| 00:01.45 | seanbright | russellb: heh |
| 00:04.56 | russellb | seanbright: i'm on to your scheme |
| 00:05.31 | seanbright | i need no favors |
| 00:35.29 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 00:35.29 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 01:12.04 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 01:12.04 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 01:29.59 | *** join/#asterisk-dev moy (~moy@CPE002719f00364-CM0026f3a1701d.cpe.net.cable.rogers.com) |
| 02:05.07 | russellb | anticipates a question from seanbright |
| 02:05.19 | seanbright | ok, i fixed a bug |
| 02:05.32 | seanbright | M18598 |
| 02:05.32 | russellb | yay |
| 02:05.34 | MuffinMan | [acknowledged] [Asterisk] Core/HTTP 0018598: Manager Event Interface w/Digest authentication does not work! reported by ksn https://issues.asterisk.org/view.php?id=18598 |
| 02:05.50 | seanbright | which was assigned to you... sorry |
| 02:05.53 | russellb | oh, awesome |
| 02:06.04 | russellb | i noticed that when documenting it ... |
| 02:06.10 | russellb | i guess i thought i might fix it someday |
| 02:06.15 | russellb | obviously didn't happen |
| 02:06.21 | seanbright | ok, so it's fixed, but then i run into an issue |
| 02:06.23 | Kobaz | ssl ami would be pretty cool |
| 02:06.26 | Kobaz | thinking of ami |
| 02:06.29 | Kobaz | i have all the code too |
| 02:06.29 | russellb | Kobaz: it's already there |
| 02:06.31 | Kobaz | oh |
| 02:06.36 | Kobaz | never noticd |
| 02:06.41 | seanbright | do you have latest 1.8 checked out? |
| 02:06.42 | russellb | :-) |
| 02:06.44 | Kobaz | yeah |
| 02:06.44 | leifmadsen | Kobaz: well that was easy :) |
| 02:06.46 | russellb | seanbright: i can |
| 02:06.57 | Kobaz | i'm not using 1.8 yet really though |
| 02:07.01 | Kobaz | i need to find this really really bad bug |
| 02:07.05 | russellb | it was in 1.6.something too |
| 02:07.13 | Kobaz | it got introduced in early 1.6.2 |
| 02:07.28 | seanbright | well, basically, there is these security events that are built in now |
| 02:07.28 | Kobaz | i don't know what it is at all |
| 02:07.56 | Kobaz | but i have lots of local channels spinning in tight loops, and some moh buffer gets overwritten and eventually asterisk crashes |
| 02:07.59 | seanbright | main/manager.c report_req_not_allowed for example |
| 02:08.00 | Kobaz | that's all i know |
| 02:08.15 | russellb | valgrind? |
| 02:08.15 | Kobaz | ah, cool |
| 02:08.18 | Kobaz | yeah |
| 02:08.21 | Kobaz | i found one thing with valgrind |
| 02:08.23 | seanbright | ok |
| 02:08.25 | russellb | seanbright: looking |
| 02:08.28 | Kobaz | that's the thing i pasted to you before russellb |
| 02:08.37 | Kobaz | setting callerid does an invalid read |
| 02:08.40 | seanbright | which in turn calls mansession_encode_sin_local |
| 02:08.50 | seanbright | but tcptls_session is 0x0 |
| 02:08.57 | Kobaz | i'm gonna make a bug report out of it since i can't figure it out |
| 02:08.59 | seanbright | i guess because it's an http connection? |
| 02:09.04 | Kobaz | but i think my valgrind is weird |
| 02:09.17 | Kobaz | i see functions in the backtrace with names like T.1073 |
| 02:09.31 | Kobaz | the valgrind bakctrace looks kind of screwy |
| 02:09.33 | russellb | seanbright: yes |
| 02:10.13 | russellb | seanbright: but actually, it shouldn't be ... it looks like the http callback sets it |
| 02:10.27 | russellb | seanbright: look at the top of generic_http_callback |
| 02:10.28 | seanbright | hmm |
| 02:10.44 | Kobaz | why would valgrind show a function name of T.1073 |
| 02:10.50 | seanbright | ser could be NULL? |
| 02:11.19 | seanbright | i guess i could check that |
| 02:11.26 | russellb | printf debugging ftw i guess |
| 02:12.04 | russellb | Kobaz: no idea why valgrind would say that ... unless it's some stack corruption involved too, which valgrind can't really help with |
| 02:14.33 | seanbright | no generic_http_callback in the backtrace |
| 02:14.44 | seanbright | auth_http_callback is there however |
| 02:14.56 | seanbright | i'm wondering if my patch i committed earlier was fucked up |
| 02:15.21 | seanbright | to fix the double-error-message-when-not-permitted thing |
| 02:17.20 | russellb | seanbright: maybe update how mansession is decalred to set tcptls_session = ser ? |
| 02:17.36 | russellb | in auth_http_callback, just like generic_http_callback |
| 02:18.40 | seanbright | yeah |
