01:20:55 WHat's the difference between this channel, and #freebsd-irc ? 01:22:18 #freebsd-irc is if you aren't registered with nickserv 01:24:33 Ah! Ok. No wonder it was so much more quiet. 01:37:00 just built a new machine and I was wondering if freebsd would be solid on a newer amd ryzen 7000 series processor, I want to use it on one of my SSDs 01:37:37 I have better luck with older hardware most of the time with BSD-based systems, so I don't want an unstable system 01:37:55 any help would be appreciated, and thanks ahead of time 01:43:33 johnm: if it's unstable the hardware is probably needs intervention 01:44:16 rtprio: so it should not have any specific errors that are related to the OS itself? 01:44:31 no you should not 01:45:10 ok, thank you, I was hoping I had not missed something obvious and just wanted to confirm 01:48:08 perhaps someone who has that specific hardware can speak up. 01:49:42 I'll be here if someone would like to add some specific results, and it would be appreciated to have that confirmation too 08:35:00 * micttyl suffers 08:36:13 whois micttyl 08:36:15 oops 08:39:26 you just increased their suffering 08:40:22 It's a cruel world out there! 08:43:10 Hi all, I'm having issues with curl(1) refusing to connect to HTTPS servers, failing with 'SSL certificate problem: unable to get local issuer certificate'; pkg(8) reports ca_root_nss is installed, I even pkg install -f ca_root_nss for good measure, but no dice 08:44:13 Perhaps something has or is serving revoked|expired certificate? 08:44:44 What is your OS version? 08:44:52 no parv , I'm talking about facebook.com, google.fr, etc. 08:45:03 12.4-RELEASE 08:46:08 https://x0.at/omiR.txt 08:47:39 is your clock off? 08:47:53 Sat Jan 14 09:47:42 CET 2023 # no 08:48:12 the error really points to missing certs, not failing checks for dates 08:48:33 i didnt read your url until now 08:48:34 yeah 08:48:39 you're missing ca certs, thats no bueno 08:55:30 FWIW, curl here (FreeBSD 13.1-STABLE) is 7.86, ca_root_nss is 3.86 08:56:34 https://x0.at/3_Sn.txt 08:56:53 ca_root_nss-3.86 curl-7.87.0 08:57:44 this seems to be a global problem across all my jails, too :( 09:01:06 *only on my jails 09:01:27 I really had other things to do today T___T 09:01:52 moviuro, Here I have 128 links (linked to in /usr/share/certs) compared to 1 on your side 09:02:40 I need to re-trigger the install script of ca_root_nss, I guess. no idea how to do that though 09:04:03 If I remember right, pkg delete -f doesn't uninstall things packages depending on the uninstalled package. Could uninstall and reinstall 09:04:33 but 09:04:42 since you did pkg install -f, I don't think there'd be a difference 09:05:21 xtile: can confirm no difference 09:05:34 * xtile nods. 09:08:09 Using The Ports system, a port can (or, could when I last tried) be removed without removing any other packages: cd security/ca_root_nss && make build && make deinstall && make install 09:08:43 parv: I uninstalled the package and reinstalled it. I have no ports anywhere, I'm *not* touching that today 09:09:00 Yeah, mixing ports and pkg is unfun. 09:09:12 🤷‍♂️ 09:09:18 I wish ports could be updated to match quarterly pkg. 09:10:16 some jails are missing a LOT of files :/ https://x0.at/Gy5K.txt 09:16:04 xtile: that will bugs you each time installing something relavant, because it will keep suggesting you to install after all 09:16:28 hmm? 09:16:46 ah, yes, as for forcibly uninstalling a package 09:17:06 I know, since a year or so ago, bash suddenly became a dependency of many packages, and I kept manually uninstalling it. Fortunately that was fixed. 09:18:41 certctl -v rehash 09:18:42 what kind of magic does one need from bash? it should be just like "oh, i didn't comply, f it ich habe no time" 09:21:28 if it was required for the dev version, that should be reasonable though 09:25:11 what happened was a bunch of packages began requiring bash-completion as a run dependency, which also then required bash. i believe the fix was a ports dev turned it into just a build dependency 09:25:27 been a while 09:29:47 https://svnweb.freebsd.org/ports?view=revision&revision=566543 09:29:48 Title: [ports] Revision 566543 09:34:33 * micttyl suffers less 11:15:30 johnm: I am using the AMD Ryzen 7 Pro 4750U on OpenBSD, and it works great. 