01:15:34 hello everyone 01:15:42 finally got suspend and resume working in freebsd 14.3 01:15:47 :) 02:46:46 hi 03:00:23 ok my sh is not in /bin/ like in /usr/local/bin... i need it to be in /bin/ because some software relies on it 03:00:57 can anyone maybe hint to where i should dig to get proper /bin/sh? 03:03:06 i did a soft link and it worked.. on some cases, and too many levels of indirection error in some other cases (from memory).. i am still investigating what's happening 03:03:19 the softwrare in question is etherpad btw 03:07:26 agent314: what do you mean a "proper" /bin/sh? 03:07:54 i feel like we're missing a lot of context here still 03:23:02 like there is no sh in /bin folder 03:23:10 i dont know if that's normal 03:23:36 actually let me get a wgetpaste 03:26:11 i'm sorry i meant bash 03:31:36 actually i think what i might do is just copy bash to /bin (sorry, wgetpaste doesn't seem to want to connect to anything nor is there really anything heer to wgetpaste) 03:34:27 i dont really see any reason why it would not work in the short run.. i think what's happening here is etherpad expect to be on linux and say in my gentoo there is a bash in /bin/ folder, but for some reason freebsd decided to not put bash in /bin/ ¯\_(ツ)_/¯ 03:38:51 Anyone have experience with this issue? I'm hitting a wall trying to set up a Python Flask-Socket.IO app (backend) behind Apache24 on FreeBSD 14.3-RELEASE. It's on HTTPS, proxied to a subpath like /mylogs/ 04:01:37 agent314: that is not normal sh belongs to the base system, so it shouldn't be in /usr/local (as that is for non-base stuff), but your system should be absolutly broken without /bin/sh 04:05:52 ah, wait your prolem is only with bash, yeah, that doesn't belong to base and sh doesn't have to be bash, software that expect sh to be bash or expects bash to be in /bin/ is just broken and you should reconsider if you want to use that software 08:23:28 * |cos| doesn't use his printer very often, and tends to get confused by lp in base. Is there a way to make FreeBSD's executable usable with contemporary (i.e. networked) printers, or is it just a legacy thing blocking cups' lp from ports to be found (with a sane $PATH)? 08:26:13 * |cos| probably spends more time to turn on+off the printer, shake the toner and load with paper than remembering to call /usr/local/bin/lp.Sso its not a big thing, but this gotcha moment happens every time. Maybe I should just alias it… 08:48:54 who was having problem with the kmods repository here - rwp maybe? apparently there was an issue where it wasn't building the gpu firmware flavours properly, but this should be fixed within the next few days (when mirrors update) 08:52:59 ivy, I was definitely having that problem. Good information. Thank you for passing it along to me! 08:54:26 rwp: i'm not 100% sure this is the cause of your problem but it's worth trying again at least 08:56:35 Understood. For the moment I am not doing anything with it because I have been traveling and travel continues for another few days. I didn't want to mess with it before I left. I'll figure it out when I get back but it's going to be another week yet. 08:57:20 I appreciate you passing along that newsworthy tidbit of information. I am also not paying attention to lots of happenings right now. 12:59:08 if you are going to do a KDE version, do a XFCE version too, if you don't mind. 13:01:37 XFCE is the least breakable DE i know. 13:07:59 stumpwm! 13:11:48 commom man, XFCE is classic! 13:12:19 and exactly this is one of the reasons why bsdinstall should not bother having a DE setup option IMHO 13:13:53 but then again, I haven't used bsdinstall in like forever 13:13:56 its useful to speed things up 13:14:19 if your going to install a DE anyway 13:15:00 im not saying to stop doing the tradicional version. Both are needed. 13:16:11 traditional* 13:17:20 https://stumpwm.github.io/images/screens/boring.jpg 13:17:22 omg 13:17:29 just run a tty already 13:17:54 run a emacs on a tty and you can get more! 13:19:29 talisc: hey, I read one of those books! 13:19:56 nice =D 13:20:02 but the print is not mine 13:24:08 of course there's a lisp book on that picture... 