| 02:18.43 | seanbright | doing that now |
| 02:18.47 | russellb | k |
| 02:19.22 | seanbright | bam |
| 02:19.27 | russellb | woot. |
| 02:19.27 | Kobaz | russellb: ah |
| 02:19.52 | russellb | Kobaz: does DONT_OPTIMIZE on or off change symptoms? |
| 02:23.24 | Kobaz | haven't checked, i'll try |
| 02:23.30 | Kobaz | i generally have it turned on |
| 02:23.56 | Kobaz | i have DONT_OPTIMIZE and DEBUG_THREADS for pretty much every setup |
| 02:25.53 | russellb | if DONT_OPTIMIZE changes it, that makes it sound like stack corruption |
| 02:26.43 | Kobaz | okay |
| 02:26.58 | leifmadsen | and how do you debug stack corruption? |
| 02:27.05 | russellb | carefully |
| 02:27.11 | russellb | and painfully |
| 02:27.23 | leifmadsen | I see, so at that point, it's not something I can ask someone to provide information for |
| 02:27.38 | leifmadsen | I wish mantis allowed me to create different statuses |
| 02:27.43 | leifmadsen | does jira allow that? |
| 02:27.48 | russellb | leifmadsen: it does, just not easily |
| 02:27.50 | leifmadsen | or to define workflows? |
| 02:28.02 | russellb | and yes, in JIRA workflow customization is extremely powerful |
| 02:28.03 | leifmadsen | russellb: right, it does if I change the code, database, and deploy new code.... :) |
| 02:28.10 | russellb | right. |
| 02:28.11 | leifmadsen | I think we should switch to jira |
| 02:28.17 | russellb | k |
| 02:28.19 | leifmadsen | yay! |
| 02:28.33 | russellb | but it's so powerful that it's complicated and we have to let Erin do it, heh |
| 02:28.50 | leifmadsen | even better! |
| 02:28.56 | leifmadsen | seraphie: ;) |
| 02:29.44 | russellb | JIRA mastah |
| 02:29.56 | leifmadsen | foe sheezy |
| 02:29.59 | russellb | we should have the test instance ready this week |
| 02:30.48 | leifmadsen | that is exciting |
| 02:31.01 | seanbright | russellb: thank you, sir. |
| 02:31.08 | russellb | thank _you_ |
| 02:31.30 | leifmadsen | thank everyone! |
| 02:31.33 | russellb | you're making me look bad |
| 02:31.37 | russellb | i need to go fix something |
| 02:31.50 | leifmadsen | this is how open source works |
| 02:32.05 | leifmadsen | a perpetual geek one-uppance |
| 02:34.55 | russellb | heh |
| 02:35.05 | russellb | i might just drink beer instead though |
| 02:36.14 | leifmadsen | I like your moxy |
| 02:36.18 | leifmadsen | I'm just going to go to bed |
| 02:36.30 | leifmadsen | probably read a couple of pages of In Cold Blood |
| 02:36.53 | leifmadsen | .aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaand out |
| 02:38.59 | *** join/#asterisk-dev Sipster_ (~Sipster@modemcable143.199-202-24.mc.videotron.ca) |
| 03:04.17 | *** join/#asterisk-dev sruffell (~Adium@asterisk/the-kernel-guy/sruffell) |
| 03:04.17 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 03:19.35 | seraphie | ohai. missed that there mention. |
| 03:25.36 | russellb | leif did it |
| 03:27.21 | seraphie | all his fault. |
| 03:27.58 | russellb | yup |
| 03:28.18 | seraphie | Bedtime for me! Night. :) |
| 03:28.33 | alecdavis | \-o-\ May the 4th be with you |
| 03:29.28 | seraphie | likes fifths better. |
| 03:29.36 | Freeaqingme | the day that will be remembered as the day digium got its connectivity |
| 03:29.49 | Freeaqingme | fyi, May 4th is the Dutch remembrance day |
| 03:29.55 | Freeaqingme | useless fact #234234 |
| 03:30.17 | alecdavis | M234234 |
| 03:30.18 | MuffinMan | Failed to retrieve issue page for 234234. Does it exist? |
| 03:32.40 | alecdavis | come on russell, give me a shipit on reviewboard 1185, then go to bed, and it will be commited by the morning :) |
| 03:33.50 | alecdavis | it's not a 1.4 or 1.6.2 issue. |
| 03:37.12 | russellb | alecdavis: it's probably fine, but i need to look at it again when i'm not so sleepy ... |
| 03:37.16 | russellb | is off for bed now |
| 03:37.17 | russellb | & |
| 04:31.53 | *** join/#asterisk-dev moy (~moy@CPE002719f00364-CM0026f3a1701d.cpe.net.cable.rogers.com) |
| 05:16.21 | *** join/#asterisk-dev moy (~moy@CPE002719f00364-CM0026f3a1701d.cpe.net.cable.rogers.com) |
| 05:46.05 | *** join/#asterisk-dev asterisk-learner (~chatzilla@77.42.241.114) |
| 06:19.08 | asterisk-learner | hello, anyone here? |
| 06:20.15 | *** part/#asterisk-dev asterisk-learner (~chatzilla@77.42.241.114) |
| 06:21.23 | *** join/#asterisk-dev asterisk-learner (~chatzilla@77.42.241.114) |
