01:46:47 So, a "pkgbaseify" works fine for FBSD 14.3. But, after that, a pkgbase update from 14.3 to 15.0 fails (it breaks pkg which segfaults). I'm just testing this on VM's, so rollback isn't a problem. 01:47:49 But, when it comes time to upgrade to 15.0, is the path going to be freebsd-install for that and then migrate to pkgbase or should we expect pkgbase to work by the time 15.0-RELEASE is announced? 01:48:52 This is just a general question, really. If anyone has been able to upgrade via pkgbase from 14.x to 15.x, I'd be curious what the process was (outside of the handbook). 01:50:22 s/freebsd-install/freebsd-update/ 01:52:05 It's going to take some time to re train my brain. 01:53:58 There will definitely be changes. But, they're not so bad. 02:29:24 ek: you need to use pkg-static for major updates at the very least 02:29:58 but yes, generally you should expect to freebsd-update to 15.0 then ?? from there 02:43:09 kevans: I am. Following the handbook for upgrade to 15.0. But, it fails. 02:43:16 Breaks pkg every time. 02:44:36 Bunch of missing libs. 02:45:09 Well, not missing. Just updated and pkg isn't linked to them. Both pkg and pkg-static. 02:46:34 I'm fine with a "freebsd-update" to 15.0. Just curious about 14.x that's already on pkgbase. 02:48:47 I have multiple 14.x using pkgbase already (I know, not technically supported.) But, when it comes to upgrade to 15.0, I'll apparently need to use freebsd-update anyway. 02:49:15 do you have BACKUP_LIBRARIES enabled? 02:49:38 pkg-static shouldn't have that same problem because it's statically linked 02:49:59 I do not. I'm just using a fresh install of FBSD 14.3, migrating to pkgbase, and trying to upgrade. 02:50:22 This is all VM testing. 02:51:52 If I do a fresh install of 14.3 (via ISO) a d then run "freebsd-update" to the latest 14.3 (-RELEASE-p4). Then "pkgbaseify" and then get those upgrades (which are newer). 02:52:03 Everything is fine. 02:52:52 But, when I try to upgrade to 15.0-ALPHA5 for "latest" main, pkg and pkg-static die with lib issues. 02:54:14 pkg-static is statically linked, lib issues aren't a possible failure mode 02:54:29 pkg-static dies with the same lib issues. 02:54:37 then it's not a lib issue 02:54:58 So, libs are being overwritten without pkg-static being updated. 02:55:06 pkg-static does not use system libs 02:55:26 kevans: Give me a second. I'll try again and show you what I'm seeing. 02:55:42 thanks 03:02:37 kevans: So, I change the ${LOCALBASE}/etc/pkg/repos/FreeBSD-base.conf to use "latest" instead of "base_release_3" (suggested by the handbook). 03:02:48 Then, I run: env ABI=FreeBSD:15:amd64 pkg-static upgrade -r FreeBSD-base 03:03:34 I get the errot about pkg not matching (which is fine) and accept. 03:04:01 Prompt for all the package updates and everything looks good. 03:04:59 Actually, I'll just pipe this to script so you can see everything. 03:05:31 I'm re-reading the above, and I think we need to clarify your wording here: is pkg-static actually dying, or is the upgrade failing? 03:05:52 pipe is good, though 03:09:31 kevans: So, I'm using both pkg and pkg-static after the failure. 03:09:58 Rolling back to the 14.3 now and starting from scratch after pkgbaseify. 04:34:14 kevans: Tried again, same core dump. 04:34:17 https://www.purplehat.org/~ek/pkg/ 04:34:48 Core dump (Likely not usable) and pkg log script log. 04:45:08 I wonder why -static would seem to exec non-static 04:46:30 the firdt crash is presumably the known bug when you agree to let pkg remove itself 04:46:43 i don't remember what version that was fixed in 04:47:09 i'm not sure that was ever fixed 04:47:20 https://github.com/freebsd/pkg/issues/2496 04:48:07 my sources say that might not have been addressed 04:48:09 =D 04:51:58 I think we don't have a good solution for 14 -> 15 via pkgbase alone until BACKUP_LIBRARIES is accounted for in the solver 04:53:09 You can carefully upgrade the kernel then pkg to the 15.x version, but we might not let that upgrade work without trying to upgrade base packages today 04:53:56 https://github.com/freebsd/pkg/pull/2524 helps here, spexifically 04:54:40 you would enable BACKUP_LIBRARIES, then you can probably get away without pkg trying to remove itself 07:35:01 magical beast called rust is here again, took 15h to build while whole world+kernel (includes clang) takes "just" 10h, ended up taking 17g to build and 1.2g to install and has some shlib that's 200m in size... might be super good highly secure praised dev env for real, but hell with needing ""windows 11" compatible" env to use it... and someone needs to do that too, despite dev might downloading from or 07:35:07 running it ... 