**** BEGIN LOGGING AT Fri Mar 10 03:00:04 2017 Mar 10 08:20:43 DocScrutinizer05: ping Mar 10 08:41:48 anybody can help me with a regular expression? Mar 10 08:41:59 i can try Mar 10 08:42:11 but i dont know advanced usage Mar 10 08:42:47 KotCzarny: QRegExp re("[\"]signature[\"],([^\\(]+)"); Mar 10 08:43:01 and your question? Mar 10 08:43:15 and "k=new Ev(f.url);f.sig?k.set("signature",f.sig):f.s&&k.set("signature",vc(f.s));" is (a part of) what I want to match Mar 10 08:43:22 I need "vc" extracted Mar 10 08:43:42 I modified it to "[\"]signature[\"],([\w+]+\()" Mar 10 08:43:50 but it matches "vc(" Mar 10 08:44:00 What are the []s around " for? Mar 10 08:44:10 this is youtube signature function extractor from cutetube :) Mar 10 08:44:34 there are two "signature" strings Mar 10 08:44:34 Presumably you want to change \() to )\( Mar 10 08:44:47 KotCzarny: yes Mar 10 08:44:56 but what I need is the second one Mar 10 08:45:11 (at the end) Mar 10 08:45:20 the "vc(f.s)" call Mar 10 08:45:29 Maxdamantus: no, it closes the matching group Mar 10 08:45:54 you want "\"signature\",(vc.*?)\); maybe? Mar 10 08:45:58 freemangordon: right, but you only want "vc", not "vc("? Mar 10 08:46:00 lemme try Mar 10 08:46:05 "\"signature\",(vc.*?)\);" Mar 10 08:46:07 Maxdamantus: yes Mar 10 08:46:15 KotCzarny: ah, it is not that simple Mar 10 08:46:20 the function name changes ;) Mar 10 08:46:24 So you expect a literal ( to appear after the matched group. Mar 10 08:46:30 ie (..)\( Mar 10 08:46:52 "\"signature\",([a-z+\(].*?)\);" Mar 10 08:47:02 Maxdamantus: in English please, I can't really grok the syntax and logic of regexp :) Mar 10 08:47:24 erm, "\"signature\",([a-z+\(]+.*?)\);" Mar 10 08:47:32 I'm guessing the (non-escaped) brackets denote the matching group. Mar 10 08:47:38 Maxdamantus: yes Mar 10 08:47:43 while escapd brackets are just literal characters. Mar 10 08:47:46 yes Mar 10 08:47:48 try the last one Mar 10 08:47:51 ok Mar 10 08:48:09 KotCzarny: your last bracket is the wrong way round. Mar 10 08:48:13 KotCzarny: hmm, that ";" at the end... Mar 10 08:48:19 Should be \( at the end, not \) Mar 10 08:48:42 oh, so we want only function name Mar 10 08:48:46 KotCzarny: remember, this is obfuscated js, we may have "," and whatnot instead of ";" Mar 10 08:48:51 KotCzarny: yes Mar 10 08:49:35 ah, wait, so if I add another "\" at the end, it will be stripped from the result? Mar 10 08:49:42 "\)" Mar 10 08:49:50 I'd probably go with: "\"signature\",([^\)]+)\(" Mar 10 08:50:05 assuming I understand the regex syntax. Mar 10 08:50:33 * Maxdamantus prefers just using the syntax used in sed/grep. Mar 10 08:50:42 BRE or whatever it is. Mar 10 08:50:42 that matches nothing Mar 10 08:50:50 er Mar 10 08:51:05 "\"signature\",([^\\)]+)\\(" Mar 10 08:51:18 since this is a string literal in some language. Mar 10 08:51:41 (should escape the backslashes so they're part of the regexp string) Mar 10 08:53:17 \"signature\",([^(:]+)\( Mar 10 08:53:24 this one, otherwise it would catch too much Mar 10 08:53:45 KotCzarny: which backslashes are meant to appear in the regexp? Mar 10 08:54:08 yeah, though php's re didnt complain about ( in [] Mar 10 08:54:32 KotCzarny: want me to try ^^^? Mar 10 08:54:39 I'm talking about "\(" == "(" Mar 10 08:54:46 dunno if that's true in PHP Mar 10 08:55:00 KotCzarny: cool, looks like working Mar 10 08:55:11 ahm, additional quotes in qregexp might be needed Mar 10 08:55:22 yes, I know Mar 10 08:55:23 fmg, but it's fragile, as always Mar 10 08:55:32 I am using https://regex101.com/ to test Mar 10 08:55:34 since js/webscraping is quirky Mar 10 08:55:38 yes, I know it is fragile Mar 10 08:56:20 there are pre matches in RE, but i have never really grasped them, yet Mar 10 08:56:29 KotCzarny: if you have a better