| 06:51.36 | *** join/#asterisk-dev schmidts (~schmidts@Lo20-rt01.lm33.sil.at) |
| 06:51.38 | schmidts | good morning |
| 06:58.35 | *** join/#asterisk-dev Corydon76-home (pink@c-69-137-80-31.hsd1.tn.comcast.net) |
| 06:58.35 | *** mode/#asterisk-dev [+o Corydon76-home] by ChanServ |
| 07:02.32 | *** join/#asterisk-dev tzafrir_laptop (~tzafrir@local.xorcom.com) |
| 07:02.32 | *** mode/#asterisk-dev [+o tzafrir_laptop] by ChanServ |
| 07:23.48 | schmidts | just a short question how can i add the -pg compiler option to asterisk? |
| 07:23.57 | schmidts | i have used this last year but didnt remeber how ;) |
| 07:48.28 | *** join/#asterisk-dev tamiel (~tamiel@213.30.183.226) |
| 07:57.21 | *** join/#asterisk-dev moy (~moy@CPE002719f00364-CM0026f3a1701d.cpe.net.cable.rogers.com) |
| 08:01.05 | schmidts | found it ;) ASTCFLAGS="-pg" |
| 08:07.57 | *** join/#asterisk-dev pnlarsson (~niklas@fw1.gml.g.icnet.infracom.se) |
| 08:31.49 | *** join/#asterisk-dev sdr_ (~sdr@teracomm-group.com) |
| 08:32.41 | sdr_ | what might be the reason for menu select not letting me enable chan_dahdi? i got dahid and libss7 installed |
| 08:32.54 | sdr_ | and they are detected according config.log |
| 08:33.58 | sdr_ | Depends on: res_smdi(M), dahdi(E), tonezone(E) Can use: pri(E), ss7(E) |
| 08:54.59 | tzafrir_laptop | dahdi is dahdi-linux, basically. tonezone is from dahdi-tools |
| 08:55.35 | tzafrir_laptop | pri is libpri. Though this is not a dependency, just an extra feature (as is libss7) |
| 08:55.41 | tzafrir_laptop | sdr_, ==^ |
| 08:55.55 | tzafrir_laptop | If you have them installed, re-run ./configure |
| 08:56.04 | sdr_ | hm |
| 08:56.23 | sdr_ | i don't wan libpri just libss7 |
| 08:56.32 | sdr_ | i do |
| 08:56.37 | tzafrir_laptop | Also: how do you have them installed? from source or from binary packages? |
| 08:57.05 | sdr_ | ./configure --with-dahdi=/space/dahdi-linux-2.4.1.2 --with-ss7=/space/libss7-git/ --without-pri |
| 08:57.16 | sdr_ | all is source |
| 08:57.59 | tzafrir_laptop | hmm.. this may require a few tweaks. See the notes in contrib/scripts/live_ast |
| 08:58.13 | tzafrir_laptop | You may need a symlink or two |
| 08:58.31 | tzafrir_laptop | libss7 should basically be the same as libpri, IIRC |
| 08:58.32 | sdr_ | aa |
| 08:59.10 | sdr_ | similar setup elsewhere works fine |
| 09:03.05 | sdr_ | omg! I'm missing dahdi-tools |
| 09:05.37 | sdr_ | solved! :) |
| 09:24.56 | schmidts | :( i think this XXX (http://www.asterisk.org/doxygen/trunk/astobj2_8c-source.html#l00646) is something we should take a look at |
| 09:26.15 | schmidts | i am looking at which ratio we use the ao2_callback and after start its like 70% wit OBJ_POINTER and 30% for traversal but after 5000 simulated calls its allready something like 50:50 |
| 10:06.35 | *** join/#asterisk-dev micols (~ident@rlogin.dk) |
| 11:17.24 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 11:17.24 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 11:55.42 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 11:55.42 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 11:57.51 | *** join/#asterisk-dev ccesario (~ccesario@187.75.139.188) |
| 12:06.19 | *** join/#asterisk-dev Jasnejac (kvirc@81.91.107.236) |
| 12:20.01 | sdr_ | how do I use the timing interface? I need to routinely check for blocked channels and try resetting them |
| 12:26.48 | sdr_ | What i need is a callback to be executed from time to time .. |
| 12:27.17 | kpfleming | that's not timing, that's the scheduler |
| 12:38.01 | sdr_ | yes .. and libss7 got one buildin but it's not exposed in 1.0.2 .. |
| 12:45.08 | sdr_ | it is exposed now in my tree ;) |
| 12:59.13 | *** join/#asterisk-dev Deeewayne (~dwayne@75-150-14-49-Atlanta.hfc.comcastbusiness.net) |
| 12:59.13 | *** mode/#asterisk-dev [+o Deeewayne] by ChanServ |
| 13:13.05 | *** join/#asterisk-dev elguero (~miguel323@174.141.211.66.inaddr.G4.NET) |
| 13:13.20 | *** join/#asterisk-dev jeffg (~jeffg@pdpc/supporter/professional/jeffg) |
| 13:37.36 | *** join/#asterisk-dev miguel3239 (~miguel323@174.141.211.66.inaddr.G4.NET) |
| 13:39.10 | *** join/#asterisk-dev fullstop (~fullstop@static-173-210-91-4.saucontech.com) |
| 13:40.06 | fullstop | Hi all. I am reviewing chan_sip.c from asterisk 1.6.2, trying to understand how the retransmit timer works. |
| 13:41.12 | fullstop | It looks like it will use 500ms, but there is a comment in sip.conf which suggests that it will use the RTT of qualify=XXXX if the peer has qualify on. |