07:35:09 ... in "cloud" 08:32:03 vim lib/libarchive/Makefile 08:32:10 er, wrong window 08:33:46 Password: 08:34:24 hunter2 08:36:46 i wonder if it's fortunate or unfortunate that reason i do c, inside or outside fbsd, is because i couldn't find any other way 08:37:20 sometimes couldn't even describe the issue 09:01:49 ugh rust 09:02:08 luckily rust people are mainly on Discord so I dont have to listen to them whining on IRC :P 09:02:35 Funny how the rust devs who moan about security use an insecure platform to yap on... 09:02:56 polarian: oh you don't like discord? could you tell me more about this? i missed it the last 37 times you mentioned it 09:03:09 ivy: get used to it :) 09:03:41 still trying to convince certain FreeBSD devs to use IRC 09:03:48 currently they are Discord-only 09:04:21 the only developer i'm aware of who is on discord but not irc is Mark Linnemon 09:04:45 probably because he mostly uses it for his weekly video thing 09:05:03 I spoke to stefano at EuroBSDCon and my friend did too 09:05:10 hes going to make his matrix bridge public 09:05:19 so fingers crossed... more people on IRC soon! 09:05:27 Stefano of bsd.cafe or a different stefano? i don't think he's a developer 09:05:48 bsd.cafe and no hes not, but he agreed with the problem :) 09:06:10 trust me to spread anti-discord propaganda 09:08:16 ivy: Is designed for kids use, not for decent people like you :) 09:08:24 *it's 09:08:56 *anti-WEB3.0 propaganda 09:08:56 mosaid: it's clearly not designed for kids use, they don't even allow children (although granted many children lie about their age) 09:09:13 i mean if people don't like discord that's fine, i don't either, but don't just make things up 09:10:48 I mean by kids "stupid people", not real kids.. 09:12:23 Discord is deigned in same way all social media designed (a lot of interactive stuff; to blow your mind) 09:12:36 mosaid: utter nonsense. the reason people use Discord is it provided a hosted solution for text + voice chat that non-technical users (especially gamers) find easy to use. that has nothing to do with people being "stupid", and also, using "kids" as a synonym for "stupid" is pretty offensive 09:12:58 mosaid: if you think people shouldn't use Discord, the onus is on you to provide something that does the same thing better 09:14:22 <|cos|> bashing languages and platforms on fictive parameters... isn't there like a #freebsd-offtopic or something? 09:16:25 ivy: sorry, you know my lang is not english, and I am trying my best to deliver the info.. After googling I think "Adult child" is the term which I was searching for 09:17:01 Not all Discord like that.. but at least 60% of it 09:19:40 this is not a language issue, your argument is just bad, discord is not primarily used by either children or stupid people, or "adult children". if you don't understand why Discord is successful, you are doomed to spend your days sitting here complaining about it and never doing anything about it 10:04:40 but we had, eg msn before, before it we had icq etc, actually, "mothers" and "sisters" used it, even i had to make account... now i have account in every developer and non-developer thing and sometimes i even use them eh 10:05:21 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=273566#c8 10:05:29 how to fix this? 10:54:23 hi 10:54:25 please 10:55:16 help, I upgraded to p3 to p4 (14.3) but some apps won't start (kicad) 10:57:50 what's the error message? 11:02:12 ketas: user input during loader is intended to interrupt it, in order to allow interrupting it at all in case a zero timeout is configured 11:02:13 <|cos|> 11:01 after upgrading from 14.3-p3 to 14.3-p4 kicad stopped working: Failed to load shared library '/usr/local/bin/_pcbnew.kiface': /usr/local/ffmpeg4/lib/libavcodec.so.58: version LIBAVCODEC_58 required by /usr/local/lib/libTKService.so.7.9 not defined 11:02:57 <|cos|> the above was posted on #freebsd-irc earlier. kicad launches on my machine, but i've never run it before. 11:03:30 that doesn't seem like it could possibly be caused by updating from 14.3p3 to 14p4 11:04:34 ketas: the "crap you're seeing" in console, does it also show up in dmesg? 11:05:42 Koston: invalid chars in dmesg? no, never 11:06:14 I mean the "gpioths0: err(INTERVAL, 40) = 60" stuff 11:06:45 oh yes, that's there 11:07:03 but i didn't paste it into shell 11:07:20 in fact i didn't even have cu running 11:07:37 and fwiw, cu(1) is really buggy, I highly recommend using eg. tio(1) or gnu screen 11:07:42 Koston: that can mean that there are loose contacts to that sensor 11:07:53 <|cos|> Posterdati: are you DemolitionMan, or do we have two reports on kicad breaking with 14.3p3 → 14.3p4 ? 