01:37:48 pathbuf canonicalize() handles "~" properly FUCK YES 02:04:27 heh 04:08:45 hello 04:09:09 why would a bridge interface in freebsd have lots and lots of Oerrs in netstat -in 04:09:24 it's also limiting packets to 10 megs in one direction 04:09:29 the underlying interfaces aren't erroring 04:31:46 that's so strange. setting sysctl net.inet6.ip6.forwarding=1 brought it up from 10 megs to 225 megs 04:32:02 but it's still slower then it should be. it should be 900+ 04:36:46 what's attached to the bridge? 08:11:39 rtprio, epairs for fnet jails and lagg0.4 08:20:43 lagg0.4 is a vlan of lagg0, and lagg0 is a failover of two intel igb adaptors 08:25:31 i've tried disabling lro, tso, hwvlantagging, hwvlanfilter, and others 08:25:43 none of those seem to have much of an effect 10:00:45 nom vs pest for making a text format parser? i wanna try making a json parser 11:00:00 or peg-rs 11:08:43 I am legally obliged to recommend nom 11:09:02 why 11:23:47 polyex: geal's my mate :p 11:32:21 i think i like peg parser more than combinator tho. to just have a simple grammar to work with is really nice 11:32:30 what you think Hecate? 11:42:58 polyex: I am a Haskeller so I'm heavily biased towards combinators, although in the end I don't know your objective: good error messages or speed? 11:43:23 good errors 11:43:37 the more custom the better then 11:44:36 the peg crate looks very declarative 11:46:01 but then you would perhaps lose control over what errors are reported and how 11:46:37 i thought i was in #rust lol 11:47:08 :) 11:47:27 I just installed btop on my freebsd boxes, lovely interface 11:48:16 btop? 11:54:49 polyex: an alternative to htop! 12:39:45 Hi there .... 12:40:13 After upgrading i'm missing mysql support in python ... 12:40:47 previously I used py-mysql-connector-python port, any suggestion for mysql connector for freebsd for python 311 ? 12:55:27 nevermind 12:55:44 recompiled ports it starts working thanks 13:19:40 Hello, all. I prevent my HDD from parking and unparking like crazy with camcontrol(8), by invoking it at boot, from /etc/rc.local . But after suspend and resume cycle the effect seems to disappear, whence two questions: 1. How can I check the current state of APM management and the value of the standby timer (to make sure they are reset after suspennd), and 2) shall I therefore invoke camcontrol /also/ from a resume script (as per acpiconf(8))? 13:23:02 ant-x I guess it is better to adjust drive options via smartctl 13:23:27 some drives does require hardware settings 13:23:29 afk 13:24:15 so you got a bridge and a tap, if I pass on tap... then traffic will be allowed to and from tap, but it goes tap --> bridge --> phy, so if I do all port filtering on the bridge, even though all packets are passed on tap they will be then caught in the bridge rules, correct? 13:28:54 nerozero, why smartctl? Does it make more persistent settings? 13:29:48 My most pressing problem now is persisting drive settings over suspend-wakeup. 14:09:42 its permanent 14:09:53 you write this setting into hdd eep 14:12:49 nerozero, I meant 1) camcontrol idle -t 0 ada0 and 2) camcontrol apm -l 254 . They seem to be external controls, not recorded into the HDD. 14:14:47 camcontrol not always communicates properly with drives 14:16:12 also look et EPC settings 14:16:22 Extended power conditions if i recall correctly 14:19:25 nerozero, I have checked that my drive does not support EPC. 14:19:34 :( 14:19:47 is it related to zfs ? 14:20:08 drives fall offline sometimes ? 14:22:04 nerozero, no, no ZFS here. I said: camcontrol works correctly until suspent-resume. Then I have repeat the two camcontrol commands above to restore the desired HDD behavior. Are you 100% sure those settings are stored insde the HDD. If they are not, the situation is explained. 