| 13:42.46 | fullstop | found it in create_addr_from_peer |
| 13:46.16 | *** part/#asterisk-dev serafie (~erin@nat/digium/x-vrxlwcioldmkkrus) |
| 14:02.06 | kpfleming | fullstop: that is correct... with a lower bound of 100ms |
| 14:02.12 | kpfleming | this 'timer T1' from RFC 3261 |
| 14:06.48 | *** join/#asterisk-dev sgimeno (~chatzilla@163.117.206.10) |
| 14:13.10 | *** join/#asterisk-dev putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 14:13.10 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:15.17 | fullstop | kpfleming: Thanks. I needed to document that to get Asterisk certified for an ITSP. |
| 14:16.10 | seanbright | kpfleming: c language question for you |
| 14:16.15 | seanbright | when you have a moment |
| 14:16.23 | kpfleming | shoot! |
| 14:16.28 | seanbright | http://pastebin.com/raw.php?i=2u1G25cT |
| 14:16.52 | kpfleming | looks broken |
| 14:17.03 | seanbright | agreed |
| 14:17.05 | seanbright | but compiles and runs |
| 14:17.09 | kpfleming | of course |
| 14:17.15 | kpfleming | it just produces undefined behavior |
| 14:17.22 | seanbright | ok |
| 14:17.29 | fullstop | Don't do it. :) |
| 14:17.38 | seanbright | heh, i won't |
| 14:17.39 | kpfleming | C compilers will in general not stop you from indexing off either end of an array |
| 14:18.08 | fullstop | In fact, that was an old-school optimization technique |
| 14:18.23 | kpfleming | wait... i misread the for loop. that code looks correct. |
| 14:18.32 | seanbright | ah |
| 14:18.45 | seanbright | well the question goes to the const char label[] in the strucr |
| 14:18.45 | kpfleming | well... hmm. |
| 14:18.47 | seanbright | struct* |
| 14:18.47 | kpfleming | yeah |
| 14:18.50 | kpfleming | that's not clear |
| 14:19.18 | kpfleming | so using a non-specified-length array is a GCC extension, this isn't valid standard C |
| 14:19.48 | kpfleming | if i had to guess, i would think it would be finding the longest value in that field in all the array elements, and then making the array in each instance of the structure be that size |
| 14:19.49 | seanbright | so it would fail to compile an ansi strict mode |
| 14:19.53 | kpfleming | yes |
| 14:20.20 | kpfleming | but if this is for asterisk, that's fine |
| 14:20.50 | seanbright | but it still produces incorrect results |
| 14:20.54 | kpfleming | what i would do there is put in a debug statement that outputs sizeof(testing[0]), and also make the string lengths vary |
| 14:21.08 | kpfleming | define 'incorrect'? |
| 14:21.19 | seanbright | ok, let me back up |
| 14:21.29 | seanbright | this is analogous to "broken code" that i fixed last night |
| 14:21.34 | kpfleming | yes, i saw that |
| 14:21.52 | seanbright | i guess my question is... |
| 14:22.12 | seanbright | what is the difference between 'const char label[]' and 'const char *label' |
| 14:22.38 | kpfleming | in the first case, the array itself is a member of the structure. in the second case, a pointer is a member of the structure. |
| 14:22.39 | seanbright | the former is allocating an array of bytes (of unknown size) and the latter is allocating just a pointer? |
| 14:22.52 | kpfleming | allocating *space for*, yes |
| 14:24.12 | seanbright | it's the size of the array that is in question in the former case |
| 14:24.17 | kpfleming | right |
| 14:24.28 | kpfleming | and adding a debug statement and varying the string lengths will tell you that |
| 14:24.36 | seanbright | ah, i see |
| 14:24.54 | fullstop | the const would put them in their own section. sizeof(testing) would be interesting to see. |
| 14:24.57 | kpfleming | the size of each member of the structure array will be the same, since it has to be |
| 14:25.22 | kpfleming | no, if the array is part of the structure, the strings will be in the structures themselves |
| 14:25.53 | kpfleming | when it's a pointer in the structure, then the string it points to will be stored elsewhere |
| 14:27.07 | *** join/#asterisk-dev elguero (~miguel323@174.141.211.66.inaddr.G4.NET) |
| 14:28.17 | seanbright | http://pastebin.com/aB0iv19R |
| 14:28.38 | seanbright | output at the bottom |
| 14:28.54 | fullstop | I think const char label[] implies that it is a pointer |
| 14:29.12 | fullstop | and those pointers point to the literal |
| 14:29.52 | seanbright | but why is 'id' getting trampled on? |
