes2 | ergo, there is no way we can change the preferred transport without breaking it for someone. then we would be looking at a new sip.conf option (and trunk only.. which means this'll wait until asterisk 12) and proper (srv/naptr like you said) contact parsing. blegh |
| 09:48.38 | *** join/#asterisk-dev schmidts (~schmidts@Lo20-rt01.lm33.sil.at) |
| 09:48.42 | schmidts | hello |
| 10:20.03 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 10:20.03 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 10:46.45 | oej | Good morning, kpfleming. Still raining? |
| 10:47.22 | kpfleming | nope, it stopped on tuesday afternoon, finally. |
| 10:51.37 | *** part/#asterisk-dev mjordan (~mjordan@user-69-1-6-49.knology.net) |
| 11:16.41 | *** join/#asterisk-dev mjordan (~mjordan@nat/digium/x-eaamqxapfyrqyozj) |
| 11:16.41 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 11:24.52 | oej | wdoekes2: Well, session reuse is important, so if we have NAT support we might have to use the way you outlined and force reuse of the way they connected to us. |
| 11:24.57 | oej | But not for IPv6 :-) |
| 11:25.14 | oej | So there are quite a lot of parameters to get this right. Some of it a bugfix to 1.8 and others a new feature. |
| 11:28.44 | oej | Going off line to go home. |
| 11:47.23 | *** part/#asterisk-dev apten (~apten@carbon.gonicus.de) |
| 12:25.36 | *** join/#asterisk-dev oej (~olle@ns.webway.se) |
| 12:25.36 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:28.37 | *** join/#asterisk-dev luckman212 (~irc@2001:470:1f07:1225:ec3d:ac06:21a2:ec00) |
| 12:32.24 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 12:32.24 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 12:55.27 | oej | Afternoon folks. Any topic for discussion this lovely afternoon? |
| 13:07.41 | leifmadsen | oej: just that freenode is going to be up and down all day likely :) |
| 13:09.16 | *** join/#asterisk-dev russellb (~russellb@asterisk/contributor-and-cool-guy/russellb) |
| 13:09.16 | *** join/#asterisk-dev jsmith-away (~jsmith@fedora/jsmith) |
| 13:09.17 | *** mode/#asterisk-dev [+o russellb] by ChanServ |
| 13:09.17 | *** mode/#asterisk-dev [+o jsmith-away] by ChanServ |
| 13:09.18 | *** join/#asterisk-dev jdoe (jdoe@falseprophet.ca) |
| 13:09.18 | *** join/#asterisk-dev wdoekes2 (~walter@wjd.osso.nl) |
| 13:10.05 | *** join/#asterisk-dev moy (~moy@99.238.186.209) |
| 13:11.58 | oej | I hate being upside down. |
| 13:12.08 | oej | What's going on with good ol' freenode? |
| 13:12.32 | dwayne | oej, http://www.thelocal.se/36002/20110907/ |
| 13:13.06 | oej | Yes, it's a cool picture. I laughed a lot this morning when I noticed it in one of my feeds. Local elks are not that fun. |
| 13:18.04 | *** join/#asterisk-dev moy_ (~moy@173.239.155.74) |
| 13:20.22 | *** join/#asterisk-dev serafie (~erin@nat/digium/x-recgxcfgnbeylitv) |
| 13:21.45 | JonathanRose | Aw man, I left all those verb statements in... |
| 13:21.46 | JonathanRose | oi |
| 13:21.48 | leifmadsen | oej: freenode is doing ircd upgrades |
| 13:23.05 | JonathanRose | Oh well, at least it was just to trunk. |
| 13:25.55 | *** join/#asterisk-dev sruffell (~Adium@asterisk/the-kernel-guy/sruffell) |
| 13:25.55 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 13:26.38 | *** part/#asterisk-dev jmls (~Julian@host217-36-208-155.in-addr.btopenworld.com) |
| 13:29.36 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 13:29.36 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 13:36.44 | JonathanRose | pabelanger: weird verbiage removed. |
| 13:40.28 | kpfleming | lol |
| 13:44.10 | oej | Color is fun. Adds spice to our software. But maybe not in verbose... |
