00:00:20 Don't make things more complicated than it needs to be. 00:02:03 hello I am looking for help on mail 00:02:39 email mail program and sending mail usually was sendmail now dma I don't know what to do also fetch mail I think it's configured somewhat right but this whole thing is not really working 00:07:25 anyone? 00:08:43 hello sorry I disconnected 00:08:47 I'm going to try again later 00:08:59 maybe it's a very extensive topic and will require a lot of reading 00:09:43 ok thank you be around later 00:11:28 rwp: hi thanks for the autossh article, i've bookmarked it :D 00:13:13 my own use case is rather simpler but yeah it's a very useful thing to have for remote admin 01:44:55 f451, Obviously I wrote that up for Debian but the same technique works great on FreeBSD too. 01:45:10 I am also using the spiped setup for servers to connect their email up to my server where I can get email notifications from them now. Since most cloud VPS hosting vendors block port 25 by default (a good thing). 03:15:30 https://youtu.be/q4PNSeNXEww?si=uqN1CFnBkEFj7L9G 03:16:07 Installing every window manager on FreeBSD 03:48:20 Ah. First base updates with pkgbase and ansible, and everything worked perfectly on first try 04:26:43 lts: itll break then youll come in here saying thats it broke and that your hate freebsd 04:26:49 lts: happens all the time 04:28:43 I'll beat that horse when it gets here 04:28:55 the horse is too big 05:55:16 that's what she said 06:09:02 rtprio, not appropriate for the desired environment 07:09:46 I apologize 08:41:10 rwp, "Who is trying to get autossh working and having problem"> it was me, but f451 has helped me solve them. And I have bothered the maintainer for no good reason. 08:43:24 rwp, many thanks for the article on /Setting Up AutoSSH/ -- it looks very good, will read. 08:45:44 rwp, I did not know SSH tunnel could user HTTPS, because it is picky about mismatching HTTP headers and actuall addresses. 08:59:01 I think my favourite part about 15.0-RELEASE is that you get to spin a roulette wheel on boot in Hyper-V 08:59:16 today it took 12 loops to boot the VM without a kernel panic 09:00:34 very thankful I do not need to *install* it in a VM right now because that is close to impossible (since hv_hid gets loaded automatically and that is a 100% crash) 09:04:07 is it expected that updating a non pkgbase system doesn't get a patched kernel but a pkgbase system does? 09:04:51 (on 15.0R) 09:12:13 * Remilia goes to check hers 09:14:20 markmcb: I think it gets weirder, I just updated my pkgbase poudriere system and rebuilt my custom kernel and it is still tagged 15.0-RELEASE 09:14:49 oh wait 09:14:58 the *kernel*, on 15.0 09:15:40 markmcb: that is probably expected behaviour 09:16:33 not sure why zfs.ko does not get updated though 09:22:52 markmcb: 15.0 kernel is only affected by the ZFS issue which usually means the kernel blob itself is entirely unchanged, so the kernel version stays the same 09:26:01 re: zfs.ko not updated: oh I am an idiot and it is my local problem 09:42:43 actually I didn't notice vmm 09:46:10 markmcb: just checked and your freebsd-update'd system should have the exact same files in /boot/kernel, with the same kernel version string, as before, with the exception of linker.hints, vmm.ko and zfs.ko 09:46:39 freebsd-update does not touch unchanged files 09:47:08 while pkg removes old files and installs new, replacing the full kernel set 09:48:05 * Remilia goes to reboot the server for p1 10:08:56 thanks Remilia. haven't dug into binary differences yet, but there was definitely a kernel package update and freebsd-update shows -p1 kernel running after reboot. non pkgbase system shows no change to kernel, only userland. 10:09:40 markmcb: ls -l /boot/kernel, you will see that the affected modules were updated 10:40:40 yeah, my point is freebsd-version -kru show different things on two fully patched 15R systems 11:43:50 netcup: 'power cycle your server to take advantage of new KVM features!!' 