idea on how to get the signature frunction from youtube base.js, I am all ears :) Mar 10 08:56:34 (think of match withing a match) Mar 10 08:57:01 *within Mar 10 08:58:05 i'm using this snippet for testing: http://pastebin.com/raw/F92Kn0mP Mar 10 08:58:37 KotCzarny: hmm, your regexp counts on ":" being present, which is not guaranteed Mar 10 08:59:02 fmg, then you should first try with : and if it returns empty do second re without : Mar 10 08:59:30 no need to escape the ( inside [] btw (just sayin') Mar 10 08:59:33 and function you have pasted looks like trigraph (a?b:c) Mar 10 08:59:59 so unless you match multiple cases, it should work Mar 10 09:00:40 this is how it looks today, tomorrow google will change it to look some other way Mar 10 09:00:49 I want ti working tomorrow as well Mar 10 09:00:57 see 'fragile' part Mar 10 09:01:09 that's why you use youtube-dl -U Mar 10 09:01:28 youtube-dl is fragile too, it's just that it's updated regularly ;) Mar 10 09:01:41 inz: it is qregexp, afaik it is a little bit different than php, those brackets were put there by the original author (marxian), I don;t want to mess with them Mar 10 09:02:22 inz: ever tried cutetube? Mar 10 09:02:46 doesnt cutetube rely on youtube-dl ? Mar 10 09:02:49 no Mar 10 09:03:03 it has has nothing in common afaik Mar 10 09:03:09 I think I once did, it didn't work at that time. Mar 10 09:03:24 there is cutetube2 which is the one maintained Mar 10 09:03:34 well, yes, I meant cutetube2 Mar 10 09:04:11 which is stupid imo and ct should be replaced with ct2, and whole ct2 just should point to ct Mar 10 09:04:50 offtopic, i've just fixed oscp to pass -Wall -Wextra -Werror ;) Mar 10 09:06:57 fmg, there is no guarantee how base.js will look tomorrow, so no matter the RE. you just can't rely on anything being stable in it Mar 10 09:07:10 imo a youtube program should just do things by executing `youtube-dl` Mar 10 09:07:54 that way the user can at least keep an up-to-date version of it in his/her PATH if everything else ends up unmaintained. Mar 10 09:07:57 maxd: does youtube-dl support spewing data to stdout? Mar 10 09:08:29 KotCzarny: I imagine so, if not conveniently, it must do with some piping hacks. Mar 10 09:08:40 ok, I think I got it, I need to call re.cap(2) for the second capturing group result Mar 10 09:08:44 It also supports just printing out a URL to download. Mar 10 09:08:45 KotCzarny, Maxdamantus : it does Mar 10 09:09:16 maxd: that's also nice, assuming googl doesnt rely on cookies Mar 10 09:09:23 but ime, Google has done things like requiring certain headers, so just passing that to curl/wget might not be reliable. Mar 10 09:11:32 ok, that's it - QRegExp re("[\"]signature[\"],((\\w+)\\s*\\()"); Mar 10 09:12:46 That outer group seems redundant. Mar 10 09:12:57 yup Mar 10 09:13:06 hmm? Mar 10 09:13:13 as do the []s Mar 10 09:13:18 in English, please :) Mar 10 09:13:38 he means: re("\"signature\",(\\w+)\\s*\\("); Mar 10 09:13:38 redundant or not, this is the first one that works correctly Mar 10 09:13:50 QRegExp re("\"signature\",(\\w+)\\s*\\("); Mar 10 09:14:00 jinx Mar 10 09:14:04 1st! Mar 10 09:14:25 I had to copy/paste/edit it in osso-xterm Mar 10 09:14:39 also, does \w catch . ? Mar 10 09:14:39 ^^^ does not work Mar 10 09:14:48 if allowing for spaces between func and paren, why not also between comma and func? Mar 10 09:15:21 inz: right Mar 10 09:15:26 freemangordon: are you looking at the same output? I'm guessing you'd need to look at a different one. Mar 10 09:15:42 yah, capture group changes from 2 to 1 with the change Mar 10 09:15:57 freemangordon: eg, if you were looking at some `match[1]` before it might be `match[0]` after removing the unnecessary capture. Mar 10 09:16:19 or 2, 1, dunno what convention Qt