13:24:10 cracauer 13:24:33 jbo: 3 of them 13:26:36 i'd really like to run a pure tty if only there was a reasonable tty web browser 13:26:55 browsh seemed promising but then they dont provide keyboard navigation out of the box 13:27:01 "reasonable" and "web" don't go well together 13:27:36 yes but browsh seems able to crunch it all down to text output 13:28:21 and it has firefox backend so it probably actually works for 99% of web browsing 13:28:44 oh, that actually looks kinda neat 13:28:51 I thought it was just another lynx 13:29:22 seti_: links 13:29:28 links is fine 13:29:30 But it uses one of those "mouse" things? 13:29:42 yes, the dreaded mouse thing seems required 13:29:57 talisc: you can find it easily if you google browsh they have a website 13:30:14 they even provide a ssh:able demo i think 13:30:24 that is "temporarily offline" 13:30:43 there is also Emacs' EWW, but i prefer links. 13:31:28 oh i thought talisc was asking for a link to browsh haha 13:31:33 no 13:31:35 haha 13:31:53 links is a cli browser 13:32:15 I think I might give browsh a try just for the fun of it 13:34:15 gave up after I ran browsh right from their container and things were broken 13:34:37 container? 13:34:59 their Docker image 13:38:02 saw now on their issue page one of the contributors mentioning that the docker build has probably been broken for years 13:42:03 I also saw that they seem to be trying to work in native vim keybindings for browsing 13:42:43 big potential like i said, its only a frontend to firefox so all the gory webcrap gets handled by whatever firefox you have on your system and browsh parses it into text output 13:43:38 bleh. it was awhile ago but I think I only used the Docker image because I ran into problems with their static binary 17:04:32 Interesting. I use caddy as my reverse proxy. I had to make all my gitea repos private due to the constant, excessive scraping. Setting up anubis was something I kept putting off. Now it turns out there is a reimplementation of anubis that is a caddy plugin called Cerberus https://github.com/sjtug/cerberus. Should be easy to xcaddy build it into a 17:04:32 new caddy bin 17:07:15 Anubis is pretty simple to implement. With either solution, you'll definitely be happy you did it. That scraping is SUPER annoying. 17:20:00 The setup instructions for anubis says it requires either (a) a separate instance of anubis running per service (sub-domain in my case, of which there are multiple), or (b) one caddy instance dedicated to TLS and to reverse proxying to anubis which then reverse proxies passed traffic to another caddy instance for all the subdomains. Not ideal for 17:20:00 me 17:22:31 It seems like some clever use of the forward_auth directive might allow it to work in a single caddy instance, but not sure. Plus Cerberus is apparently an option, so i think i will try that out 18:10:50 Let us hope that the chinese company publish the specs so that *nix drivers can be created https://www.tomshardware.com/pc-components/gpus/chinas-first-6nm-gaming-gpu-matches-13-year-old-gtx-660-ti-in-first-geekbench-tests-lisuan-g100-surfaces-with-32-cus-256mb-vram-and-300-mhz-clock-speed 18:18:13 scoobybejesus_tl: I have one instance of Anubis running and it covers all my top-level domains as well as all sub-domains. 18:19:21 I do reverse proxy it, though. 18:20:04 I'm pretty sure Dan's write-up is still available. 18:22:45 Or this. Basically the same thing: https://herrbischoff.com/2025/07/how-to-defend-against-aggressive-web-scrapers-with-anubis-on-freebsd-14/ 18:33:47 Interesting, thank you. Yeah, I've read Dan's write-up, but I hadn't seen the second one, and I found another as well that Dan had linked 18:35:52 Maybe I will spin up a single anubis just to protect gitea and go from there. Or maybe Cerberus. Hm. We'll see. 18:36:46 * hodapp forgets he has a gitea 18:36:48 hm. I should use that 18:55:19 Yep. That's kinda how I started with it. Simple things just getting scraped like crazy. Forums and repos were the worst. 20:09:02 i've been okay for cgit so far, wouldn't want gitea/forgejo on the internet 20:16:12 scoobybejesus_tl, https://dan.langille.org/2025/05/03/implement-anubis-to-give-the-bots-a-harder-time/ 20:21:54 that's the one, thank you :) 20:41:38 does anyone have an idea how to get endurance information out of an NVMe ssd? none of the available nvmecontrol log pages seem to return anything relevant 21:43:07 ivy: ssd manufacturer provides the "TBW" value (terybytes written), then you read SMART data with smartctl to see how much your wrote up to this point 21:43:34 aic: how do you read SMART data from an NVMe device? 21:44:05 smartctl --all /dev/nvme ? 21:45:37 ah, i was using /dev/nda0, that works better 21:49:23 cool, i bluffed it with linux name for nvme devices