11:43:57 also netcup: resets EFI vars 11:45:56 fun times today, everyone upgrade FreeBSD-15.0 to p1 for your shiny new rtsold without an RCE 12:22:11 at least updates are quick now with pkgbase :) 12:24:09 the main question still stands... how do we pkgbaseify old setups? 12:33:47 with pkgbasify? :) 12:40:58 where you convert a regular FreeBSD system to pkgbase 12:43:14 its a tool 12:45:29 antranigv: https://github.com/FreeBSDFoundation/pkgbasify 12:45:53 antranigv: its very good(tm) I did all my pkgbasey stuff with it 12:46:06 make sure you have the lastest pkg installed beforehand 12:46:09 anybody running 15.0-RELEASE and /latest/ pkg branch 12:46:22 then it should run fine from 14.3 12:46:27 last time I ran it on my server it crashed my server. luckily I was able to revert the BE 12:46:28 dch: yes 12:46:37 but I will try now 12:47:08 I'm getting "pkg: sqlite error while executing grmbl in file update.c:154: NOT NULL constraint failed: packages.path" https://www.irccloud.com/pastebin/JMt3JcdJ/pkglolwat 12:47:27 leah2: can you do a pkg update and see if you get the same error please? 12:47:31 sure 12:47:36 I'm on pkg 2.4.2, amd64 12:47:53 `pkg update -r FreeBSD-ports` should be sufficient 12:48:38 I don't see this error on the arm64 systems so I'm guessing something new is borked; the arm64 builders are usually behind a bit 12:48:44 indeed 12:48:56 https://l2.re/kViXnE 12:49:17 ta, I do a PR for it 12:49:24 curiously the upgrade earlier worked fine tho 12:49:37 let me try latest pkg from src first tho 12:49:49 yes its new last couple of days I think 12:51:24 nope its present in pkg 2.4.990.3-c642c9442 too 12:52:08 so... pkgbasify a BE, not the running syst 12:52:37 MelMalik: pkgbasify will do a BE for you 12:53:06 and check you have enough disk space *cough* 12:53:17 I still need to figure out how to prevent pkgbasify from giving me Everything 12:53:31 just remove packages later 12:53:38 and ensure you're using latest pkg (2.4.2) first as well 12:53:40 very important 12:53:51 yeah i mentioned that already 12:53:56 pkg: sqlite error while executing grmbl in file update.c:154: NOT NULL constraint failed: packages.path 12:53:58 thats weird 12:54:00 as if I would know which packages to remove because removing the -set- ones does not lead to autoremove doing anything 12:54:28 do I need to purge my pkg db, or is this the remote db which is being fetched having an error 12:54:43 shouldnt have corrupted the local db 12:54:44 dch: oh wait you literally just said it 12:54:50 xD 12:54:54 I think this is a remote db issue then 12:55:18 the "remote db" is literally a huge wodge of jsony-yamly-goop 12:55:27 so its a pkg issue 12:55:30 oh right 12:55:37 I didnt update pkg in the last few days tho 12:55:48 last update was Dec 5 12:55:48 that issue happens for me too when poudriere tries to fetch repos 12:55:56 and it worked fine until today... 12:56:04 weird... a sleeper bug :p 12:56:21 waited a few weeks before making itself known 12:56:41 I dunno it feels like a repo issue to me 12:56:51 because pkg works just fine with my own repos 12:57:02 Remilia: well thats what I thought but dch is a core developer so I assume he knows a lot more than I do :p 12:57:31 only the first part is true, just because I have a pointy hat doesnt mean theres anything under it ... 12:58:11 wait 12:58:18 I think I know what happened 12:59:03 yeah I am pretty sure this is a case of double repo conf 12:59:07 at least for me 12:59:52 Remilia: not for me, I checked my /etc/pkg/FreeBSD.conf 13:00:00 I have a /usr/local/etc/pkg/repos/FreeBSD.conf with url: "pkg+http://pkg.FreeBSD.org/${ABI}/latest", 13:00:19 and /etc/pkg/FreeBSD.conf with FreeBSD-ports url: "pkg+https://pkg.FreeBSD.org/${ABI}/quarterly", 13:00:32 one would surely take priority over the other? 