uses. Mar 10 09:16:35 IIRC 0 is the full match Mar 10 09:16:41 yes Mar 10 09:16:54 Mk, that's at least consistent with JS. Mar 10 09:17:12 * Maxdamantus always found tha illogical though. Mar 10 09:17:13 btw. fmg, you need: http://www.regular-expressions.info/lookaround.html Mar 10 09:17:24 but as i've said its advanced territory ;) Mar 10 09:17:29 s, what about [\"]signature[\"]\s*,\s*((\s*\w+)\s*\() Mar 10 09:17:40 *so Mar 10 09:17:54 I don;t really care if it is optimized or not Mar 10 09:18:10 \s*([^\s]+)\s*\( ? Mar 10 09:18:12 as long as it works correctly Mar 10 09:18:29 what would \w catch? Mar 10 09:18:32 KotCzarny: no, function name must be word Mar 10 09:18:48 it cannot contain dots, or columns etc Mar 10 09:18:54 *colons Mar 10 09:19:00 \s*([a-z0-9]+)\s*\( ? Mar 10 09:19:02 the fuck, what was the word Mar 10 09:19:11 :alpha: Mar 10 09:19:39 \w is equal to [a-zA-Z0-9_] Mar 10 09:19:55 but since it's obfuscated and minimized js i bet \s* around the \w is redundant Mar 10 09:20:05 no, it is not Mar 10 09:20:13 they put spaces sometimes Mar 10 09:20:38 probably obfuscator failing sometimes Mar 10 09:20:44 could be Mar 10 09:20:50 KotCzarny, qt does not support lookbehind Mar 10 09:21:25 inz: so, what about [\"]signature[\"]\s*,\s*((\s*\w+)\s*\() ? do I miss something obvious? Mar 10 09:21:40 fmg, bad \s inside Mar 10 09:22:15 you dont want your inner match to containe space Mar 10 09:22:24 since function name cant contain space Mar 10 09:22:44 not that it would ever match, because * is greedy by default Mar 10 09:22:48 KotCzarny: it is ok, it doesn't match "v c" Mar 10 09:23:11 that too, though i never know which direction RE starts being greedy Mar 10 09:23:34 guys, please, I don;t really want to learn regexps now :) Mar 10 09:23:40 liar Mar 10 09:23:41 yes you do Mar 10 09:23:45 ;) Mar 10 09:23:55 no, I don;t and I never wanted to ;) Mar 10 09:24:02 You need more problems! Mar 10 09:24:04 regexps are handy Mar 10 09:24:18 I would prefer a loop every time over a regexp :P Mar 10 09:24:20 Yes, they are an awesome tool. Usually the wrong one, but awesome nonetheless! Mar 10 09:24:37 (now some oldtimer should chime with 'in the DOS days we had two wildcards and we were happy') Mar 10 09:24:38 ;) Mar 10 09:24:47 yes, that's true Mar 10 09:24:58 you can use fnmatch() then Mar 10 09:25:05 :P Mar 10 09:25:15 it's a nice alternative to regexps Mar 10 09:25:26 the original code uses regexp, and I don;t wan't to change it much, unless really needed Mar 10 09:25:52 anyway, do you see anything wrong with ^^^ regex besides being suboptimal? Mar 10 09:25:55 yup, RE are better in webscraping, fnmatch is nice for simple things though Mar 10 09:26:18 fmg, it works, its ok, but i would change \s*\w+ to \w+ Mar 10 09:26:29 since it's not doing anything Mar 10 09:26:48 ok Mar 10 09:27:05 I'd probably rather do webscraping with DOM, but usually I'd just do it using command-line tools so it would be piped into sed which uses regexp. Mar 10 09:27:07 still works, after that change Mar 10 09:35:46 hmm, when does 'read battery bar' show in cssu? Mar 10 09:36:17 i have 5/1282mAh in status, got 'battery low' message, but it's still green bar Mar 10 09:37:05 5% of 1282 is ~64, so imo it should start showing red there Mar 10 09:37:33 freemangordon: what's the problem? Mar 10 09:37:38 (regex) Mar 10 09:37:41 yeah Mar 10 09:37:50 but it seems we fixed it so far Mar 10 09:38:19 fmg, i have latest cssu-thumb Mar 10 09:38:26 cool Mar 10 09:38:52 check if you didn;t "fix" something some 2-3 years ago first :) Mar 10 09:39:11 you shouldn't have green bar on low battery Mar 10 09:39:32 fmg, it was ok before update Mar 10 09:39:36 (2012) Mar 10 09:40:01 KotCzarny: https://github.com/community-ssu/status-area-applet-battery/blob/master/status-area-applet-battery.c