11:18:03 I would guess when it works on OpenBSD, there's a high chance it will work on FreeBSD. The problem I've usually had is hardware not working on OpenBSD but working on FreeBSD. 11:25:48 I am excited to build a new computer sometime this year. My current one's one I built in 2012, and it doesn't even count as x86-64-v2. https://en.wikipedia.org/wiki/X86-64#Microarchitecture_levels I heard some GNU/Linux distros have been moving to x86-64-v2 and dropping original x86-64 support, actually. 11:25:49 Title: x86-64 - Wikipedia 11:26:25 Reminded by Ryzen and the fact the new ones have AVX-512 and stuff. 11:33:42 No clue what AVX-512 is. Not a big hardware guy. I took a chance on AMD because my other laptop is Intel, and in a crippled state to address all the issues with their recent CPUs. 11:35:15 At least lost 50% of its capacity. 11:35:25 dang 12:06:14 xtile: that sounds like a big break. and feels a lot faster than the move from i386 to i586 / i686 12:08:24 perhaps... It means those distros no longer support my current hardware. But when I asked all my friends about their hardware, no-one else was affected by this 12:34:55 I did a pkg search latex and didn't find a package that looked right, what's the recommended way to install latex on FreeBSD (13.1) 13:00:11 A quick web search hints that texlive-full may be a good one for that. 13:42:25 crb: iirc I use tex-formats for latex 13:45:25 * meena uses https://tectonic-typesetting.github.io/en-US/ 13:45:35 but no one has put that into ports yet 14:24:03 hello people someone can help to me i wiil try run from freebsd 13.1 the re::engine::RE2 and i get error https://bsd.to/E1DK 14:24:05 Title: dpaste/E1DK (Plain Text) 14:28:48 meena: do you compile tectonic on freebsd? 15:42:51 ld: error: undefined symbol: vesa_unload_ioctl 15:42:53 >>> referenced by vesa.c:1965 (/usr/src/sys/dev/fb/vesa.c:1965) 15:42:56 >>> vesa.o:(vesa_mod_event) 15:43:11 i was trying to compile a PAE minimal kernel on i686 and I got this 15:43:45 (by "i686" I mean "Haswell running in 32-bit mode." I am not stupid enough to run a Pentium Pro in 2023.) 16:08:34 git grep -F hints it's in syscons, not vt. 16:08:54 After a run of "portupgrade -a" I get a long list of "No such file or directory @ rb_check_realpath_internal - /usr/local/lib/compat/pkg/libstdc++.so.6" for various packages all in /usr/local/lib/compat/pkg what is this telling me? 16:36:15 V_PauAmma_V: so i shouldn't nodevice sc, got it 16:36:27 merci 16:46:43 wwt: I believe to have done that before, yes 16:50:43 "created … using an unholy mix of python code and manual editing" https://github.com/freebsd/freebsd-src/commit/017de00844ae9954214cefb72e350845c928ab83 16:50:44 Title: Add a SVG version of vector-Beastie. · freebsd/freebsd-src@017de00 · GitHub 17:02:04 meena: freebsd foundation doesnt even have the rights in the beastie image 17:02:16 so it is not allowed to manipulate it 17:14:14 wwt: i knew that, but i didn't know Marshall Kirk McKusick held it: http://www.mckusick.com/beastie/mainpage/copyright.html 17:14:16 Title: BSD T-shirts 17:22:26 kusick should waive his rights and make it free 17:29:18 * V_PauAmma_V . o O ( "should"? ) 17:39:58 must 18:10:48 I'm currently updating zabbix5-server to zabbix6-server. with that comes an update from mysql5.x to mysql8.0. my problem is that the zabbix-server seems to be able to connect to the mysql database via the /tmp/mysql.sock socket (as it can successfully query the database schema version) but as soon as it wants to alter the table (performing the update) it looses the connection and reports: "Can't connect to local MySQL server through socket /tmp/mysql.sock (61)" 18:10:51 this runs in a jail 18:11:30 does anybody have any idea? this was working fine for many years now. I have been digging for over an hour now but nothing seems to work. The PHP webinterface seems to have no problems connection to the database via the socket using the same credentials. 18:13:41 before that it literally reports: "Lost connection to MySQL server during query" 18:13:48 tct: can you do it? with the MySQL Client? and do tube thing it tries (and fails)? 18:14:06 meena, what do you mean by "do the tube thing"? 18:14:16 I can login with the same user, same credentials, same socket via the CLI client. 