13:00:56 they are different 13:01:10 one is FreeBSD-ports and the other is FreeBSD so it tries to update both 13:01:41 anyway experiments showed that it's latest with issues 13:02:03 quarterly updates fine 13:03:06 quarterly is behind though 13:03:07 I guess no conflicts there because it just gets two different repos for ports, I disabled the old one (I think FreeBSD was renamed to FreeBSD-ports?) 13:03:25 I don't really care about it being behind as long as it has working poudriere 13:04:01 thus, solved my issue for now hahaha 13:04:07 Oh others are seeing the same pkg error, thank dog. Thought I was going to have to hand-repair pkg databases 13:06:12 is anyone else unable to update the latest FreeBSD-ports repository on a FreeBSD 15.0 system? 13:07:08 crest: https://github.com/freebsd/pkg/issues/2575 13:07:11 no 13:08:00 dch: so you have the same problem 13:08:00 alright buildworld time 13:08:10 Remilia: use `pkg repositories` to see what all the resolved pkg repos are 13:08:18 quarterly still works, but i don't want to downgrade 13:08:18 for those using IPv6 please bare in mind there is an RCE in rsold 13:08:31 polarian: link? 13:08:40 dch: https://www.freebsd.org/security/advisories/FreeBSD-SA-25:12.rtsold.asc 13:09:12 ta. aah cool "from systems on the same network" 13:09:27 yeah it couldn't be WAN 13:09:31 as your router would drop them 13:09:38 ~hopefully~ 13:09:51 for the moment, my kids are not at this level of hacking sadly. But one day! 13:10:20 still something I wont leave unpatched, takes one compromised device on the network to then use the RCE to pwn you :p 13:10:29 also public wifi... 13:11:01 polarian: I thought FreeBSD SAs were posted to mailing lists and RSS feeds 13:11:04 well just because the vuln exists doesn't make it exploitable, but again not something to leave up to chance 13:11:07 Remilia: they are 13:11:15 I got the email from the security mailing list last night when I was going to bed 13:11:48 I have just booted my laptop this morning to find pkg not working, and was updating my src tree to buildworld/buildkernel 13:11:52 so did I and people have been posting about it here too hmm 13:12:13 I guess we'll just see it repeated for the next few days haha 13:13:16 @polarian: FreeBSD-SA-25:12.rtsold reads as bad as ShellShock *yikes* 13:14:10 crest: I was going to bed and saw "RCE" and thought "oh shit" until I realised its router solicitation and that is LAN-only 13:14:25 and my laptop is on a separate VLAN 13:14:31 and your wifi? 13:14:32 I mean how many ipv6 enabled systems do you know 13:14:42 do you never connect to any untrusted wifi networks? 13:14:42 Remilia: all my systems :0 13:15:02 crest: of course... but I am at home right now, so I am fixing it, but if I was not home I would have been "oh shit" 13:15:16 if say, I was attending a conference, I would be concerned 13:15:39 but you can also disable router solicitation in rc.conf while patching and enable it after 13:16:29 or install dhcpcd 13:16:43 meh 13:34:36 slaac systems unaffected yeah 13:38:12 most interfaces have nd6 options=23 but rtsol/rtsold isn't running on their systems or what they connect to 13:47:11 f451: router solicitation is needed for SLAAC 13:47:13 afaik 13:47:46 the only time you dont need router solicitation (and advertisement gateway side) is when you are statically assigning IPv6 13:49:09 neighbour solicitation and advertisement is always required for the functioning of IPv6 as its a requirement for NDP 13:50:17 so if it's using ipv6 it'll be using rtsol at the very least? 13:50:35 f451: do you have it enabled in /etc/rc.conf? 13:50:37 it is a daemon 13:50:51 no its not enabled 13:50:55 then you dont need to worry 13:51:33 but if you are using slaac you really should use solicitation as its used to request network information from the router 13:52:14 rtsol seems to be misnamed from how I understand Background and Workaround in the SA, it reads like rtsol handles Router Advertisement if ACCEPT_RTADV is set on the interface 13:52:16 the only ipv6 stuff in my rc.conf is ifconfig_bridge1_ipv6="inet6 -ifdisabled auto_linklocal accept_rtadv" 13:53:02 nimaje1: afaik it handles the router advertisements, and sends solicitation 13:53:15 but I haven't read the source, so I cant say for sure 13:53:50 nimaje1: it is also possible that accept_rtadv passes to rtsol if rtsold is not running? 