| 13:44.55 | russellb | that's what comment blocks are for |
| 13:45.03 | JonathanRose | I prefer for my frustrated ramblings that I add to test patches to stay out of the code base. |
| 13:45.44 | oej | We should deprecate overlap dialing. Where do we have the list of stuff to deprecate? Fax, overlap dialing, dtmf... |
| 13:45.50 | kpfleming | oej: yeah |
| 13:45.54 | kpfleming | enbloc dialing for the win |
| 13:46.01 | *** join/#asterisk-dev jeffg (~jeffg@pdpc/supporter/professional/jeffg) |
| 13:46.16 | oej | We have variable length phone numbers in Sweden. Imaging how fun that is... |
| 13:46.51 | kpfleming | fires up webex |
| 13:47.37 | oej | thank kpfleimng for the reminder |
| 13:48.34 | oej | http://vimeo.com/16175616 |
| 13:48.43 | *** join/#asterisk-dev anonymouz666 (~anonymouz@187-28-37-118.poolip.RJO.embratel.net.br) |
| 14:16.48 | *** part/#asterisk-dev pnlarsson (~niklas@fw1.gml.g.icnet.infracom.se) |
| 14:26.04 | anonymouz666 | I was reading an old post from 2008, that states chan_sip should be using thread pool pattern already is present in chan_iax2... something has changed since this post? |
| 14:26.36 | oej | I don |
| 14:26.43 | oej | I don't think that was ever completed |
| 14:26.46 | kpfleming | it has not |
| 14:26.48 | oej | File started with a branch |
| 14:27.13 | *** join/#asterisk-dev putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 14:27.13 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:27.38 | anonymouz666 | so, chan_iax could perform better than chan_sip specially on high load due that? |
| 14:28.03 | file | that's not a very good comparison to make |
| 14:28.06 | kpfleming | no, it won't, because chan_iax also has to handle all media. chan_sip does not handle media, media is handled by the rtp stack, which is already multithreaded. |
| 14:28.09 | oej | That's very hard to say? The channels are very different, so are the protocols |
| 14:28.39 | kpfleming | that was the big reason for chan_iax to become multithreaded, not so that it could more call setups and teardowns |
| 14:30.42 | russellb | i started down that road with chan_sip once, too |
| 14:30.54 | russellb | i started but just moving the scheduler into its own thread, leaving do_monitor how it was |
| 14:30.58 | oej | The road with no return. |
| 14:31.00 | russellb | and it crashed all over the place |
| 14:31.14 | oej | The question is if it's easier with separate sockets for TCP and UDP stuff? |
| 14:31.18 | russellb | the module is _huge_ and it makes so many assumptions about the threading model ... |
| 14:31.43 | russellb | well, it's already multi-threaded if you bring TCP into the picture ... |
| 14:31.44 | anonymouz666 | I am about to say that, in my setup, I've got better results using chan_iax. phoneSIP -> asterisk1 -> iax2 -> asterisk2 -> e1 then using sip in the middle |
| 14:32.02 | russellb | that's interesting... |
| 14:32.10 | kpfleming | anonymouz666: define 'better' |
| 14:33.37 | *** join/#asterisk-dev The_Boy_Wonder (~vossel@asterisk/batman-developer/dvossel) |
| 14:33.37 | *** mode/#asterisk-dev [+o The_Boy_Wonder] by ChanServ |
| 14:34.06 | anonymouz666 | kpfleming: I noticed that the load become a little bit lower |
| 14:34.27 | anonymouz666 | and with high load very strange things happened with stuck channels from one machine to another |
| 14:34.39 | kpfleming | that could easily be caused by the use of iax2 trunking, which uses fewer sockets to transfer multiple media streams between the same two boxes |
| 14:34.56 | anonymouz666 | lots of retransmissions |
| 14:35.00 | russellb | s/sockets/packets/ |
| 14:35.36 | kpfleming | russellb: both, actually :-) |
| 14:35.42 | russellb | well, true :-) |