| 14:30.04 | seanbright | i mean... the moral is: this is wrong |
| 14:30.08 | seanbright | but the results are interesting |
| 14:30.45 | kpfleming | no, 'const char label[]' is an array. there is no 'implication' in C, everything is explicit. |
| 14:31.35 | kpfleming | ok, well, that's pretty telling. |
| 14:31.37 | seanbright | but according to the sizeof it has size of 0 |
| 14:31.53 | kpfleming | sizeof(struct testing) is 4, which is only enough space for the int. |
| 14:32.04 | seanbright | right |
| 14:32.07 | kpfleming | so the character array has no storage, but the compiler is putting the string contents in anyway |
| 14:32.29 | kpfleming | so... this code is bogus, and the compiler should be complaining about the lack of space for the strings in the initializer list. |
| 14:32.45 | seanbright | i have a non-new version of gcc |
| 14:32.55 | seanbright | 4.4.3 |
| 14:32.57 | *** join/#asterisk-dev The_Boy_Wonder (~vossel@asterisk/batman-developer/dvossel) |
| 14:32.57 | *** mode/#asterisk-dev [+o The_Boy_Wonder] by ChanServ |
| 14:33.02 | seanbright | not bleeding edge or anything |
| 14:33.04 | kpfleming | that's pretty new |
| 14:33.07 | kpfleming | close enough, anyway |
| 14:33.28 | fullstop | test.c:6: warning: invalid use of structure with flexible array member |
| 14:33.28 | kpfleming | so this is a bug in this GCC extension... switching to 'const char *' will cure the problem |
| 14:33.36 | kpfleming | ah-ha! |
| 14:33.47 | *** part/#asterisk-dev asterisk-learner (~chatzilla@77.42.241.114) |
| 14:33.50 | fullstop | with -Wall and -pedantic |
| 14:33.53 | seanbright | ah |
| 14:34.20 | seanbright | test.c:5: warning: ISO C90 does not support flexible array members |
| 14:34.30 | kpfleming | yeah, but that's different |
| 14:34.35 | fullstop | sorry.. |
| 14:34.45 | seanbright | well good times |
| 14:34.46 | fullstop | I had -std=c99 as well |
| 14:34.50 | seanbright | ah |
| 14:34.59 | *** join/#asterisk-dev moy (~moy@CPE002719f00364-CM0026f3a1701d.cpe.net.cable.rogers.com) |
| 14:35.01 | fullstop | without it, I get 2 warnings. |
| 14:35.06 | kpfleming | well, does using just -Wall -Wextra produce any warnings? |
| 14:35.22 | fullstop | nope |
| 14:35.29 | fullstop | gcc 4.4.5 here |
| 14:35.35 | kpfleming | then that is a bug |
| 14:35.41 | kpfleming | there should be a diagnostic issued for this |
| 14:36.05 | kpfleming | but --std=c99 is not a bad thing... we might want to add it to the asterisk build |
| 14:36.44 | fullstop | Do you think that Clang (the compiler) has any legs? |
| 14:36.49 | kpfleming | yep |
| 14:37.41 | kpfleming | it's fast, produces good code, produces excellent diagnostics, and it's very standards compliant |
| 14:37.45 | fullstop | I hope so, because it feels like gcc is getting slower with each release. |
| 14:38.02 | kpfleming | i bet you haven't even tried gcc 4.6 with LTO yet... it's really slow |
| 14:38.06 | fullstop | have you built asterisk with it? |
| 14:38.15 | kpfleming | i have not... i've been on asterisk scf for quite some time now |
| 14:38.17 | fullstop | No, I've not gone past 4.4.5 |
| 14:38.35 | file | is used to slow builds now |
| 14:39.10 | leifmadsen | file: are sword fights in your future? |
| 14:39.23 | file | leifmadsen, always! |
| 14:40.09 | leifmadsen | heh |
| 14:40.13 | leifmadsen | oh xkcd <3 |
| 14:40.54 | fullstop | When Randall was a kid, he lived right down the street from where I work now. |
| 14:44.06 | fullstop | clang with seanbright's C test from earlier: http://pastebin.com/BJ01tDyt |
| 14:44.30 | kpfleming | yep. awesome diagnostics from clang. |
| 14:44.44 | fullstop | Yeah, it underlined the whole literal. |
| 14:47.50 | seanbright | nice |
| 14:49.34 | *** join/#asterisk-dev datachomper (~Adium@nat/digium/x-olfdfahsybamadvu) |
| 14:52.41 | *** join/#asterisk-dev citywok (~citywok@67-134-194-33.dia.static.qwest.net) |
| 14:55.42 | sdr_ | ok lads I got chan_dahdi + libss7 initialize correctly |
| 14:55.56 | sdr_ | should i submit patch or something? |
| 14:57.30 | sdr_ | i'm using stable dahdi, stable libss7 and svn asterisk from branch 1.6.2 |
| 14:58.07 | sdr_ | i've modified libss7 in the process |
| 15:05.01 | sdr_ | also fixed group unblocking of cics. Added group blocking answers to libss making group blocking requests usefull. this was already in trunk libss7 but it's too ahead for asterisk 1.6 |