13:53:54 infact slaac isn't running either ;) 13:53:56 the rtsol part and what I understand rtsold_enable to enable is just sending the Router Solicitation messages 13:54:31 nimaje1: welp from the SA, it states if you have accept_rtadv you are vulnerable 13:54:51 so I would assume rtsol/rtsold is in use in some shape or form if a interface has a the accept_rtadv option 13:54:56 i patched anyway 13:54:56 s/option/flag/ 13:55:07 buildworld still running for me :p 13:55:49 i wanna call what rtsol does chirping because i'm a dork 13:56:04 i just patched then make install in /usr/src/usr.sbin/rtsold 13:57:15 f451: theres also erattas to zfs and vmm 13:57:35 but because I use my /usr/obj for jails, I rather have a complete world 13:57:53 vmm one doesnt affect me cos i dont use passthrough 13:58:03 need to look at the zfs one 13:58:06 meh I dont bother reading erattas I just patch them :p 13:58:17 I only tend to read security 13:58:27 ~even if it doesnt affect me~ 13:58:54 i read em all if i can cos im kinda ocd about it 14:00:03 yeah, I didn't find documentation how Router Advertisement are handled, but as I understand rtsold_enable, that part is just about sending Router Solicitation messages 14:00:49 f451: I did think that buildworld preserved /usr/obj on patch releases 14:00:51 but obviously not 14:00:55 I was decieved 14:01:16 :D 14:01:20 yeah 14:01:26 I wonder if I had make in /usr/src/usr.sbin/rtsold if it would have wiped my obj too 14:01:29 as differing timestamps 14:01:37 prob would have :/ 14:01:41 ive been bitten by preserving obj. and by meta mode 14:01:54 meh, I am going out for a pint soon, will leave makeworld going 14:02:00 it makes updating my jails really easy too 14:02:07 best to blat /usr/obj. i know it takes more time 14:02:44 I will provision a build server at some point, I have lying around resources, just lazy :p 14:03:40 going from 14 to 15 especially - blat obj ccache meta modr turn off filemon move src.conf asige build generic 14:03:59 yeah but major releases always need a fresh obj tree 14:04:04 also install compat14x 14:05:03 i broke everything all the different ways the first time. thankfully on a disposable vm 14:07:08 f451: also you said you compiled rtsold 14:07:14 you didnt compile rtsol? 14:07:19 which is also vulnerability 14:07:30 was just thinking that 14:08:12 they are in the same dir 14:08:41 ah, should have looked a bit lower in rc.conf(5) rtsol handles Router Advertisement messages for interfaces which have accept_rtadv 14:09:13 f451: and this is why I just buildworld :p 14:09:36 nimaje1: HAH 14:09:38 I knew it! 14:09:51 I assume with rtsold is running rtsol passes it to rtsold 14:10:15 oh wait nvm 14:10:17 no 14:10:46 nimaje1 was correct in their assumption, rtsold is exclusively handling sending router solicitation, while it appears rtsol actually only handles router advertisement 14:10:53 so the whole "Misnaming" thing makes sense 14:10:56 rtsol is actually rtadv 14:11:07 polarian: the Makefile in the rtsold dir controls both rtlol and rtsold 14:11:17 f451: ah alright 14:11:26 I haven't inspected the repo, I was just asking :p 14:11:53 well, it isn't really documented, just rtsol_flags saying it is used for rtadv 14:16:17 i dont see how the rce should work tho 14:18:19 leah2: how so? 14:18:29 could the duplicated CuraEngine-debuginfo package be caused by a bug in sub package support? 14:18:44 nimaje1: maybe something to bring up with the docs people then? 14:19:03 one of the duplicated CuraEngine-debuginfo packages has no path attribute as noticed by @dch 14:53:12 crest I wondered but there’s nothing I can see in https://cgit.freebsd.org/ports/tree/cad/cura-engine/Makefile about sub packages am I right? 