| 14:35.54 | russellb | i was thinking about iax2 trunking vs iax2 without trunking ... |
| 14:36.03 | russellb | not IAX2 vs. SIP and RTP |
| 14:36.12 | kpfleming | sure |
| 14:38.25 | *** join/#asterisk-dev acidfoo (~nib@modemcable094.94-70-69.static.videotron.ca) |
| 14:50.02 | jkroon | Hi guys, busy running asterisk under valgrind and seeing some wack races with ast flags. |
| 14:50.52 | jkroon | looking at the macros it seems that there are read-read and read-write sequences, without any kind of locking, which in itself I don't think is a problem, but valgrind is complaining, and I think it can be avoided ... worth trying or am i wasting my time? |
| 14:51.58 | kpfleming | don't quite understand what you are saying. what 'flags' are you referring to? |
| 14:53.17 | jkroon | the ast_set_flag and ast_test_flag macros in include/asterisk/utils.h |
| 14:54.14 | jkroon | http://pastebin.com/hX2iX5nF |
| 14:54.15 | kpfleming | yes, because those macros test and flip bits, there will frequently be multiple reads and writes involved for the memory that holds the flags |
| 14:54.17 | jkroon | as example |
| 14:55.05 | kpfleming | i have never seen that sort of thing from valgrind before, so i have no clue how it is determining there is race. how can it know what regions of memory are protected by locks? |
| 14:55.56 | jkroon | --tool=helgrind |
| 14:56.07 | jkroon | looking for a bug in chan_sip losing a bunch of peers, just spotted that. |
| 14:56.20 | oej | How many do you got? |
| 14:56.28 | jkroon | two or three cases of that. |
| 14:56.37 | oej | How many peers? |
| 14:56.42 | jkroon | mostly i'm seeing invalid unlocks in manager.c though, already filed a bug on that one. |
| 14:56.57 | oej | Does that affect sip peers? |
| 14:56.58 | jkroon | seen some "possible races" in chan_sip, but still investigating. |
| 14:57.07 | oej | Ok, good luck then. |
| 14:57.36 | jkroon | no, chan_sip is the thing that originally got me to use valgrind, just spotted that as part of the valgrind output. |
| 14:59.11 | oej | Cool |
| 15:00.05 | jkroon | not too worried, if anything it's a false positive from a quick read of the code ... but i've learned from past experience to not ignore anything that valgrinds spits at me. |
| 15:03.47 | jkroon | i think the issue that valgrind moans about is that it sees sequences such as thread one reading, then writing back, and interleaved with that another thread doing the same thing, so t1-read, t2-read, t1-write type of thing. |
| 15:04.46 | jkroon | which based on the code is perfectly possible and would actually cause some ast_set/clear_flag calls to go missing. |
| 15:08.42 | kpfleming | except that the data structure that the flags are embedded in should also contain a lock, and the lock should be used to synchronize those accesses. if that's not happening, that's bad. |
| 15:10.02 | jkroon | kpfleming, the global asterisk_options is def not protected by any locks that I can see. |
| 15:10.27 | jkroon | *ast_options sorry |
| 15:11.23 | jkroon | and all the cases so far is on ast_options |
| 15:11.28 | russellb | but in reality they only get set on startup IIRC |
| 15:11.40 | jkroon | that's why i say not too worried. |
| 15:11.46 | jkroon | http://www.uls.co.za/valgrind/valgrind.helgrind.log for my full valgrind run. |
| 15:15.08 | *** join/#asterisk-dev infobot (~infobot@rikers.org) |
| 15:15.08 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Astricon 2011 in Denver, Colorado: AstriDevCon on October 24. Be there! |
| 15:17.06 | oej | I have too few screens to follow webex meeting, compile and follow e-mail at the same time. Need a fourth screen. |
