00:08:11 destination path needs to end in . lol. so cp -RpP /tmp/foo/ /home/foo/. 00:08:14 so weird 00:08:31 guess it makes sense but the trailing / should convert how symlink is treated into a dir 00:08:47 if we can touch /home/foo/bar.txt, should be able to cp into it 01:44:34 hi 01:44:59 Is there something like `ldd` that also shows dynamically called linked dependencies? - Like `dlopen` style? 02:36:48 is there anyway to see the -px (x being the version) subreleases... 02:38:19 are they announced anywhere I mean* 02:53:10 I think polyex has found a minor bug in the cp handling of -R. It can be workedaround by using "cp -R src/ dst/." instead of "dst/" pushing even further through the symlink to the "." pointed to. 02:53:44 Normally using foo/ forces foo to be a directory and implies "foo/." but that seems to be getting snarled up there with cp -R for some reason. 03:50:29 now you've got me thinking about grep optimization. Sigh. 04:29:31 I've been known to troll people into action in the past, so don't feel bad 05:07:37 taking /dev/random and the pattern [1234567890]+1234567890, FreeBSD is slightly faster than the l4w (also the l4w is 2950x vs freebsd on 1900x). 05:08:16 ... but l4w might be getting in it's own way. I'll have to bench this on some equivalent VMs or something. 05:09:08 1.3G/s (freebsd) vs. 1.1G/s (kali linux on l4w). And the linux had the faster processor. 05:20:09 zBeeble: and you're definitely testing /usr/bin/grep and don't accidentally have gnugrep from ports installed? 05:21:04 I will check. I don't usually istall gnu grep. 05:21:16 Maybe the pattern isn't involved enough to trigger? 05:24:25 I'm certainly not giving up because _one_ test was good. I have several reasons to believe that it is slow. 05:25:25 (I was just idly poking at things while I fixed up the pspp poudriere mess) 05:30:53 to be fair /dev/random also isn't necessarily representative of the kinds of workloads grep typically operates under 05:31:33 usually we're talking reasonably-sized lines that we need to search, /dev/random may have random \n scattered around enough that it's not obvious 05:36:00 nope ... not gnugrep. 05:36:36 true, true. I'll think about testing it with the pspp log ... see how well that goes. 07:42:22 hi.. i got stuck when i did "pkg search nginx" 07:42:24 how to fix? 07:46:04 I have cerated jail with iocage but when I try to use pkg -j it says jail ... not found. any tip ? 07:46:13 or is bettre to use ioncage pkg ? 07:48:28 iocage not ioncage ... :) 07:50:17 Gurar: if you use the same 'JID' as listed in the 'iocage list' for 'pkg -j X ...' ? 07:51:05 yes 07:51:15 Im using name not jid 07:51:18 ups 07:52:23 is possible to use name instead of jid ? 07:54:47 yes. Use 'jls -N' to get name rather than jid 07:55:17 iocage apparently prefixes the jailname with 'ioc-'. 07:56:02 what do you suggests to use iocage pkg or use pkg directly ? 07:57:28 I'd use pkg directly. 'iocage pkg' has (atleast for me) had some issues with the output. 08:03:42 norrland thanks! 08:05:30 al1r4d: what's the output? Can you post the command and the output in https://dpaste.org/ ? 08:05:31 Title: dpaste 09:34:02 after following the handbook on freebsd upgrade minor version i get freebsd-version -kru 14.1-RELEASE 14.1-RELEASE 14.1-RELEASE-p1 how come -kr doesnt show patch level? i have restarted and run last freebsd-update install but still same 09:45:52 Puzfire: patch level can differ for patchsets that only change the userland or libraries, since unames values are derived from sysctls, which are only updated when the kernel binary changes. 