11:08:03 i'm seeing it from time to time as well (my sensor is connected using "el cheapo" dupont cables) 11:08:03 cu should be fixed then 11:08:31 i'm not sure why anyone would use cu(1) nowdays, if you want to connect to a serial line, use tip(1) 11:08:38 yes i'm having it hooked up using unshieldes cheapest headers and kapton tape 11:08:39 although i hope anyone aware of bugs in cu(1) has reported them 11:09:02 right, tip(1) is also much preferred over cu(1) 11:09:22 any others reasons i wonder? 11:09:23 and tio(1) is a more modern take on tip(1) 11:10:15 or even comms/picocom 11:10:24 i never checked into driver what those errors were, i just waved them off to transient data errors from hw 11:10:53 i personally like kermit (comms/kermit) but basically anything is preferable to cu(1), which is intended to let users call a remote system using a modem bank 11:11:17 isn't kermit an independent serial protocol too? 11:11:44 hmm i never realized cu is bad 11:11:47 Koston: it has its own file transfer protocol (similar to XMODEM/YMODEM/etc.) but it also provides a general serial line interface 11:12:48 okay but another issue, how could running cu in one sc ttyv* emit any data into input buffer of another ttyv*? 11:13:05 ivy: neat. I faintly recall the name from decades ago, from actual modem days 11:13:31 and how does data self-echo back from machine when cu doesn't even run 11:13:49 are you sure you didn't have a serial client running? 11:14:21 when cnars came into other terminal, i had, otherwise no 11:15:03 also i would habe never noticed if of those issues 11:16:02 luckily chars don't come from closed serial to the console, that would be really bad 11:22:43 ivy: fwiw, I mostly agree with you wrt. Discord, but it does also promote much more dumbed down dialogue due to the way both servers and clients are at their default settings. compared to other modern social media though, it's still relatively smart. 11:22:43 at least one of those chars emitted from there is a bel, which is happily played out 11:23:22 despite nothing is ringing and nevermind into other "window" 11:24:18 Koston: that doesn't match my experience... i don't currently use Discord, but i've found the "threads" feature (which they stole from Slack) is pretty useful to allow in-depth discussions that can't easily happen on IRC. i've been in a couple of philosophy-related Discords and found this really helped to explore topics in depth. 11:25:34 I noticed that Newsboat is missing from FreeBSD 14 (amd64). Is it common with delays like this in FreeBSD? (I'm a new user) 11:25:36 for example, if we keep discussing Discord here, someone will probably tell us to shut up, whereas on Discord you can just create a thread and continue there 11:26:57 ivy: imho Discord is the best out of current widely used such services (vs. slack, teams etc), I'm just not really a fan of "autoplay videos/gifs by default" etc 11:27:38 newsboat is apparently console rss reader 11:28:04 Web feed reader to be exact. ;) 11:28:12 unsure why that would be missing in 14 ports 11:28:14 Koston: i believe there's a way to turnoff autoplay of gifs, perhaps hidden in "accessibility" settings 11:28:32 ketas: Khard is missing as well. 11:28:43 Both are used basically daily by me. :( 11:28:59 Hund: what "delay" are you talking about? does the package exist but it's too old? 11:29:05 (and I'm pretty sure that they're both quite popular) 11:29:18 dep on rust gmm 11:29:21 hmm 11:29:21 ivy: It exist in version 13. No version at all in 14. 11:29:43 I really don't care about "old" software, as long as it works. :) 11:29:49 freshports tells it's on 14 11:29:52 ivy: you can do a lot on both server and client side to enforce more civil and less spammy dialogue, but too much of it depends on latter because you should always expect the users to do absolutely nothing 11:30:25 ketas: I'm looking at freshports right now. It's not there? :| 11:30:25 Koston: right, but that's equally true on IRC. Discord can only provide the technical means, it's up to users to enforce policy 11:31:21 https://www.freshports.org/www/newsboat 11:31:23 boat 11:31:56 and it's not in repo? 11:31:57 ivy: irc doesn't provide official clients though 11:32:03 Koston: the FreeBSD Discord is terrible for this sort of thing, but that's because they refuse to moderate it even i suggested it several times, which is why i'm no longer there. it's not Discord's fault, it's the server owners' fault 11:32:27 yeah I left that discord long time ago for the same reason, lol 11:33:18 but discord is easy to follow? 