| 15:23.34 | *** join/#asterisk-dev moy (~moy@CPE002719f00364-CM0026f3a1701d.cpe.net.cable.rogers.com) |
| 15:28.34 | *** join/#asterisk-dev eelcob (~eelco@87.249.110.67) |
| 15:32.28 | jkroon | kpfleming, ast_debug moans about that flag the most ... |
| 15:36.30 | russellb | tell it to hush |
| 15:38.31 | kpfleming | russellb: i just finished (finally) my OLF presentation |
| 15:38.41 | russellb | awesome |
| 15:38.49 | *** join/#asterisk-dev Fritz09 (~Adium@pop1-3802.catv.wtnet.de) |
| 15:38.57 | russellb | in plenty of time, too |
| 15:39.08 | kpfleming | still 36 hours after the deadline, though |
| 15:39.43 | russellb | most speakers edit slides up until the presentation, those deadlines are lies, i tell you |
| 15:40.11 | file | it's TRUE |
| 15:46.21 | jeffg | russellb, i think most speakers START their slides on the flight, actually |
| 15:47.08 | kpfleming | well, when flying from HSV to CMH, that doesn't provide a lot of time |
| 15:50.02 | leifmadsen | that just means you should have a short presentation :) |
| 15:50.42 | kpfleming | leifmadsen: i didn't use all the stuff you gave me (it was too long), but i did grab some stuff from the cookbook as well |
| 15:50.56 | leifmadsen | nice |
| 15:51.08 | leifmadsen | for some reason I do like 45 slides in 30 mins ;) |
| 15:51.13 | leifmadsen | (I know that breaks all the rules) |
| 15:51.33 | sruffell | lol |
| 15:52.03 | sruffell | I'll play the upright piano at leif's next presentation |
| 16:23.24 | oej | Aren't flights purely for writing presentations? |
| 16:27.51 | kpfleming | if anyone has not yet purchased their airline tickets for astricon, it appears fares just dropped about 25% in the past 48 hours |
| 16:31.06 | oej | is there a message hidden in that fact? |
| 16:31.11 | oej | Good negotiation by Digium? |
| 16:31.16 | kpfleming | i wish |
| 16:31.20 | kpfleming | we all just bought our tickets over the weekend |
| 16:33.54 | leifmadsen | ugh, I just bought mine on Tuesday! |
| 16:35.01 | Corydon76-home | kpfleming: probably, you just made the flight hit its profit mark, so remaining tickets are all pure profit |
| 16:35.16 | kpfleming | Corydon76-home: the airfare reduction affected all flights, not just the ones i chose |
| 16:35.35 | oej | Good for you, otherwise I would be really paranoid |
| 16:35.40 | Corydon76-home | Oh, okay. ;-) |
| 16:35.45 | oej | "Oh we have a kpfleming on the flight, sale time". |
| 16:36.44 | Corydon76-home | "Oh, we have that expected corporation to pay whatever rate on the plane; now to target the individuals." |
| 17:02.54 | Kobaz | i'm driving out |
| 17:03.07 | Kobaz | no groping and xrays for me |
| 17:07.03 | *** join/#asterisk-dev irroot (~irroot@41.54.128.72) |
| 17:10.36 | kpfleming | Kobaz: i'm sure those could be arranged for you |
| 17:11.20 | irroot | lo folks from umhlangla rocks |
| 17:13.42 | Kobaz | heh |
| 17:20.01 | *** join/#asterisk-dev DocAwesome (~DocAwesom@asterisk/documenteur-extraordinaire/blitzrage) |
| 17:20.01 | *** mode/#asterisk-dev [+o DocAwesome] by ChanServ |
| 17:20.41 | irroot | see there was no reoslution to iLBC in 1.8.7-rc |
| 17:20.47 | kpfleming | nope |
| 17:20.50 | kpfleming | still working on it |
| 17:21.02 | leifmadsen | coughs out 1.8.7.0-rc1 |
| 17:21.30 | leifmadsen | irroot: ya figured we can get it in before we do full 1.8.7.0 release, but no point in holding up testing since it's not a key module for many people |
| 17:21.45 | irroot | +1 |
| 17:22.20 | irroot | leifmadsen rather cough it up than express it via some other oriface |
| 17:26.30 | *** join/#asterisk-dev DocAwesome (~DocAwesom@asterisk/documenteur-extraordinaire/blitzrage) |
| 17:26.30 | *** mode/#asterisk-dev [+o DocAwesome] by ChanServ |