18:14:22 same, the same thing 18:15:37 meena, I'd rather not run the same alter table statement manually as this is part of an upgrade procedure performed by zabbix6-server. don't want to get into more problems 18:18:09 is the code not idempotent? 18:18:39 I do not understand what that means :/ 18:20:37 I've also updated the jail from 13.0-RELEASE to 13.1-RELEASE 18:22:20 https://en.m.wikipedia.org/wiki/Idempotence#Computer_science_meaning 18:22:21 Title: Idempotence - Wikipedia 18:23:04 zabbix 5 to 6 upgrade is seamless. the upgrade of database is handled by zabbix itself. 18:23:24 * otis putting net-mgmt/zabbix* maitainer hat on. 18:23:47 otis, exactly - and zabbix cannot do that: https://pastebin.com/dZqxQJc4 18:23:48 Title: 11211:20230114:180906.199 Starting Zabbix Server. Zabbix 6.0.12 (revision 126aa - Pastebin.com 18:23:53 otis, heh, thanks for maintaining this :) 18:24:38 hm, in fact, i don't use it with mysql, only with postgres. you could check mysql error log perhaps? 18:24:40 otis, if I set incorrect credentials or connection settings in the zabbix server config. it also errors but then isn't able to query the database version. 18:25:06 otis, yeah, this is the last and only server I have with mysql. I migrated to psql some time ago but can't/don't-want-to do this with this one here right now. 18:25:43 otis, trying to locate the corresponding mysql log files... 18:25:43 or clone the database (or at least affected tables) and perform the query manualy to see what's the error. 18:27:51 treefrom 18:27:56 otis, just checked mysql logs. nothing relevant to this. it just says that everything is fine :< 18:28:03 oops, sorry 18:32:11 otis, if I set an incorrect DBPassword= in the zabbix server config, then restart the zabbix_server service, it tells me that it couldn't authenticate. so what I have shown in the paste would clearly indicate that it CAN communicate with the mysql server. (especially given that it can successfully read the version from the database) 18:37:38 so I did what otis & meena suggested: login with the same user & password (also throught he socket) and ran the same query: https://pastebin.com/T0Z8ALrW 18:37:39 Title: zabbix_server:/root@[18:20] # mysql -u zabbix -p -S /tmp/mysql.sockEnter passw - Pastebin.com 18:38:10 (I did a 'use zabbix;' inbetween to select the 'zabbix' database. 18:39:26 seems like _something_ kills the connection 18:43:36 tct: did you do any mysqlcheck to see if the database isn't corrupted? 18:44:36 otis, mysqlcheck -c zabbix returns all "OK" 18:45:08 otis, but then it looses connection again: https://pastebin.com/rB5iHqP8 18:45:09 Title: zabbix_server:/root@[18:29] # mysqlcheck -c zabbixzabbix.acknowledges - Pastebin.com 18:45:43 so clearly there's something wrong with your mysql, i'd say. 18:45:53 hence I came to ask for help :D 18:46:05 :-/ default log is in /var/log/mysql/mysql.err 18:46:17 or mysqld.err 18:46:18 that does not exist 18:46:45 log files are in /var/db/mysql for me 18:47:20 ok, so you should check them. 18:49:13 otis, it would appear that the zabbix.dashboard table is corrupt 18:49:44 mysqlcheck -r --auto-repair maybe? 18:50:09 yeah I am trying to run the repair but "The storage engine for the table doesn't support repair' 18:50:16 I remember why I switched to psql... 18:51:45 changing the engine on only that table doesn't work either as: "because the table participates in a foreign key constraint" 19:01:46 error log on MySQL might have to be enabled 19:16:26 otis, I restored everything to the state before I touched something, then did the same procedure again. On first start of zabbix6-server: https://pastebin.com/aVvmJYjR 19:16:27 Title: 37039:20230114:190112.344 Starting Zabbix Server. Zabbix 6.0.12 (revision 126aa - Pastebin.com 19:17:02 so clearly everything is fine other than the corrupted database. 19:17:10 not sure if it's worth fixing or starting over with psql 19:18:21 * meena votes for PostgreSQL 19:19:10 * otis votes for PostgreSQL 19:19:42 it has one very small drawback: it's more than advisable to track every major version of postgres and update in a timely manner. 19:20:12 esp. when you decide to go with timescale 19:21:44 I wonder if there's any way to migrate the data… https://pgloader.io/ 19:21:45 Title: pgloader 19:22:39 it's ⬆️ written in Common Lisp 20:01:46 It must be most excellent then :-) 20:31:23 cracauer: it's rock solid.