Mar 10 09:40:32 battery_status_plugin_update_icon Mar 10 09:41:16 lemme share the nice regex stuff I found Mar 10 09:42:53 let's start with alias man=':(){ [ "sed" == "$1" ]||{ /usr/bin/man $1;return; }; less -f <(/usr/bin/man sed) <(wget http://sed.sourceforge.net/sed1line.txt -q -O -); };:' Mar 10 09:44:43 ~fiasco Mar 10 09:44:43 L4-compatible real-time microkernel capable of running Linux in usermode. URL: http://os.inf.tu-dresden.de/fiasco/ Mar 10 09:46:08 http://stackoverflow.com/questions/22937618/reference-what-does-this-regex-mean/22944075#22944075 Mar 10 09:47:54 ^^^ excellent! Mar 10 09:48:13 can someone confirm that using 0xffff is safe to flash both Vanilla_PR_EMMC_MR0 and PR_COMBINED_MR0 images ? Mar 10 09:48:31 yes, is safe Mar 10 09:48:38 DocScrutinizer05: thanks Mar 10 09:48:55 if it fails, pester pali ;-9 Mar 10 09:49:14 i will :) Mar 10 09:50:02 If it breaks your device, you get to keep both pieces. Mar 10 09:57:09 N900 is basicaly unbrickable, so safety of 0xFFFF is almost only a question of host system it runs on Mar 10 09:57:47 in that regard I think 0xFFFF must be safer than my lazyflashing mess Mar 10 09:58:16 ~lazyflashing Mar 10 09:58:17 well, lazyflashing is http://wiki.maemo.org/Updating_the_tablet_firmware#The_Lazy_Approach Mar 10 09:59:09 at least I'm pretty sure 0xFFFF doesn't try LD_PRELOAD tricks ;-P Mar 10 10:00:46 hehe, that statusbar only goes red when value is 0 Mar 10 10:01:18 so with 0mAh reported ;) Mar 10 10:01:34 should be named 'empty' not 'verylow' :P Mar 10 10:02:36 that's with bme_replacement, without it it's set in 0..80 range Mar 10 10:03:50 my bme is 0.8.55+0m5 though Mar 10 10:03:56 weird Mar 10 10:05:09 peetah: you nevertheless might want to give lazyflashing a try, and pester me if it doesn't help you Mar 10 10:07:39 ~lazyflashing Mar 10 10:07:39 well, lazyflashing is http://wiki.maemo.org/Updating_the_tablet_firmware#The_Lazy_Approach Mar 10 12:48:54 Hello, I'm still trying to install NITDroid on my N900. Don't want to bother anybody, but if you may help... Mar 10 16:39:44 Hello, I am reading the following page: Mar 10 16:39:45 https://wiki.maemo.org/Nitdroid_easy_install_on_EMMC Mar 10 16:40:13 There's a simplified procedure for installing NITDroid on the N900. However links called by the installer are broken. Mar 10 16:40:29 I have been able to find some packages here: Mar 10 16:40:58 http://repository.maemo.org/extras/pool/fremantle/ Mar 10 16:41:30 But when trying to install kernel-power-settings, two more packages are required, namely: Mar 10 16:41:48 kernel-feature-overclock Mar 10 16:41:57 kernel-feature-smartreflex Mar 10 16:42:15 They are not in the mentioned repository. Does anybody know where to find them? Mar 10 16:48:03 Enrico__: those are provided by kernel-power package Mar 10 16:48:28 or kernel-power-flasher, can't remember Mar 10 16:49:14 I already installed kernel-power. So probably I need kernel-power-flasher. I will try soon. Mar 10 16:50:44 Enrico__: why not use HAM to install stuff? it will pull the required dependencies for you Mar 10 16:51:02 You mean the package manager? Mar 10 16:51:29 (Once started up Maemo, in the graphical environment?) Mar 10 16:51:29 yes Mar 10 16:51:44 It does not work. Sits there and doesn't download anything. Mar 10 16:51:45 HAM == hildon-application-manager Mar 10 16:51:58 give it some time, like 10 minutes :) Mar 10 16:52:04 Ah sorry. DIdn't know the acronymous. Mar 10 16:52:08 10 minutes??? Mar 10 16:52:16 if you want it to be fast, install CSSU Mar 10 16:52:35 ~speedyham Mar 10 16:52:35 SpeedyHAM is 30 times faster than HAM http://maemo.merlin1991.at/cssu/community-devel/pool/free/h/hildon-application-manager/hildon-application-manager_2.2.73-2_armel.deb. SpeedyHAM