| 17:29.29 | pabelanger | leifmadsen: indeed. I just check, the same flights now are $25 cheaper |
| 17:29.58 | kpfleming | well, that's not much for yours. mine dropped from $424 to $314. |
| 17:30.28 | *** join/#asterisk-dev irroot (~irroot@41.54.128.72) |
| 17:30.37 | file | mine dropped by ... a lot |
| 17:30.52 | kpfleming | it may be worth rebooking then |
| 17:31.25 | file | it would end up costing more |
| 17:31.31 | kpfleming | yeah... mine too |
| 17:32.03 | leifmadsen | ya I can't cancel my flight without losing the whole amount |
| 17:32.31 | kpfleming | that seems unnecessarily harsh. |
| 17:32.42 | leifmadsen | agred |
| 17:32.54 | leifmadsen | I could have booked it at a refundable rate, but it was like $1200 |
| 17:33.19 | file | leifmadsen, eh? |
| 17:33.21 | kpfleming | well, non-refundable doesn't mean you lose it... it just means you can only use the value of the ticket towards another ticket within 12 months, after paying a huge change fee. |
| 17:33.26 | file | yeah |
| 17:33.40 | leifmadsen | oh gotcha |
| 18:00.51 | leifmadsen | russellb: I keep forgetting you're in here :) |
| 18:01.03 | russellb | i see how it is! |
| 18:01.04 | russellb | JK. |
| 18:01.25 | leifmadsen | basically I'm seeing if there is something obvious I can fix in ASTERISK-18137 |
| 18:01.29 | leifmadsen | because I'm seeing the same thing |
| 18:01.34 | DocAwesome | [ASTERISK-18137] [Status: Triage] Configurations saved in ~/.asterisk.makeopts ignored - https://issues.asterisk.org/jira/browse/ASTERISK-18137 |
| 18:02.49 | russellb | 'tis a bug looks like |
| 18:03.12 | leifmadsen | indeed |
| 18:03.21 | leifmadsen | trying to figure out where it is |
| 18:03.28 | leifmadsen | is learning menuselect it looks like... |
| 18:03.30 | russellb | it'll take menuselect changes |
| 18:03.30 | kpfleming | leifmadsen: right there ^^^^ |
| 18:03.40 | leifmadsen | kpfleming: oh gosh you're right! |
| 18:03.57 | russellb | i can't mess with it right now ... but maybe if you bribe me later |
| 18:04.29 | leifmadsen | ya no worries I wasn't looking for a fix but, I don't even know what :) |
| 18:04.39 | leifmadsen | in case you were like, "oh ya, go look at file XX" |
| 18:04.41 | russellb | sanity check? |
| 18:04.49 | russellb | it's a bug, it's not trivial |
| 18:05.08 | kpfleming | nothing that touches menuselect is trivial |
| 18:05.11 | irroot | russellb sanity check went for one they found none |
| 18:05.14 | russellb | kpfleming: +1 |
| 18:05.39 | russellb | that code got out of hand, oh well |
| 18:05.42 | irroot | multi platforms / builds .... |
| 18:05.49 | leifmadsen | russellb: thanks for the input! |
| 18:05.51 | russellb | np |
| 18:05.54 | irroot | its not good to mess with make |
| 18:05.59 | kpfleming | hey, it's better than what we had before, which was nothing |
| 18:06.23 | kpfleming | before menuselect, you got everything that was buildable, period. |
| 18:06.35 | russellb | and even some things that weren't |
| 18:06.50 | russellb | and not some things that were |
| 18:06.57 | Kobaz | hmm |
| 18:06.57 | russellb | (but that was solved by configure) |
| 18:07.07 | Kobaz | where's the main function that handles dialplan pattern matching |
| 18:07.24 | leifmadsen | pbx_config.so ? |
| 18:07.32 | russellb | ifneq ($(wildcard /usr/include/somelib.h),) YAY BUILD IT HEHE ^_^ endif |
| 18:07.34 | Corydon76-home | No, main/pbx.c |
| 18:07.43 | Kobaz | i wanna make a new function like, DIALPLAN_MATCH(${PATTERN},1XXX,2XXX,4XXX) |
| 18:07.50 | Kobaz | and it will return the index of the pattern matched |
| 18:07.59 | leifmadsen | neat |
| 18:08.12 | Kobaz | so my external apps can make use of asterisk matching |