11:33:41 all those images, emojis, sidechats, and so on 11:33:50 that a modern im brings 11:33:55 ketas: https://ibb.co/Z6gwCVDp ? 11:35:15 ivy: but the FreeBSD discord really *should* know better. since they can't do it right, how do you expect some average bunch of gamers to do it? 11:35:32 there are are also irc clients that willy happily fetch every url and embed it into chat, while only showing last 10 lines 11:36:16 Koston: i don't know, but you can't expect the platform to solve moderation, that's not their job. if there were "Discord Moderators" would tried to enforce this sort of rule, everyone would (rightly) be complaining about that 11:36:58 why do we ever need huge moderation anyway? 11:37:22 ketas: no one is saying we need "huge moderation" 11:37:24 ivy: I'm not talking about moderation, I mean things like someone posting a link and everyone else having link auto-preview on (by default) then gets half of their screen's worth of whatever the client thinks is appropriate as preview 11:37:35 sharing a simple link shouldn't mean spamming everyone 11:38:16 yeah that's hard to follow comms medium 11:38:40 That's why I like IRC and XMPP. You can choose whatever clients (and features that) fits you. 11:38:41 ok i've used that, sometimes it's good 11:40:44 ketas: What does freshports show you? 11:44:21 py311-khard having pkg's 11:44:36 i never checked the actual repo 11:44:59 at worse you can build it i guess 11:45:47 wait 11:46:03 amd64 11:46:11 was missing for newsboat 11:46:19 Yes 11:48:55 i wonder if it didn't build 11:49:03 but khard is there 11:49:23 Good. It was missing a few days ago the last time I checked. 11:51:08 khard is 3 months old, so? 11:51:28 |cos|: it was me, DemolitionMan = Posterdati 11:54:20 ketas: It's missing from quarterly. 11:54:38 Failed to load shared library '/usr/local/bin/_pcbnew.kiface': /usr/local/ffmpeg4/lib/libavcodec.so.58: version LIBAVCODEC_58 required by /usr/local/lib/libTKService.so.7.9 not defined 11:54:48 :( 11:54:51 I haven't dared switching to a version other than the default version of FreeBSD. 11:55:01 gmake: *** [Makefile:193: src/utils.o] Error 1 11:55:14 no wonder it's not there 11:56:04 https://pkg-status.freebsd.org/beefy20/data/143amd64-quarterly/24fedaeb4e97/logs/newsboat-2.40_1.log 11:56:24 Ohh 11:57:01 and khard has no logs at all? 11:57:06 no idea 11:57:14 i don't use either 11:57:44 Things like this always happens for me. :P 11:57:57 :p 11:58:19 but there are older pkg's? 11:59:07 i use ton of console stuff, but not those 11:59:14 I need to go back and check that out then. I'm on my phone right now. 12:00:08 You don't have contacts and calendars? :D 12:00:22 sadly this broke: https://www.freshports.org/astro/weatherspect/ 12:00:38 reminds me fun i used to run 12:00:57 i do, kind of, or not 12:01:11 or i don't console them :p 12:01:38 that error, unless reported, should be told to maintainer 12:03:41 I prefer using the computer for everything. :) 12:03:47 I'll look into that later. 12:09:34 i have no idea why python 3.11 port is not built tho 12:09:36 need a built-in command that will add a slight random delay to some cron jobs so they don't all run at the same time on every machine 12:10:39 if [ ! -t 0 ]; then sleep "`jot -r 1 0 10`"; fi 12:13:17 idea came from periodic 12:16:12 where i also found lockf 20y too late after implementing my own (oh hey, unix way) 12:16:23 dch: ^ 12:16:28 cool, today I learned jot(1) exists 12:16:36 lololol 12:16:47 cpdup, i also found recently 12:16:59 ketas: thats a good idea 12:17:26 now all my cronjobs have that bit 12:17:38 I think I need a tool called "jig" that randomly sleeps between 0 & x seconds 12:17:44 the ! stdin jot sleep 12:17:54 jig? 12:17:57 it's criminal how many simple but useful tools BSD has had since the dawn of time, but not included in POSIX or GNU 12:17:58 and if x is not supplied on command line, assume 60 s 12:18:03 its going to be a new tool 12:18:33 column -t 12:18:43 oh base is filled with joy 12:18:48 Koston: such as asa(1). although, this one is in POSIX, just no one knows how useful it is 12:18:56 column -t is my favourite "old new" tool since ages 12:19:06 I use it all the time from vim to line up stuff 12:19:22 from /etc/fstab to random data from zfs/zpool tools 12:19:41 ivy: I did not know about asa(1) either, thanks..! 12:19:46 i never found it until now 12:19:50 also, who knew about leave(1)? 12:20:02 i found so many random utilities trying to sort out pkgbase... 