14:53:30 not directly 14:53:31 like where does this -debuginfo suffix even come from? 14:57:38 has to be from something in Mk/Features/debuginfo.mk 14:58:12 dch: did you do something, pkg upgrade works now? 14:58:18 oh wait nvm! 14:58:36 i think the WITH_DEBUGINFO=yes line enables this 14:58:58 ups 14:59:00 44:WITH_DEBUGINFO= 1 # Will segfault if Release 14:59:05 in cad/cura-engine/Makefile 15:01:10 aaah I missed that probably because I read it on my phone 15:17:40 but WITH_DEBUGINFO is something for a ports user to set, not for a port to set 15:22:29 wait 15:22:30 https://cgit.freebsd.org/src/commit/?id=6759fbb1a553 15:22:33 does that port maybe want to hard code CMAKE_BUILD_TYPE instead? that seems like something a port is allowed to set 15:23:24 - /* TODO: remove security bug */ 15:23:24 + if (security_bug) return; 15:24:32 nimaje1: could be an env var name collision between the ported software and the port 15:24:40 wrt `pkg update` issue, this has been fixed on the main build server, and we are waiting for this to trickle out to mirrors atm 15:27:52 https://github.com/freebsd/pkg/issues/2575#issuecomment-3665793754 15:31:27 if somebody who does reddit wants to update the thread over there that would be a great help 15:34:04 it's not a variable name collision because the string [DEBUGINFO](https://github.com/search?q=repo%3AUltimaker%2FCuraEngine+DEBUGINFO&type=code) doesn't appear in the upstream repo 15:34:36 so this port is abusing the WITH_DEBUGINFO ports feature to get specific build flags 15:46:49 Other than using some variant of -g and some variant of the linker / splt command to break out the debuginfo file how is this port abusing WITH_DEBUGINFO? 15:47:40 the a comment in the port makefile says that Release builds of Cura-Engine will segfault 15:48:02 WITH_DEBUGINFO= 1 # Will segfault if Release 15:48:05 in line 44 15:48:40 this should be harmless 15:49:05 but because of the problem in the ports tree infrastructure it isn't 15:49:18 instead of just generating one more package 15:49:38 so three instead of two 15:50:23 There should be other ways to fix a segfault 15:50:57 and one of the two "CuraEngine-debuginfo" packages lacks the required "path": "" attribute in the json file "data" inside the "data.tzst" file 15:51:20 download https://pkg.freebsd.org/FreeBSD:15:amd64/latest/data.tzst if you want to confirm the problem 15:51:42 extract the "data" file from the tarball 15:51:46 I'm more interested in debuginfo and segfaults ;-) 15:52:00 and run it through jq 15:52:25 something quick and dirty like: cat data | jq -C '.packages[] | select(has("path") | not)' | less 15:52:30 (yes it's cat abuse) 15:53:10 crest: "it's cat abuse" glad it's not about animal cruelty. 15:54:06 run this command if you want to see both packages: 15:54:07 < data jq -C '.packages[] | select(.name == "CuraEngine-debuginfo")' 15:56:10 one package has the path and repopath field and the other lacks exactly those two 15:56:18 "path": "All/Hashed/CuraEngine-debuginfo-5.7.0_5,1~ba3686b059.pkg", 15:56:25 "repopath": "All/Hashed/CuraEngine-debuginfo-5.7.0_5,1~ba3686b059.pkg", 16:23:57 the fixed repo has been replicated to all mirrors 16:24:08 https://people.freebsd.org/~dbaio/pkg-master-report.html 16:24:19 freebsd:15:x86:64 latest 2025-12-17 15:18:41Z 1.02 100 present 17:30:56 How can I check whether the patches were applied on my system? 17:35:08 check /var/log/messages 17:35:28 freebsd-version -ru and compare the output to "Fixed in ...-p" in the announcement email. 17:53:11 hmm I wonder if I could have a clean pkgbase install in a separate BE and then switch to that 17:54:13 V_PauAmma_V: this approach will not work for vmm/zfs patches with freebsd-update, and might scare the user into thinking they have not updated when they have 18:02:07 JetpackJackson: freebsd-version -u -k should be -p1 18:03:54 polarian: kernel version will not be -p1 if not pkgbase 18:12:28 Remilia, duly noted. 