| 18:08.12 | russellb | you could abuse pbx.h APIs to do that |
| 18:08.21 | Kobaz | and not have to reimplement the matching |
| 18:08.23 | russellb | create a temp context with 1 extension, then check canmatch() or whatever |
| 18:08.52 | Corydon76-home | Kobaz: without an underscore, those are all exact matches |
| 18:08.59 | Kobaz | er |
| 18:09.00 | Kobaz | right |
| 18:09.09 | Kobaz | ...assume i wrote it right |
| 18:09.12 | russellb | or you could expose the matching code directly, but that sounds like more work :-) |
| 18:10.38 | Kobaz | it would make having external ivr scripts via agi quite nice |
| 18:11.19 | Corydon76-home | Kobaz: but why? If you're doing it externally, why use POSIX regex? |
| 18:11.20 | russellb | i guess, but if you're writing in AGI, you have regex ... |
| 18:11.26 | russellb | Corydon76-home: junx |
| 18:11.26 | Corydon76-home | jinx |
| 18:11.27 | russellb | or jinx |
| 18:12.05 | Kobaz | yeah but, what if you want to keep patterns consistent |
| 18:12.15 | Kobaz | why have asterisk patterns in dialplan and regex in the ivr |
| 18:12.36 | Kobaz | i've came across situations previously where i've wanted to match on a pattern |
| 18:12.44 | Kobaz | but didn't want to have it in the dialplan as an exten |
| 18:12.46 | Corydon76-home | Because we don't support regex in the dialplan |
| 18:13.22 | russellb | i started adding regex to the dialplan once |
| 18:13.33 | russellb | and it was way too invaisve to pbx.c so I threw it away ... |
| 18:13.49 | Corydon76-home | I'd agree on creating a temporary context and using the match API... because if more than one pattern matches, you want the best match available |
| 18:14.34 | Kobaz | yeah, i want the whole best matching logic too |
| 18:14.37 | Corydon76-home | Steve Murphy completely revamped the pbx core a couple years ago. Even had his own tests to verify the improvement in speed |
| 18:14.42 | Kobaz | regex wont do that |
| 18:15.02 | Corydon76-home | Well, regex will, if you write your own core... |
| 18:15.38 | Corydon76-home | Asterisk pattern matching is far easier as a method of scoring than regex |
| 18:16.24 | Corydon76-home | You just have to remember two things. Length takes priority over particularity |
| 18:16.28 | Kobaz | wouldn't creating temp contexts and extens be a bit of a hit? |
| 18:16.48 | russellb | it's the best way to do it if you're going to test matches against multiple patterns |
| 18:16.51 | Corydon76-home | Creating them, no. Comparing them, yes. |
| 18:16.52 | kpfleming | not really. you have to store the data somehow. |
| 18:16.56 | Kobaz | you would need to hold a lock the dialplan, create stuff, check, delete stuff, unlock |
| 18:16.56 | russellb | otherwise it's going to be a ton of work |
| 18:17.15 | kpfleming | no, you can create contexts independently of the live dialplan. |
| 18:17.19 | Kobaz | ah okay |
| 18:18.27 | Corydon76-home | kpfleming: he can, but I don't think he can compare it using the standard APIs... |
| 18:18.33 | kpfleming | bummer |
| 18:19.03 | Corydon76-home | Has to be merged into the live dialplan due to the API not having those handles |
| 18:19.22 | russellb | seems like it wouldn't be hard to add if it's not possible yet ... |
| 18:19.36 | russellb | root = arg ? arg : live_dialplan_root; |
| 18:20.22 | Corydon76-home | Yep |
| 18:21.13 | Kobaz | are you saying to monkey up the matching functions to take an arg for which dialplan to check |
| 18:21.21 | Kobaz | so you can have a private one and not hold up the show |
| 18:21.33 | Corydon76-home | Right |
| 18:21.38 | Kobaz | that shouldn't be too bad |
| 18:21.55 | kpfleming | but don't use an actual monkey |