is included in CSSU now Mar 10 16:52:48 Ok ok I can wait... I just thought it didn't find anything! Mar 10 16:53:16 And what about multiboot-kernel-maemo and multiboot-kernel-power? Same way? Mar 10 16:53:30 also, stock HAM tries to use Nokia repos, which are dead for a long time Mar 10 16:53:38 multiboot is deprecated Mar 10 16:53:52 also, never used it so dunno Mar 10 16:54:06 I tried with u-boot, but created many problems. Mar 10 16:54:12 u-boot is the way Mar 10 16:54:26 shouldn;t create any problems Mar 10 16:54:39 Ok. The problems are: Mar 10 16:54:57 you just neet to install the required -bootimg (like kernel-power-bootimg) and it will created the needed entries for you Mar 10 16:55:21 *create Mar 10 16:55:53 - first, the interface. I have been unable to access it by usb, so I have to manage by the small N900 keyboard and screen Mar 10 16:56:05 which interface? Mar 10 16:56:06 - and the keyboard mapping is not standard Mar 10 16:56:16 I mean the shell. Mar 10 16:56:20 Sorry bad language. Mar 10 16:56:32 u-boot shell? why do you need it at first? Mar 10 16:56:44 *first place Mar 10 16:57:01 I am trying to understand how it works. Just to be able to put files where I want. Mar 10 16:57:08 ah, I see Mar 10 16:57:11 ~uboot Mar 10 16:57:11 N900 uBoot is a siamese twin binary [uBoot+stockMaemoKernel] that resides in kernel NAND partition /dev/mtd3 aka "kernel". You can't uninstall it, rather you'll nuke it when you flash/install another kernel like stock maemo kernel or powerkernel. To start other than stock maemo kernel via uBoot, you have to provide the according kernel image files, or http://talk.maemo.org/showthread.php?t=81613 Mar 10 16:57:46 Enrico__: look at ^^^ Mar 10 16:57:54 the thread at TMO I mean Mar 10 16:57:56 I am reading. Mar 10 16:58:18 also, install kernel-power-bootimg, to give you an example on how to create u-boot entry Mar 10 16:58:48 you just copy it, edit to your needs, run u-boot-update-bootmenu and thats it Mar 10 16:59:24 Ok I already understood something about u-boot in the last few days. I didn't find the name of the NAND partition, now I have it. Thanks for that. Mar 10 16:59:35 Enrico__: look in /etc/bootmenu.d Mar 10 16:59:39 NAND? Mar 10 16:59:41 are you sure? Mar 10 17:00:19 What is written a bit up? Not "kernel NAND partition"? Mar 10 17:00:20 there are 5 OneNAND partiotion, and all of them are used by maemo Mar 10 17:00:51 Enrico__: read the thread, it is all there Mar 10 17:00:59 A minute... Mar 10 17:01:17 Enrico__: /me gtg, ping Pali if you need help re u-bbot Mar 10 17:01:24 *u-boot Mar 10 17:01:49 I already had many talks with Pali. Also via email. He helped me to unbrick my device. Mar 10 17:02:01 as expected :) Mar 10 17:02:04 But he says he does not know about NITDroid. Mar 10 17:02:13 actually he is the u-boot maintainer Mar 10 17:02:23 me neither, sorry Mar 10 17:02:30 anyway, bye Mar 10 17:02:34 I know. But NITDroid has not any automatic procedure for installing on u-boot. Mar 10 17:03:42 I need to understand how to access the NAND partition where the /root directory is (but that's not mandatory to be honest), and I don't know exactly what to pass to NITDroid through u-boot to obtain a correct startup. Mar 10 17:04:31 I have been able to begin the boot, but NITDroid kernel doesnt' find many things (I think the root directory and the initial file system.) Mar 10 17:33:22 xes, why does tmo direct link to post breaks formatting? looks like runaway
or something, can it be fixed? Mar 10 17:33:39 in related new, oscp has now manual on tmo Mar 10 17:33:42 *news Mar 10 17:53:13 KotCzarny: an example? Mar 10 17:53:58 just any post, for example http://talk.maemo.org/showpost.php?p=1460158&postcount=3 Mar 10 17:55:27 KotCzarny: have you tried