18:42:32 polarian: yippee it is 18:48:42 one downside of pkgbase in jails: you cannot update some packages without allow.chflags 18:48:59 though I guess you can pkg -r .... from the host system 18:54:05 tbf that problem exists without pkgbase, too 18:54:28 libc and other critical things are noschg no matter the install mechanism 19:09:02 https://www.instagram.com/p/DSYC-jMDglu/?utm_source=ig_web_copy_link&igsh=NTc4MTIwNjQ2YQ== <-- kookies & bred. 19:11:28 hello, later on with 15.0 -> 16.0 will freebsd-version still exists or? also will branch releng/15.0 get updated patches etc? 19:15:22 ? 19:15:36 who knows 19:15:49 mzar: forget what I wrote :D 20:29:52 slimmed down my pkgbasified system by vicious, prejudiced use of bectl jail 21:19:19 all right, writing this down for the future: `BSDINSTALL_CHROOT=/jail/path/here bsdinstall pkgbase --jail` 21:19:44 unticking every set gives you a 121 MB jail 21:21:32 is that how you can whip out a jail? that's amazing 21:21:58 if you are wise. 21:22:12 you can do a 40MB jail with OccamBSD 21:22:55 it's not even particularly dark magic 21:23:21 i don't like openbsd as much 21:23:23 but it's a whole build 21:23:50 nono https://github.com/michaeldexter/occambsd 21:24:23 it is a script to turn off every build option that can be turned off, and then turn on what you need 21:24:40 thankfully I do not need this 21:25:09 Remilia, Thanks for that! I am writing it down to try later... 21:25:30 you are wise, Remilia :) 21:25:46 rwp: I wish I read that bsdinstall man page two days ago but nevertheless, pkg -r also works hahaha 21:25:56 i would love to convert some of these bhyve freebsd vms into jails 21:26:21 rwp: if you copy the repo keys over to the jail directory, you can `pkg -r /path/to/jail install FreeBSD-set-minimal-jail` and that's basically the same 21:26:53 you don't even need anything else, that pkg command will do everything needed 21:27:11 i reeber back in the day when APC was APC and not APAC schneider asking them for a battery they told me my UPS is not replaceable, then I managed to find the exact same battery the apc website said i needed on officedepot.com 21:27:17 but bsdinstall is definitely more convenient 21:27:51 cpet: they use generic lead batteries iirc 21:28:13 at least for BackUPS series 21:28:14 cant find 12V 6Ah only 5 at lowes 21:28:17 i still have not diagnosed my eaton :'( 21:28:29 i got new batteries and they did not help 21:28:34 was surprised that office depot had the exact pn 21:29:07 the UPS is old and it was connected to my cable modem, switch and voip 21:29:15 so I was upset when it ble up at 2am heh 21:29:28 :( isn't that the way 21:29:43 but APC support is beyond shit 21:30:10 but then again most companies which get bought end up 21:30:15 well shit :D 21:30:20 Remilia, Thanks! I am still on 14.3R everywhere but as soon as I dig myself out of my current task queue debt I will be testing out the new stuff. There is some cool stuff in 15 that I really want to try out. 21:30:55 desktop hasnt crashed in over a day 21:31:06 maybe ill be the one submiting a PR for updated gpu-firmwares 21:31:41 rwp: in other things I learnt today: pkgbasify WILL bloat your system (it installs every single set), and for clean and safe trimming you can do the following (I will probably write a weblog post tomorrow): 21:31:48 cpet, APC makes both user replaceable battery UPS units and supposedly user NOT replaceable UPS units. I think the disposable ones should be outlawed as a crime against the environment. But anyway, as you find it is possible to replace those batteries too. 21:32:28 opened it up and d/c the cables 21:32:34 i would call that user replaceable 21:32:44 if they were dolered on i would classify that as user no replaceable 21:32:45 heh 21:33:13 after upgrading to 15 create a new BE, `bectl jail -U` it and leave the BE; then `pkg -r /tmp/be_whateveritgaveyou set -v0 FreeBSD-set-[every set you do not need]` followed by `pkg -r .... autoremove` 21:33:16 but I bought the battery at 58 and I should receive it on 29th of Dec 21:33:31 I think all of the batteries have spade lugs which are just slide off and slide on connectors. But you had to use a screwdriver to disassemble it, right? That made it disposable rather than user replaceable. 