09:46:18 freebsd-version -kur will tell you the kernel, userland, and running versions respectively 09:49:08 so all upgrade is done correctly even so : # freebsd-version -kur 09:49:08 14.1-RELEASE 09:49:08 14.1-RELEASE 09:49:08 14.1-RELEASE-p1 09:49:08 uname -a -> 14.1-RELEASE FreeBSD 14.1-RELEASE releng/14.1-n267679-10e31f0946d8 GENERIC amd64 09:50:24 if 14.1p1 is the newest patch set, then yes 09:50:51 ye just did update :) wasnt sure as old upgrade on 14.0 did look different ;) 09:51:03 so was unsure 09:51:31 so now i can venture into the jails, now im sure host is done correct. need to document it so i can remember 10:34:28 Im creating jails with iocage but I can access internet from jails so I cant even install pkg's etc. Im creating jail like this iocage create -n postgresql ip4_addr="192.168.2.20" -r 14.1-RELEASE 10:53:52 inside a jail can we create our own zfs datasets separate from those of the jail's host? 11:43:04 yes 11:54:24 Gurar: have you read https://iocage.readthedocs.io/en/latest/networking.html ? 11:54:25 Title: Networking — iocage 1.2 documentation 11:55:17 norrland: thank you. But looks like my problem was solved 16:07:21 i am super happy about the sys/fs/p9fs commit in CURRENT (and the "MFC after: 3 months" note) :) 17:40:39 https://www.theregister.com/2024/06/20/systemd_2561_data_wipe_fix/ 17:40:40 Title: systemd 256.1: Now slightly less likely to delete /home • The Register 17:40:42 thank you freebsd ;] 17:40:45 i get to keep my homedir 18:03:39 haha yikes 19:35:17 dch do you know how to create zfs datasets inside a jail? 20:26:19 good evning! I thinking of trying to install FreeBSD on a macbook air mid 2012, and wondering is it something special I need to think off? or can I just download the image to an usb and install it? 20:33:23 polyex: yes, there are really just a couple of things you need to do 20:34:11 zfs create -o canmount=off -o mountpoint=none zroot/jailed # make a containing dataset to keep things tidy and clean, not strictly required, but very helpful 20:35:07 zfs create -o mountpoint=/home/jenkins -o jailed=on zroot/jailed/jenkins # create the dataset that will *in future* be jailed. it needs mountpoint to be relative to the jail chroot 20:36:25 zfs jail jenkins zroot/jailed/jenkins # do this just after jail creation & before startup, which mounts the dataset inside the jail 20:37:10 if you use jail.conf then it looks a bit like this: https://git.sr.ht/~dch/ansible-jails/tree/main/item/templates/instance.conf.j2.default#L28 20:37:11 Title: ~dch/ansible-jails (main): templates/instance.conf.j2.default - sourcehut git 20:41:43 dch ah. what i was doing was i had my zfs clone -o jailed=on -o volmode=dev zroot/jails/templates... zroot/jails/jail/testjail. but that didn't work 20:42:59 polyex: ok yeah you can probably "jail" block devices but I don't think that would be secure 20:43:14 why not? jc 20:45:35 <|cos|> nwe: i've never tried an air, but my learnings from attempting on a macbook pro is that apple hardware leaves a bit to be desired. 20:46:58 <|cos|> nwe: i got 14.0 running "okay", but convenient features such as resume after suspend and sound is where i'm currently stuck. 20:48:27 <|cos|> nwe: for network i needed to go with wifibox, which runs an alpine in a bhyve and load some binary-blob linux-only drivers. 20:51:24 |cos|: thanks for the heads up :) Ihmm wifibox, whats that? 20:52:20 <|cos|> nwe: a virtual machine running a minimal linux, just for the linux wifi-drivers. or in abstract terms, a pragmatic solution. 20:53:33 <|cos|> nwe: i used a usb wifi-adapter while installing, just to get to the phase where i could replace it with wifibox. 20:59:13 |cos|: okey, regarding the sound is it because lack of driver support? 21:01:04 <|cos|> sound works with headphones. i read people managing to get speaker sound by fiddling. i wrote down the link, let me find it. 21:02:22 Hi, I'm trying to capture packets using WireShark on FreeBSD 14.0. The issue I'm seeing right now is only incoming packets are captured. This happens with and without firewall enabled. 