| 15:08.30 | *** join/#asterisk-dev sruffell (~Adium@asterisk/the-kernel-guy/sruffell) |
| 15:08.30 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 15:33.46 | *** join/#asterisk-dev jeffg (~jeffg@pdpc/supporter/professional/jeffg) |
| 15:38.35 | *** join/#asterisk-dev moy (~moy@CPE002719f00364-CM0026f3a1701d.cpe.net.cable.rogers.com) |
| 16:04.00 | *** join/#asterisk-dev rgagnon (~rgagnon@rrcs-71-42-183-54.sw.biz.rr.com) |
| 16:04.29 | *** join/#asterisk-dev sgimeno (~chatzilla@163.117.206.10) |
| 16:04.58 | *** join/#asterisk-dev jeffg (~jeffg@pdpc/supporter/professional/jeffg) |
| 16:46.28 | *** join/#asterisk-dev elguero (~miguel323@17.140.211.66.inaddr.G4.NET) |
| 16:52.38 | *** join/#asterisk-dev mvanbaak (~michiel@asterisk/contributor-and-bug-marshal/mvanbaak) |
| 16:52.38 | *** mode/#asterisk-dev [+o mvanbaak] by ChanServ |
| 17:08.47 | *** join/#asterisk-dev elguero (~miguel323@17.140.211.66.inaddr.G4.NET) |
| 17:34.42 | seanbright | so |
| 17:34.59 | seanbright | you open a file with open and get back an fd |
| 17:35.06 | seanbright | open(), sorry |
| 17:35.16 | seanbright | and then pass that fd to fdopen and get back a FILE * |
| 17:36.04 | seanbright | is there any guarantee that an operation on the fd will be immediately visible to the FILE *? |
| 17:38.24 | kpfleming | no. you should not do any operations directly on the fd once you have a FILE * for it. |
| 17:39.10 | seanbright | ok, that's very telling |
| 17:39.12 | seanbright | thanks |
| 17:39.41 | kaii | i'm going to add a new feature to app_dial, which handles multiple interfaces as a unit. (when a single user has a single extension but multiple devices) with the dial option set, Dial() returns busy if the user is busy on one of his devices. also Dial() stops ringing the other devices, if the user forwards the call from one device. |
| 17:40.16 | kaii | what i need is a good name for that feature. dial option namespace is nearly full and i cant find a good name. |
| 17:40.29 | kaii | internally we call it "multicall" |
| 17:40.32 | kaii | any suggestions? |
| 17:41.01 | kaii | it is already working, but i would like to give it a good name before checkin in to my branch |
| 17:45.23 | *** join/#asterisk-dev cVsup (~cVsup@189.83.152.237) |
| 17:50.22 | leifmadsen | kaii: I think your options for letters is pretty limited, so I think you just pick one :) |
| 17:50.47 | *** join/#asterisk-dev malcolmd (~malcolmd@pdpc/sponsor/digium/malcolmd) |
| 17:50.47 | *** mode/#asterisk-dev [+o malcolmd] by ChanServ |
| 17:51.35 | leifmadsen | Maybe something like.... "Single busy" or something... ? |
| 17:51.53 | seanbright | you aren't limited to letters |
| 17:51.56 | seanbright | i would use * |
| 17:51.58 | seanbright | heh |
| 17:52.00 | leifmadsen | :) |
| 17:52.07 | leifmadsen | actually, b and B are free |
| 17:52.10 | seanbright | or maybe you are limited, i dunno |
| 17:52.57 | leifmadsen | I'd suggest B, and add the description, "When calling multiple devices, if any of the devices return a status of being busy, stop ringing all devices, and return a ${DIALSTATUS} of BUSY" |
| 17:53.00 | leifmadsen | something like that.... |
| 17:53.24 | leifmadsen | I'd describe it as something like, 'Stop on busy' |
| 17:53.27 | leifmadsen | dunno |
| 18:06.41 | *** join/#asterisk-dev joshaidan (~brianj@host-216-26-197-55.tbaytel.net) |
| 18:06.53 | *** join/#asterisk-dev guilhermebr (~Guilherme@200.103.96.98) |
| 18:11.00 | russellb | The_Boy_Wonder: what was the resolution to the func_jitterbuffer issue? |
| 18:12.53 | leifmadsen | russellb: make it more awesome |
| 18:13.05 | russellb | ah, ok. |
| 18:13.22 | The_Boy_Wonder | russellb: the simple answer is that i relaxed when a frame is allowed to be pulled from the jitterbuffer. It used to force retrieval until after the time the next frame was expected, the problem with this is that our timing is not always perfect so it would not always work as expected... now i loosened it up when we are allowing frames to be pulled from the jitterbuffer if the time they are requested is at least within the i |
| 18:13.23 | leifmadsen | actually he was describing some change I didn't understand and it didn't sound like The_Boy_Wonder totally understood either |
| 18:13.27 | leifmadsen | ya that |
| 18:13.32 | leifmadsen | The_Boy_Wonder: btw I was wrong, I don't have a meeting today |
| 18:13.35 | leifmadsen | have at the system |