21:33:32 you will probably have to manually `pkg -r ... remove FreeBSD-tests` etc. 21:33:59 Remilia, debloating the result of pkgbasify requires a jail? 21:34:01 set-base includes everything 21:34:11 rwp: well that is stupid now is it 21:34:13 ant-x: I prefer to do it in a separate BE 21:34:18 to be safe 21:34:26 OK. 21:34:27 bectl lets you jail a BE 21:34:29 FreeBSD ist hardcore 21:34:36 Lol 21:34:56 ximon: didnt you say yesterday that you would go back to arch cause you couldnt find netoworkmanager ? 21:35:03 ximon: now you say freebsd is hardcore ? 21:35:41 ant-x: basically I am very paranoid and it is a production system with a ton of stuff on it 21:35:54 so I switched to the new BE and kept the old one for now 21:36:58 for some reason ive noticed that people jut like to do thing just to do things rather to do things when they need to do things rather than just using the computer 21:38:28 Cpet yes 21:38:49 FreeBSD is simple and self-evident 21:38:59 I get stumped every time I have to deal with Linux 21:39:18 Remilia try nixos 21:39:19 Linux was not made for dumb people like me 21:39:20 Boot Environments rock! I can't believe I lived without them for all of the while before. 21:39:41 ximon: I will not, ever 21:39:44 Remilia how comes I'm dumb I can use Linux bot not 21:39:51 But not freebsd 21:39:54 Lol 21:40:22 ximon, Illusion of control, because of an initially low-effort install? 21:40:35 Illusia 21:40:36 * Remilia has FreeBSD for amd64 platforms, IRIX for her Tezro, AIX for the good old RS/6000, and Windows for the home PC 21:40:48 Illusion of controll was always there never was gone 21:41:10 Nobody verify the code your running 21:41:11 ximon: different type of dumb 21:41:16 I guess Solaris is there on the Netras 21:41:24 Remilia, I always have spare systems that I test things on before doing anything I haven't done before on production systems. When possible. I have some systems that I inherited to maintain and they are one-off unique systems without a development area and I always dread working on those systems live. 21:41:53 if you have a production system ill be dumb enough to try tech preview stuff on it 21:41:58 that is on you though 21:42:05 rwp: I know that feeling, it overwhelms me every time BoT/MUFJI ask me to knock sense into their COBOL code 21:42:35 because they spent 30 years migrating off System/360 and are not done yet 21:42:57 I find that configuring FreeBSD is really nice, since so much can be achieved simply by editing /etc/rc.conf, instead of having to go to a million different config files all of different formats on Linux... Things get a bit more difficult when you have to debug things though, since there's less documentation and a lot of things can error out in quite mysterious ways, like with NFS 21:43:11 Remilia, It's a journey not a destination! :-) 21:43:13 MUFJ, the I is extra there 21:43:17 DaliborFox: i dont like json 21:43:23 and UCL 21:43:43 JSON is better than XML, anyway. 21:43:43 rwp: they really wish it wasn't because as it turns out nothing modern can replace COBOL and Erlang 21:44:13 it cost too much to replace it 21:44:13 I think they're emulating System/360 now and that's their best solution 21:44:24 then they realize they have to replace it cause no one knows cobol any more 21:44:30 Putting COBOL and Erlang in the same era in the same sentence blows my mind! I mean, decades between them, right? 21:44:34 ant-x: XML is has schemata, JSON does not 21:44:34 except for retired old school IT peoples 21:44:38 at least use TOML 21:44:47 -is* 21:45:15 i still like the bleh = yes 21:45:23 rwp: Erlang is 1986 21:45:26 Remilia, true, that. But JSON is so much more readable, editable, a parseable. 21:45:27 same age as me 21:45:28 think there were talks in making rc.conf either UCL or jason 21:45:29 * DaliborFox spices things up with S-expressions! 21:45:43