| 18:22.26 | Kobaz | good to know |
| 18:22.41 | file | unless you are a professional |
| 18:22.45 | kpfleming | in fact, we should add that to the coding guidelines. |
| 18:22.51 | Kobaz | what if it's a dedicated service monkey? |
| 18:22.57 | russellb | kpfleming: lol. |
| 18:23.12 | kpfleming | our coding guidelines currently lack humour. |
| 18:23.30 | Kobaz | "please do not let service monkeys have access to the code" |
| 18:23.31 | Corydon76-home | I think he meant 'monkeywrench' anyway |
| 18:23.40 | kpfleming | normally, we'd blame such things on the person who most recently left the department... |
| 18:23.58 | russellb | O.O |
| 18:27.37 | russellb | but today we'll blame file ? |
| 18:27.44 | file | nope |
| 18:28.21 | file | you could blame my dog |
| 18:28.22 | irroot | blame me ... |
| 18:28.51 | irroot | im on holiday not too bothered ... i did delete the props from trunk :P so im a candidate |
| 18:29.12 | kpfleming | file: it would be interesting if you brought your dog to astricon |
| 18:29.24 | file | she would go crazy |
| 18:38.29 | *** join/#asterisk-dev malcolmd (~malcolmd@pdpc/sponsor/digium/malcolmd) |
| 18:38.29 | *** mode/#asterisk-dev [+o malcolmd] by ChanServ |
| 18:45.26 | irroot | anyone want to +1 this move function VALID_EXTEN to DIALPLAN_EXISTS [from app_readexten.c to func_dialplan.c] |
| 18:45.39 | irroot | https://reviewboard.asterisk.org/r/1396/ |
| 18:53.21 | Kobaz | +1 |
| 18:54.50 | irroot | Kobaz i want to move it and mark it deprecated as they both do almost the same thing |
| 18:56.25 | *** join/#asterisk-dev luckman212 (~irc@2001:470:1f07:1225:ec3d:ac06:21a2:ec00) |
| 18:56.29 | leifmadsen | +1 |
| 19:00.11 | *** join/#asterisk-dev malcolmd (~malcolmd@pdpc/sponsor/digium/malcolmd) |
| 19:00.11 | *** mode/#asterisk-dev [+o malcolmd] by ChanServ |
| 19:51.30 | *** join/#asterisk-dev infobot (~infobot@rikers.org) |
| 19:51.31 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Astricon 2011 in Denver, Colorado: AstriDevCon on October 24. Be there! |
| 20:03.29 | *** join/#asterisk-dev serafie (~erin@nat/digium/x-onhewgxuoqacrgcg) |
| 20:04.21 | sruffell | leifmadsen: I cannot change the topic in #asterisk to update with the new DAHDI versionâ¦.not a channel operator. |
| 20:10.27 | *** join/#asterisk-dev serafie (~erin@nat/digium/x-wexzvidepkzgtpbo) |
| 20:13.13 | *** join/#asterisk-dev philipp64|laptop (~chatzilla@50-39-153-143.bvtn.or.frontiernet.net) |
| 20:24.18 | russellb | sruffell: i'll fix that |
| 20:27.55 | sruffell | thx |
| 20:30.48 | *** join/#asterisk-dev tzafrir (~tzafrir@212.179.75.202) |
| 20:30.48 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 20:32.24 | *** join/#asterisk-dev serafie (~erin@nat/digium/x-dgqftebbyneshgdn) |
| 20:38.31 | *** join/#asterisk-dev serafie (~erin@nat/digium/x-tcrqrqioqnxcwhrq) |
| 20:44.14 | *** join/#asterisk-dev pabelanger (~pabelange@asterisk/contributor-and-bug-marshal/pabelanger) |
| 20:44.14 | *** mode/#asterisk-dev [+o pabelanger] by ChanServ |
| 20:47.00 | *** join/#asterisk-dev pabelanger (~pabelange@asterisk/contributor-and-bug-marshal/pabelanger) |
| 20:47.00 | *** mode/#asterisk-dev [+o pabelanger] by ChanServ |
| 20:54.28 | *** join/#asterisk-dev malcolmd (~malcolmd@pdpc/sponsor/digium/malcolmd) |
| 20:54.28 | *** mode/#asterisk-dev [+o malcolmd] by ChanServ |
| 20:55.04 | *** join/#asterisk-dev malcolmd (~malcolmd@pdpc/sponsor/digium/malcolmd) |
| 20:55.04 | *** mode/#asterisk-dev [+o malcolmd] by ChanServ |
| 23:20.59 | *** part/#asterisk-dev seraphie (~erin@75.76.38.159) |
| 23:21.08 | *** join/#asterisk-dev seraphie (~erin@75.76.38.159) |
| 23:29.28 | *** join/#asterisk-dev n0tk (~n0tk@109.123.117.162) |