21:06:01 <|cos|> nwe: https://forums.freebsd.org/threads/no-audio-on-my-system.57473/ 21:06:02 Title: No Audio on my system | The FreeBSD Forums 21:06:37 <|cos|> nwe: the second page has stuff like `hint.hdac.1.cad0.nid18.config="as=1 seq=0"`. let me know if you get it working, especially if you truely understand it. (: 21:08:45 |cos|: I will first create a backup, and when I have time to install it and get it to work I will reach out to you! 21:10:41 <|cos|> nwe: backup seems like a great idea, actually 21:12:21 <|cos|> unfortunately macOS is probably the only thing one can run if wanting to use a macbook as ones daily driver. not having suspend-resume gets old quite quickly... 21:14:49 |cos|: the sad thing is that I cant upgrade that macbook.. 21:17:14 <|cos|> nwe: i understand, and believe it will become a neat freebsd machine. just want you to be fully prepared on what to expect. 21:20:14 nwe: Have you looked at OpenCore Legacy Patcher? https://dortania.github.io/OpenCore-Legacy-Patcher/BUILD.html 21:20:15 Title: Building and installing OpenCore | OpenCore Legacy Patcher 21:21:20 mariuss: nope, didn't know it exists :P 21:26:48 dch why it insecure to jail block device? 21:41:31 Well, this is odd. I'm working with a client's systems where /etc/rc.conf has multiple ifconfig*alias* lines. However, when the booting system hits one line, it just hangs. Any thoughts on what sort of thing triggers this? 21:42:53 paste 3 lines. the line above it, the line that halts, and the next line 21:49:47 ifconfig_genet0_alias0="inet fqdn.domain0.net netmask 255.255.255.255" 21:49:58 ifconfig_genet0_alias1="inet fqdn.domain1.com netmask 255.255.255.255" 21:50:08 ifconfig_genet0_alias2="inet fqdn.domain2.net netmask 255.255.255.255" 21:50:31 shouldn't gqdn.domainx.x be ips not domains? 21:50:31 Actual names substituted.... 21:50:47 All fqdns exist in /etc/hosts 21:51:02 change to ips and try 21:51:44 Did that for fqdn.domain1.com, and got it working. Just trying to figure out why just the one won't work. 22:10:01 in bsdinstall's installerconfig i set some env vars. but bsdinstall is ambiguous on the format. quotes or no? like nonInteractive="YES" but ZFSBOOT_VDEV_TYPE=stripe. wtf? 22:10:18 man bsdinstall is ambiguous* 22:15:37 In many circumstances, quotes are optional. 22:17:40 If your data includes things like spaces or single quotes, wrapping it in double quotes disambiguates it. Otherwise, if your data is a simple alpha-numeric string the quotes do nothing. 22:21:48 ty 22:23:35 Wrapping in double quotes is a best common practice. It allows including spaces in data while also allowing variable expansion. From this, no quotes or switching to single quotes, or adding backslash escapes cover the uncommon exceptions. 22:25:12 just always wrapping env var values in double quotes even when not needed? 22:25:55 will wrapping value in double quotes ALWAYS be valid no matter what or are there more rules? 22:26:49 Yes. It comes from historical development of the shell, back they'd resolved the requirements. 22:27:22 is there such a thing as export FOO=1 is different from export FOO="1"? 22:29:59 Yes. In test comparisons, the former is numerical, while the latter is a string. 22:31:01 so what if i "double quote everything" but then the prog using the env var expects numerical 1 instead of string "1"? 22:31:01 Given shells are loose with respect to types, there exists options for pushing for the opposite during specific tests. 22:31:05 i just passed wrong type 22:32:20 Possibly. That's an uncommon case though. 22:32:41 Shells are loose on types, so that's an easy override. 22:32:42 man that sucks 22:32:49 the ambiguity is really stressful 22:33:13 It can be. Fortunately that's rare. 22:34:09 I think I've run into "1" vs. 1 something like a half-dozen times since I got started on Unix. 22:34:26 * tmp_ started on Unix in 1988. 22:55:18 looking at https://github.com/freebsd/freebsd-src/commits/releng/14.1/ looks like some updates since it came out. have those gone through any other release for vetting or are they new code? 22:55:20 Title: Commits · freebsd/freebsd-src · GitHub