another browser or a different user agent? Mar 10 17:55:45 hmm, it might be that dark-classic them im using. hrm Mar 10 17:55:56 can you click theme on the bottom? Mar 10 17:56:29 http://i.imgur.com/0SXVxWi.png Mar 10 17:56:34 that's how it looks for me Mar 10 17:57:11 yep. It's an issue of the dark theme Mar 10 17:57:26 eh. any chance of fixing that theme? Mar 10 17:58:48 Mar 10 17:58:54 i will try.. Mar 10 17:58:55 this looks like a culprit Mar 10 17:59:48 and while it might be genuinely used for centering post on the screen, that classic themes might miss uncentering the content after that Mar 10 18:00:54 why not just scroll to post instead of displaying single-post? Mar 10 18:03:08 ie. those little links on thread would look better this way: http://talk.maemo.org/showthread.php?t=94590#postcount1460158 Mar 10 18:03:37 and has a bonus of being able to read previous/next posts without changing page Mar 10 18:04:35 btw. that last link already works Mar 10 18:19:04 showpost is fubar, s/post/thread/ and you're fine Mar 10 18:19:44 literally Mar 10 18:20:34 actually showpost always been fubar Mar 10 21:50:38 FUCK OFF PULSEAUDIO!!! Mar 10 21:50:54 kernel: [14525.614165] pulseaudio[5521]: segfault at 18 ip 00007ff7a1cd16d9 sp 00007ffdad0eefb0 error 4 in libpulsecore-5.0.so[7ff7a1c6f000+8f000] Mar 10 21:50:56 it crashed Mar 10 21:51:32 and now when I want to start it again it again it just show me error message: N: [pulseaudio] main.c: User-configured server at {b80eeeba890737b39f0c899a00000008}unix:/tmp/pulse-AcBAanBtuBoD/native, refusing to start/autospawn. Mar 10 21:51:37 and nothing is running Mar 10 21:51:42 so no sound Mar 10 21:52:39 great software... looks like if it do not do clean shutdown, then it is not possible to start it again Mar 10 22:00:32 Pali: what distro is that? Mar 10 22:00:39 ubuntu? Mar 10 22:00:41 yes Mar 10 22:00:46 12.04? Mar 10 22:00:50 yes Mar 10 22:01:01 seems user here have same problem http://unix.stackexchange.com/questions/282048/pulseaudio-how-to-restart-without-restart-x11-if-delete-tmp Mar 10 22:01:06 but I did not deteled anything Mar 10 22:01:08 isn't it still supported? Mar 10 22:01:14 just pulseaudio crashed Mar 10 22:01:24 12.04 is still supported Mar 10 22:01:45 well, at least send error report to canonical Mar 10 22:01:53 they might fix it Mar 10 22:02:16 I doubt... and I do not have reproducer Mar 10 22:02:31 this is first time it crashed in way that it is not able to start Mar 10 22:02:49 you should have apport (or something like that) Mar 10 22:02:55 disabled Mar 10 22:02:59 ah Mar 10 22:03:12 coredump? Mar 10 22:03:39 do not have it Mar 10 22:03:51 but I'm angry as I cannot start it again Mar 10 22:04:07 basically now it is not running and it fail to start Mar 10 22:04:13 yeah, got it Mar 10 22:04:30 maybe killing X will suffice? Mar 10 22:04:46 crash does not matter here... but nonworking start is wrong! Mar 10 22:04:58 sure Mar 10 22:05:05 I have open apps which I do not want to close now... so restart X not right now Mar 10 22:05:15 but maybe it is X related... Mar 10 22:05:20 going to google Mar 10 22:05:54 bingo! pax11publish -r it fixed Mar 10 22:06:07 :) Mar 10 22:06:07 https://wiki.parabola.nu/PulseAudio/Troubleshooting Mar 10 22:06:42 pax11publish -d show me some registered path to crashed and non-running PA process Mar 10 22:06:50 that -r clean it Mar 10 22:07:04 and now pulseaudio server started again Mar 10 22:07:11 stupid error message Mar 10 22:07:25 and even stupid detection if server is running or not Mar 11 00:01:40 https://linux.die.net/man/1/pax11publish *COUGH* >>FUCK OFF PULSEAUDIO!!!<< indeed Mar 11 00:05:54 >>Authors: The PulseAudio Developers << THAT EMAIL ADDR!!! Any questions Watson? **** ENDING LOGGING AT Sat Mar 11 03:00:01 2017