14:23:58 The new so called " green drives " do spindown by automatically 14:24:18 that is a consumer grade drives, chipo onces 14:24:34 its controlled by EPC 14:24:56 as far as I know freebsd does not automatically spindown drives 14:25:22 also suspend on bsd is very poorly implemented 14:26:32 My drive is: WD5000LPCX . All I know, is that [camcontrol apm] and [camcontrol idle -l ] commands do persist their effect across suspends. Should they? 14:40:14 This page: mentions several official Russian-language FreeBSD lists: ru-freebsd-annouce, ru-freebsd-chat, ru-freebsd-doc, ru-freebsd-questions , but I see no such lists now. Does anybody rember what has become of them, and why? 18:00:40 hi all 18:00:59 how can I set default system compiler to gcc in freebsd 14.1 ? I looked at src.conf and make.conf man pages but anything regarding gcc settings were removed from man pages 18:03:53 Set the cc environment variable to point to the full path to gcc as installed by ports, I believe. 18:04:27 I am curious what you're hoping to achieve with it, though. 18:05:13 while possible, results may vary with gcc since i'm not aware of anyone actively testing it 18:05:47 playing and learning what can i do and cant (learning freebsd by breaking, etc learning from man pages and handbook) 18:06:33 how can i compile the kernel and world with gcc tho ? there seems to be no definition left in man pages of src.conf and make.conf for gcc 18:14:01 ok, but why? 18:19:59 rtprio, whyy not :) 18:57:43 CountryBall0, what -- the default compiler /not/ GCC in 14.1? 18:59:22 ant-x, I'm trying to change default compiler of 14.1 to compile kernel and world and ports with gcc14 19:02:47 CountryBall0, I thought your question implied that GCC was /not/ the default compiler on FreeBSD 14.1, and found that /very/ unusual. GCC very slow, bloated, complicated, and requires special options sucn as --pedantic to accept the actual C, but still it is the de facto starndard. 19:04:33 No, gcc is not the default compiler, and hasn't been for some time 19:08:24 Oh, just as well. 19:09:10 Can devd in daemon mode be caused to log events to a file? 19:16:34 ant-x, thx . i just wanted to see if I'm able to change compiler for the system.. gcc was the known and most up-to-date option to try 19:38:54 ant-x: freebsd has used llvm as a toolchain for most things for a while now, and is currently in the process of (or maybe has already) switched some of the elf tools to it too 19:39:24 for a while there were only a few of us using gcc while clang still had issues with debug info, but they're largely the same now... so there's not much point in using gcc 19:44:31 I have a newbie shell question: I am trying to log devd events into a file. First I found that devd -d prints to stderr, and was able to filter it with: devd -d 2>&1 | grep 'Processing event' . This printed event list to the console. Then I tried to redirect the result to a file: devd -d 2>&1 | grep 'Processing event' >devd.log , which does not work: the events are handled, but devd.log is empty. Why? 19:54:17 buffering most likely 19:55:51 llua, You mean when I press Ctrl+C to exit from the devd running in foreground, the descriptors are not flushed? 19:57:23 /msg greybot !faq buffering 19:58:51 llua, Thank you. Will read: . 21:39:38 When upgrading 13->14, is there a new version of the bootloader I should also install? 21:39:50 I remember that was very important for 12->13 22:13:26 llua, re: buffering and that FAQ question: it says the buffering is performed by the writer, and that `grep` avoid it for console output, but buffers if its output it piped futher. How does it distinguish between these situations internally to decide wheter to buffer or not? I thought it simply wrote to stdout using the standard functions... 22:18:16 nerozero (not here, but may be other can help), I cannot find any evidence that the effects ( and values) of `camcontrol apm -l ` and `camcontrol idle -t ` persist in the HDD epp, as you say. Could you please point me to the relevant documentation? 22:53:37 man 3 isatty 22:53:43 ant-x: ^ 22:56:38 Thanks. 23:07:15 What do you guys think is the better or more idiomatic way to persist settings across power-offs /and/ suspends -- calling a `persist' script 1) at boot (/etc/rc.local) and 2) from a generic rc.d script (as per rc.resume) or from the devd event?