| 18:14.53 | The_Boy_Wonder | so now, if the next frame is say 100 and we ask for it at 99, that's cool as long as it is with in 80-100 |
| 18:15.10 | The_Boy_Wonder | otherwise asking for it at 99 would fail and have to wait until later to retrieve it |
| 18:16.10 | russellb | what if it's at 101? |
| 18:16.22 | russellb | i.e., we ask late |
| 18:16.45 | The_Boy_Wonder | you get it, that used to be how it worked, you always had to ask either on time or later |
| 18:17.13 | russellb | so late is still acceptable |
| 18:17.18 | The_Boy_Wonder | yes |
| 18:17.22 | russellb | excellent, |
| 18:17.27 | russellb | makes sense to me |
| 18:17.28 | russellb | yay |
| 18:18.01 | leifmadsen | ya sounds right to me |
| 18:18.02 | The_Boy_Wonder | yeah, well ideally it should be in perfect intervals |
| 18:18.12 | The_Boy_Wonder | or at least really close intervals |
| 18:18.19 | leifmadsen | we should introduce some jitter and test it |
| 18:18.25 | leifmadsen | see if it actually does stuff :) |
| 18:18.31 | russellb | testing is overrate |
| 18:18.32 | russellb | d |
| 18:18.34 | russellb | so is documentation |
| 18:18.40 | The_Boy_Wonder | so, if it was 20ms intervals. i'd expect to see 19 or 21 actually happen sometimes |
| 18:18.44 | leifmadsen | I agree |
| 18:19.28 | The_Boy_Wonder | but that's not even the case, sometimes it is completely screwed up. like 10 - 30 ms intervals when it shoudl be 20ms. overall they all average out to 20ms |
| 18:19.44 | The_Boy_Wonder | but they don't all occur super close to 20ms like i'd expect |
| 18:20.10 | The_Boy_Wonder | anyway, i don't fully understand that, but my fix makes the jitterbuffer sound better for ConfBridge |
| 18:20.40 | The_Boy_Wonder | or it did for me. i suppose the real test is when we have the next conference call |
| 18:20.52 | russellb | we could have a call ... now! |
| 18:21.17 | The_Boy_Wonder | omg! actually, let me svn up leifmadsen's box |
| 18:21.26 | leifmadsen | ok |
| 18:21.26 | The_Boy_Wonder | you're good with a restart right leifmadsen |
| 18:21.28 | leifmadsen | I'm in 31337 |
| 18:21.31 | leifmadsen | yep go! |
| 18:21.36 | leifmadsen | I hung up |
| 18:23.21 | *** join/#asterisk-dev anonymouz666 (~anonymouz@187-28-37-118.poolip.RJO.embratel.net.br) |
| 18:27.07 | The_Boy_Wonder | ok restarted |
| 18:27.10 | The_Boy_Wonder | 31337 go ! |
| 18:27.49 | russellb | i'm the only person in this conf. |
| 18:30.16 | kaii | leifmadsen: thx for suggestions |
| 18:30.40 | leifmadsen | np |
| 18:32.57 | russellb | someone removed the merge properties from the 1.8 branch |
| 18:32.58 | russellb | :-( |
| 18:33.04 | leifmadsen | I hope it wasn't me... |
| 18:33.10 | leifmadsen | my merges today were taking..... a long time |
| 18:33.19 | leifmadsen | using merge68 and merge8trunk |
| 18:33.44 | russellb | did you control+C merge68? |
| 18:33.48 | leifmadsen | nope |
| 18:33.53 | leifmadsen | I just let it sit there until ti finished as far as I know |
| 18:37.19 | russellb | oh well, no big deal |
| 18:37.21 | russellb | was easy to fix |
| 18:38.42 | leifmadsen | sorry :( |
| 18:38.59 | leifmadsen | no idea how that happened ... has never happened to me before |
| 18:41.19 | *** join/#asterisk-dev joshaidan (~brianj@host-216-26-197-55.tbaytel.net) |
| 18:54.58 | *** join/#asterisk-dev elguero (~miguel323@17.140.211.66.inaddr.G4.NET) |
| 19:00.41 | *** join/#asterisk-dev miguel3239 (~miguel323@17.140.211.66.inaddr.G4.NET) |
| 19:05.36 | The_Boy_Wonder | leifmadsen: the box confbridge is on is a dedicated machine right, not a vm |
| 19:05.57 | seanbright | Corydon76-home: ping |
| 19:08.03 | leifmadsen | The_Boy_Wonder: yes |
| 19:08.09 | leifmadsen | The_Boy_Wonder: it has VMs on it, but Asterisk is on the host OS |
| 19:08.22 | The_Boy_Wonder | ok |
| 19:08.30 | leifmadsen | which is installed directly on hardware |
| 19:20.53 | *** join/#asterisk-dev miguel3239 (~miguel323@17.140.211.66.inaddr.G4.NET) |
| 19:30.03 | *** join/#asterisk-dev elguero (~miguel323@17.140.211.66.inaddr.G4.NET) |
| 19:44.35 | *** join/#asterisk-dev rogersja (~RogersJa@S0106000f6695039f.gv.shawcable.net) |
| 19:45.00 | *** join/#asterisk-dev moy (~moy@CPE002719f00364-CM0026f3a1701d.cpe.net.cable.rogers.com) |
| 19:52.41 | Corydon76-home | seanbright: pong |
| 19:52.58 | seanbright | afternoon. i am fixing an issue with mysql realtime. |