12:20:15 grdc? 12:20:18 pom? 12:20:20 :) 12:20:24 leave(1) is a funny one 12:20:35 ivy: clearly you're not making good use of leave(1) yet tho I see your 5am UTC commit this morning 12:20:45 ketas: i have at least two commits to usr.bin/pom :-) 12:20:52 why? 12:21:00 why what? 12:21:14 well i could look what they did 12:21:21 dch: fixed waking hours are for liberals 12:21:32 lmao ivy 12:21:37 pom, grdc were supposedly to be removed 12:21:39 and people with children 12:21:42 but unsure 12:21:55 in other news I have binge-eaten a packet of crisps 12:22:00 could be a port too 12:22:02 and I feel no guilt whatsoever 12:22:08 * ivy adds that to the list of reasons not to have children 12:22:22 should you? 12:22:28 feel guilt 12:22:54 dch: i assume you mean one of those big bags of kettle chips meant for several people, otherwise i think that's pretty normal? 12:22:59 i know i don't have any due being good eating thin 42yo boy 12:23:21 in my defence this bag is only 120g 12:23:23 does that count 12:23:29 s/is/was/ 12:23:32 having kids definitely compromises the ability to stay up every night until 5am 12:23:53 unless you have endless supply of methamphetamine and spare livers and kidneys 12:23:59 I guess 12:24:12 Koston: presumably if you have kids, they have livers and kidneys 12:24:24 touche 12:24:35 but those organs are like in use... 12:24:40 :p 12:25:16 so tip is better than cu? 12:25:34 i looked both 12:25:48 ketas this is why I have several kids, its like a Redundant Array of Insolent Children 12:26:10 :p 12:26:24 what we really need is some sort of subscription organ service, like Netflix for kidneys 12:26:36 or livers, whichever the one is that explodes if you drink too much 12:26:44 yeah like all gals ran away so how do i even make them now... 12:26:45 you could borrow one for a big week and then return it 12:27:42 oh and why they ran, because i forgot them over fbsd 12:28:31 Back.. ivy: yeah, I am just saying that my expressions isn't the best (I was trying to say: Adults behaves in a stupid and irresponsible way) sorry for the bad argument.. Discord is bad not for it's features, but the way it works (nearly all the WEB3.0) on the psychological/mental and technical levels.. We can talk about that for days, but let me explain the simplest level which is the... 12:28:33 ..."technical", and this question will answer nearly half the thing: What is the best? taking the info in it's plain state/arranged in a pretty simple way.. or that bloated interface, makes you consume more computing power and brain cells just for simple audio/video and text chat. That what makes everything goes more back, not forward.. this the same for all the social media (many useless... 12:28:34 ...stuff here and there, makes taking the simplest info hard; So the focus on the value of thing goes less.. and this normal now days, companies is just making money ;) And IRC VS Discord, is like Local Market VS Mall) I can't even run discord well here on SeaMonkey. I don't really have time, but surly it could be there a something like improved IRC: Neat, Organized, More options and maybe... 12:28:36 ...hosted.. but still simple and can run nearly anywhere.. 12:28:37 Also, for moderation.. I want you to know that many Discord people have it's own way to behave vary from IRC; That explains why the moderation also different.. I know there is a real production communities there, but I am just talking about the most I saw.. and most my talk in end is on discord itself 12:28:54 i am not going to read all that, sorry 12:29:00 This only my opinion 12:29:22 ivy: XD 12:31:13 mosaid: if only world worked that way, but people *want* instant gratification type media 12:31:59 Yeah, and companies feeds that for more money 12:32:13 kids these days, with their livers and kidneys... 12:33:01 recently after a few interviews, I told a headhunter I will not be using a webcam because it's counterproductive and I've concluded the best way to assert this is just to never use one. he told me it was a requirement for getting hired, so bye bye. 12:34:08 dch: how much do you care about D53016 and D53015? i think they are not critical (it's like ~2MB in total, right?) but i can land them now if you need it 12:34:51 mosaid: yes! 12:35:14 however i don't even use ""social" social media" thay way 12:35:34 ivy: its less about the size and more about understanding where they come from 12:36:03 I'm fine if they only go into BETA2 but its marginally more work for cperciva to cherry-pick back 12:36:04 ivy: please could you take a look at this tool, I want to use it https://vermaden.wordpress.com/2021/04/02/ufs-boot-environments/ 12:36:19 dch: yeah, i think we sorted that though, right? :-) i'm just wondering what you want from me (if anything) since i'm trying to track pkgbase issues for 15.0 12:36:30 ketas: I don't even have a smartphone.. 12:36:40 ivy: nothing other than my gratitude from making it happen 12:36:44 and I will never thing of buying one 12:36:47 if the answer is "nothing" that's fine, i'll just check with Colin when the next release is... i'm actually not sure what's happening with BETA1 12:36:57 wasn't that meant to come out this morning? 12:36:59 *think 12:37:05 i have them, running connectbots 12:37:16 i find phone situation sad 12:37:17 mosaid: no, i am far too busy and don't really care about UFS boot, sorry 12:37:28 mosaid: but it seems like a good idea, so feel free to post on a mailing list or something 12:37:40 No, problem I will try it 12:37:47 hw is ok, sw is ok, but politically phones suck 12:37:57 I installed, and I think it will work 12:38:48 there has never been a technical reason we don't support UFS BEs, it's literally just that no one wrote support for it 12:38:59 I will try simple install just boot, kernel and the necessary stuff 12:39:16 ivy: its this morning colin time, so our afternoon 12:39:23 That tool will really help in my May upgrade to 14 12:52:10 seems to rely on rsync for copying filesystem contents to another partition 12:54:24 Yeah 12:54:36 as long as you understand the limitations of doing that, should work fine 12:55:02 rsync is gold, then you simply chroot and do what do you want.. 12:55:32 rsync is not in base, so any implementation of UFS boot environments would need to not rely on that 12:55:58 it's just a simple script 12:56:04 not full thing 12:57:29 I can't be arsed to read the whole script, but I don't see it remounting source disk read-only or other means of guaranteeing data consistency of the copy 12:57:56 that shouldn't be a problem in most cases, but you need to understand what that means (vs. ZFS snapshots) 12:59:30 Yeah, I will read more about till may 12:59:42 I am still new to ZFS 13:01:08 these days Just Use ZFS™ unless you know well why you shouldn't 13:01:19 koston: also remember my Home is not any more on /, so it's safe for me 13:01:52 your home directory is least of your concerns; contents of that shouldn't change at all over OS upgrade 13:02:41 huh, please explain more.. 13:03:27 just read the Handbook first, then ask if you don't understand something specific 13:04:21 mosaid: freebsd-update will change the kernel, userland tools, and optionally source & ports tree. but it does not change your user files. 13:04:37 it only updates the bits that were installed via the FreeBSD installer. 13:04:44 ah I know that 13:05:01 but why you saying that my home directory is least of your concerns 13:05:11 Is this something bigger than that 13:05:22 no, just that the upgrade won't change your homedir 13:05:30 I know 13:05:54 I am saying if the partition corrupted, I could install a fresh version easily 13:06:02 root "/" 13:06:34 partition itself doesn't become corrupted as such, but if the system files change during rsync copy, then the copy could end up being inconsistent with itself 13:06:52 typical problem when plebs try to backup databases by copying the actual db files 13:10:57 anyway, I gotta get ready for elderly nerds mtg session, haven't played that game since the 90s.. 13:14:04 Koston I will surly test that copy before the upgrade, I take backups of all my configs also before it.. I am saying if the fresh install became the easier and time safer way, I could do it 13:22:46 mosaid: I think you would love zfs boot environments for upgrades 13:25:29 I took 3 months holiday from my work; to fix many stuff, includes my PC, which even my native Lang layout was not working well on it, also never updated since 13.2 no pkgs, rely only on ports for new apps.. hardware problems (never changed my PS/2 keyboard for about +15 years, it became useless; so got a new one).. a lot of stuff and the last upgrade took so long time from me; bec I was... 13:25:30 ...using 13.2 and I jumped to 13.5, many of my old ports was not working well, need to fix. (and my forgetting about setting sys.cons 3 years ago, took the most of time). So thinked boot environments will safe a lot of time for me 13:28:34 dch: could I use it only for 2 partitions or it needs the whole hard drive? 13:28:54 mosaid: its simplest to use when you give it the whole cake 13:29:05 :( 13:29:07 so I would start off with a full zfs default install 13:29:38 once you know your way around you can split things up 13:29:58 So I would use it only for my new drive, and I will continue using UFS for / 13:30:14 then you can't use boot environments 13:30:21 Yeah 13:30:33 so I use thinking about that script 13:31:16 the UFS stuff linked to above is a half-arsed workaround for what's built into freebsd + zfs since 9.3. 13:32:26 Yeah, I know.. but I can't move all that huge data on my hard drive which continues / anywhere now.. 13:33:21 * dch nods 14:02:57 what's ufs "be"? 14:03:02 i have them 14:03:10 i wish it would be better 14:07:21 i think everyone who needs multiboot, has rolled their own 14:07:48 or they just use multiple machines which are even better 14:23:47 oh right otis was the gpioths dude 14:26:31 wait it wasn't 14:28:13 was Him 14:29:01 actually severals did it 14:39:39 actually i switched from vt to sc because terminals flickered on every switch 14:40:00 maybe should have tried to find better vt config 14:41:05 but still no idea why data would go back on closed port or how it jumps terminals 14:50:35 could test more 15:09:16 btw how to even use tip with device name :) 15:09:41 i did use others before when i needed *modem 15:53:01 btw obsd has new cu 15:53:24 import that in or something 15:54:54 but it does come with drawbacks 15:57:38 er, what's wrong with cu? 16:03:21 supposedly it's bad and buggy and source of all my problems :) 16:08:08 hi 16:08:49 please help, I upgraded 14.3 from p3 to p4 and kicad stopped to work: Failed to load shared library '/usr/local/bin/_pcbnew.kiface': /usr/local/ffmpeg4/lib/libavcodec.so.58: version LIBAVCODEC_58 required by /usr/local/lib/libTKService.so.7.9 not defined 16:08:58 thanks for any help 16:10:24 Posterdati: im guessing something bigger than just a p3->p4 patch happened 16:10:47 zi: like what? 16:11:09 port/package version changes? 16:11:46 also, it doesnt look like you installed ffmpeg from packages/ports 16:11:50 so that's gonna be a problem as well 16:11:57 no, I did not 16:12:10 I used pkg 16:12:37 it must be some third party package then--ours installs that library to /usr/local/lib/libavcodec.so.58 16:13:08 is there 16:13:29 then why is whatever _pcbnew.kiface is looking for it in /usr/local/ffmpeg4/lib/libavcodec.so.58? 16:13:50 I do not know 16:13:58 is kicad installed from ports/freebsd packages? 16:14:02 no 16:14:05 using pkg 16:14:51 pkg --version --> 2.3.1 16:15:17 what does ls /usr/local/ffmpeg4 show? 16:15:55 bin include lib libdata share 16:16:21 it seems like you built ffmpeg from source instead of using the package or port 16:16:33 no, I used pkg 16:16:44 pkg doesnt install that to /usr/local/ffmpeg4 16:16:46 where'd you get the package? 16:17:27 again I only use pkg 16:17:37 im gonna try one more time before i give up 16:17:41 pkg check ... it 16:17:51 freebsd packages for ffmpeg do not install stuff to /usr/local/ffmpeg4 16:18:06 so you either got it from not the official ports tree/freebsd package repo 16:18:11 or something else is going on 16:18:45 (like you built it from source yourself) 16:18:46 pkg check ffmpeg4 -> Checking ffmepg4: 100% 16:18:48 have you make-installed non-ports? 16:18:56 no 16:19:06 pkg which /usr/local/ffmpeg4/lib/libavcodec.so.58 16:19:13 i tend to got for own prefix if ever do 16:19:40 hmm 16:20:09 and also: pkg which /usr/local/bin/_pcbnew.kiface 16:20:24 -> /usr/local/ffmpeg4/libavcodec.so.58 was installed by package ffmpeg4-4.4.4_25 16:21:07 there's something super funky going on here with this box 16:21:27 also, your packages are wicked out of date. that 4.4.4 release is from 2023 16:21:29 i have it in usr/local/lib/libavcodec.so.60 16:21:42 -> /usr/local/bin/_pcbnew.kiface was installed by package kicad-9.0.4,2 16:21:42 right, it should be in /usr/local/lib 16:21:43 and 6.1.2_16,1 16:22:30 Posterdati: yea your packages arent in sync with eachother 16:22:41 Posterdati: as ketas mentioned, latest ffmpeg is 6.1.2 16:22:44 p3->p4 won't do that 16:22:52 so you somehow have kicad from 2025, ffmpeg from 2023 16:23:21 why is the ffmpeg old? 16:23:29 i would pkg delete ffmpeg4 (which may uninstall some other packages). then you'll want to: pkg ffmpeg4 kicad again and see if that fixes it 16:23:46 pkg search ffmpeg returns also ffmpeg-6... 16:23:53 if it's not on purpose, you should upgrade it all 16:24:08 i doubt it was an official freebsd package in the first place 16:24:14 zi: it will disinstall all kde 16:24:31 what's pkg info for it? 16:25:33 what are all other outdated packages? 16:25:53 and why would it break kde 16:26:09 i mean kde has deps in that i guess 16:26:15 Posterdati: bottom line is your installed packages arent in sync with eachother. this probably means you upgraded a couple of packages at some point, but not all. so, you'll need to get them all in sync with eachother 16:26:47 so you either pkg delete -f it which can break part of kde 16:26:58 also how did all this happen 16:27:00 pkg install kicad ffmpeg 16:27:05 it tries to install 16:27:17 ffmpeg 6, ffmpeg 4, kicad 9 16:27:18 Posterdati: running pkg upgrade -f will at least get everything in sync but you could run into other surprises with version changes 16:27:45 save pkg info & pkg info -qoa outputs if needed 16:27:55 before anything 16:28:18 could save others, could even backup all, if needed 16:28:43 but now the kicad is already broken :) 16:28:52 it is installing 1088 packages :) 16:28:57 eh 16:29:04 upgrading? 16:29:12 pkg upgrade -f 16:29:20 k 16:29:31 at least everything will be current/in sync after that finishes 16:29:33 unsure if -f was needed 16:29:33 then you can retry kicad 16:29:40 but yeah 16:30:16 I unsed pkg upgrade only, without -f 16:30:19 in the past 16:30:22 ketas: freebsd 14.3 didnt exist in 2023 so 16:30:42 he probably had a mix of packages built against multiple different freebsd releases 16:30:49 hence needing the -f to correct that 16:31:03 never upgraded anything ever? 16:31:23 and by accident? 16:31:30 some stuff got upgraded but only by dumb luck with version changes 16:31:52 so some things were built against freebsd 14.3, some 14.older 16:32:03 perhaps 14.0, which was released in 2023 16:32:11 that's all fine until libs crap up 16:32:26 hence -f 16:32:26 :) 16:33:02 i did bit better thing, built ports fot armv7 in 13 and used 16 base with it 16:33:14 that needed openssl symlinks 16:33:31 but that was on purpose 16:42:05 still have no idea how stuff ended up in /usr/local/ffmpeg4. even the pkg-plist from the 4.4.4 port shows those libs in /usr/local/lib where they should be: https://cgit.freebsd.org/ports/diff/multimedia/ffmpeg4/pkg-plist?id=f0dadc529fe9c4ee69dd96db92e2d0572d2092aa 16:42:07 Posterdati: still going? 16:47:15 finished downloading :) 16:47:21 checking integrity... 16:47:25 rgr 16:48:21 will it upgrade ports too? 16:48:31 depending on what you're running, it might make sense to reboot when you're done to ensure what's running is in sync with what is now on disk 16:48:52 I used ports only for sbcl, which is now a binary package :) 16:49:03 done (0 conflicting) 16:49:19 yeah--that will install the latest versions of everything from the freebsd package servers 16:49:34 good 16:49:45 this is an old i5 16:49:58 it has got a second life with FreeBSD on it 16:50:04 just for the future, you should get in the habit of doing a pkg upgrade -f anytime you change freebsd versions (so 14.2->14.3, etc) 16:50:10 I can use kicad too! 16:50:33 I did freebsd-update fetch 16:50:39 and then install 16:50:49 after that pkg install without -f 16:50:51 yeah but at some point in the past, you did freebsd-upgrade 16:50:56 and then didnt do a pkg upgrade -f 16:51:03 never used -f 16:51:54 i know 16:53:03 ffmpeg4 shouldn't be installed I think 16:53:05 but 6 16:54:27 once the upgrade is done, you can try pkg delete ffmpeg4--it will complain if something depends on it 16:54:40 but id probably just restart and then try kicad when the system boots back up 16:55:20 Reinstalling all 16:55:29 [7/2980] 16:55:33 lol 16:56:11 damn dude 16:56:39 pkg install kitchen-sink 16:56:45 :-) 16:58:32 [17/2980] 16:58:34 lol 16:58:43 go123 16:58:59 I do not remember to ever used go language :) 17:00:44 likely something else you installed depended on go 17:01:26 sure it is 17:01:55 I played oolite too on that laptop 17:02:09 an old k55v :) 17:03:05 curl -L -s https://dl.google.com/go/go1.24.8.freebsd-amd64.tar.gz | tar -x -C /usr/local/opt -f - && /usr/local/opt/go/bin/go version 17:03:49 aic: there's also a go124 here :) 17:19:06 [347/2980] 18:03:36 zi: [1449/2980] :) 19:04:45 zi: [2186/2980] 19:41:59 Posterdati: hah 20:16:21 I'm ready for the next update. Gotta be wrapping up at some point, though a conflict could have arisen and the whole thing needed to start over 20:47:51 anyone has configured IKEv2 who can help me figure out why I can't see traffic in plain when I tcpdump enc0? I'm terribly stuck at this... 20:52:42 antranigv: dont suppose you've skimmed `man 4 enc`? 21:21:30 zi: fixed! Thanks 21:21:56 zi: pkg upgrade -f now asked me to upgrade 1625 packages 22:57:33 Posterdati: welcome. you dont want to randomly run it--it will always ask to do stuff 22:57:48 Posterdati: only after bigger OS upgrades 23:04:08 antranigv: I've set up IKEv2 for a "VPN provider" thingie and I don't understand your question.