| 19:53.14 | Corydon76-home | Lucky you |
| 19:53.36 | seanbright | yeah. there was code introduced 100 years ago by M5881 |
| 19:53.39 | MuffinMan | [closed] [Asterisk] Core/Configuration 0005881: [patch] Allow use of separate MySQL servers for SELECT and UPDATE queries reported by voipgate https://issues.asterisk.org/view.php?id=5881 |
| 19:53.58 | seanbright | which you wrote |
| 19:54.01 | Corydon76-home | Ah |
| 19:54.24 | seanbright | and you made a change where instead of writing into a MYSQL_ROW directly, you ast_strdupa it first |
| 19:54.49 | seanbright | is that because it's not safe to modify the MYSQL_ROW in place? |
| 19:54.58 | seanbright | or has it been so long that you don't remember? :-) |
| 19:55.19 | Corydon76-home | I believe the design of the mysql interface makes the data opaque |
| 19:55.35 | Corydon76-home | So it's not clear whether it's safe, so I assumed it was not safe |
| 19:55.48 | seanbright | fair enough |
| 19:56.35 | seanbright | currently my patch removes the ast_strdupa and acts on the row directly... but i'm considering putting it on the heap |
| 19:57.20 | Corydon76-home | which branch/line number? |
| 19:58.02 | seanbright | trunk |
| 19:58.20 | seanbright | 340ish |
| 19:58.41 | seanbright | asterisk/trunk/addons/res_config_mysql.c |
| 19:58.58 | seanbright | patch is attached to M19090 |
| 19:59.00 | MuffinMan | [feedback] [Asterisk] Addons/res_config_mysql 0019090: [patch] 'voicemail show users for default' with a lot of entries in realtime voicemail cores asterisk reported by mickecarlsson https://issues.asterisk.org/view.php?id=19090 |
| 20:00.46 | Corydon76-home | seanbright: Yeah, why don't you use an ast_str there, instead? |
| 20:01.11 | Corydon76-home | Possibly threadstorage, even |
| 20:01.37 | seanbright | agreed. probably safer. |
| 20:01.41 | seanbright | thanks for the input. |
| 20:03.56 | Corydon76-home | BTW, problem is unrelated to the issue number you identified. Threw me off a little |
| 20:05.19 | Corydon76-home | But it's another one of the strdupa running out of space thingies |
| 20:06.38 | seanbright | how is it unrelated? |
| 20:06.49 | seanbright | the strdupa commit came from 5881 |
| 20:07.09 | Corydon76-home | I was thinking about the read/write multiple specification |
| 20:07.17 | seanbright | ah, right |
| 20:08.07 | Corydon76-home | Are you fixing both load routines? |
| 20:08.41 | seanbright | that is the intent, yes |
| 20:18.43 | seanbright | actually, i changed my mind |
| 20:19.58 | seanbright | since you never have to free 'row' (or pass it back into the mysql api), i'm feeling pretty good that i can modify it |
| 21:16.20 | Deeewayne | russellb, I'm can't volunteer to be seanbright-active, but if you guys decided on the community supported 1.4 branch option that has been discussed on the mailing list, I would volunteer to do some work. |
| 21:20.17 | russellb | Deeewayne: cool. If anyone steps up that is willing to lead the effort, I'll pass your name along. |
| 21:20.36 | russellb | So far nobody has stepped up with a level of commitment necessary to lead it. |
| 21:20.52 | Deeewayne | yeah, leading the effort is something I can't volunteer for |
| 21:21.00 | russellb | no problem, understood |
| 21:21.06 | russellb | i'm not terribly optimistic anyone will |
| 21:21.27 | Deeewayne | me neither. I would have volunteered to help a few days ago if I thought it was actually going to get traction |
| 21:22.42 | seanbright | seanbright-active? pffft |
| 21:22.58 | Deeewayne | that's like asterisk -vvvvvvvvvvvvvvvgc |
| 21:40.07 | *** join/#asterisk-dev mvanbaak (~michiel@asterisk/contributor-and-bug-marshal/mvanbaak) |
| 21:40.08 | *** mode/#asterisk-dev [+o mvanbaak] by ChanServ |
| 21:50.11 | *** join/#asterisk-dev micols (~ident@rlogin.dk) |
| 22:31.08 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 22:31.08 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 23:03.55 | *** join/#asterisk-dev citywok (~citywok@67-134-194-33.dia.static.qwest.net) |
| 23:15.36 | *** join/#asterisk-dev miguel3239 (~miguel323@17.140.211.66.inaddr.G4.NET) |
| 23:34.33 | *** join/#asterisk-dev tzafrir_laptop (~tzafrir@bzq-218-155-145.cablep.bezeqint.net) |
| 23:34.33 | *** mode/#asterisk-dev [+o tzafrir_laptop] by ChanServ |
| 23:38.39 | *** join/#asterisk-dev tzafrir (~tzafrir@212.179.75.202) |
| 23:38.40 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |