eyalroz | gnarface: I followed up our exchange yesterday with a forum post: http://dev1galaxy.org/viewtopic.php?id=4558 | 00:02 |
---|---|---|
adhoc | morning all | 01:19 |
hyrcanus | can one machine on the ent function as an apt-get proxy (mirror?) for another machine on the local net? | 02:03 |
rrq | hyrcanus: yes. add 'Acquire::http::proxy "http://$IP:$PORT";' to your /etc/apt/apt.conf | 02:07 |
hyrcanus | thanks awesome | 02:08 |
rrq | the mirror needs to "be" a mirror of course.. eg via apt-cacher-ng | 02:09 |
hyrcanus | right, and it doesn't need to grab a whole repo, i assume? | 02:10 |
hyrcanus | cacher implies that | 02:11 |
rrq | right. and that mirror host would proxy for itself via localhost:3142 (default) | 02:12 |
hyrcanus | what should i look for, to cache for a different os variant than the one running on the proxy host? | 02:19 |
hyrcanus | if you can think of any terms to narrow my search rrq | 02:20 |
rrq | the one apt-cacher-ng can serve as cacher for "foreign" arch as well | 02:23 |
rrq | it's a repo cacher for apt, and it caches for all sources | 02:24 |
rrq | there are other cacher, but that's the one I'm more familiar with | 02:25 |
mason | apt-cacher isn't as slick but might be a bit more reliable. apt-cacher-ng can sometimes drop into a slightly corrupt state where you have to manually remove cached files | 02:25 |
mason | That said, apt-cache-ng is convenient enough that the spectre of occasional manual purging is worthwhile. (It's what I run locally.) | 02:25 |
mason | FWIW, in /etc/apt/apt.conf.d/02proxy: Acquire::http::Proxy "http://my.internal.cache:3142"; | 02:27 |
rwp | I also use apt-cacher-ng too and every so often see it fail sticky. | 02:34 |
rwp | I have tried debugging it but have just given up. I still use it. | 02:34 |
rwp | I find the easiest thing is "apt-get purge apt-cacher-ng; apt-get install apt-cacher-ng" to fix it. | 02:34 |
sadsnork | What are the typical symptoms one would see when apt-cacher-ng is having these problems? | 02:39 |
mason | rwp: Ah, so far I've not had to go that far. I just delete the bad repo directory content. | 02:39 |
mason | sadsnork: So, in this case, it would refuse to see a package as being the right length/checksum, and thus fail to install even though the source package is correct. | 02:39 |
mason | And no amount of telling it to purge seems to help. | 02:40 |
mason | This has only happened to me once, but it's known to happen occasionally. From what I can tell it won't ever happen silently, which is why I'm willing to put up with it. | 02:41 |
sadsnork | So on the "client" machine I would see errors when apt-get updating or upgrading? | 02:41 |
rrq | hmm I've never had that as a persitent problem; I thought it happens when a source transiently is "bad", and then I have had to wait .. many minutes .. and then it works again | 02:44 |
mason | Yes. | 02:44 |
mason | rrq: This one lasted for weeks. I was waiting to see if it'd clear up and it never did. | 02:44 |
sadsnork | Well thanks for the tips, I was thinking of using it and am glad I know what to watch for. :-) | 02:44 |
onefang | A transient problem might occur if it hits a package mirror during it's syncing window. | 02:50 |
mason | onefang: Yeah, and it suggested this, but it happened with a small, local repository. | 02:54 |
onefang | I'm talking about the cache hitting an official mirror during the official mirrors sync window. | 02:57 |
mason | Right. Just saying, it suggested that this might have been the cause, but in this case it wasn't. | 02:57 |
mason | Wish I'd saved the error now. Sigh. | 02:58 |
onefang | Lasting for weeks isn't that problem, unless you get unlucky and hit a mirror that's having a lengthy problem. | 02:59 |
onefang | Any given official package mirror being that out of date for that long is very very rare. | 03:00 |
mason | Right. But this was my own repository in the house, so I knew it wasn't that. :P | 03:01 |
mason | (I just cache everything by default, and didn't make an exception for my local apt repo.) | 03:02 |
Afdal | Seriously why do my speakers audibly pop nearly every time I open a new pulseaudio server for a given program | 08:20 |
Afdal | I don't remember this being an issue on *buntu | 08:20 |
CAPTCHA_REQUIRED | thankyou for the ack gnarface | 08:33 |
CAPTCHA_REQUIRED | Afdal, i think it's due to a lack of caps on your sound card | 08:34 |
CAPTCHA_REQUIRED | higher end soundcards don't seem to have that problem | 08:34 |
CAPTCHA_REQUIRED | uh ok, liberachat broke again | 08:34 |
CAPTCHA_REQUIRED | anyways, i get the same problem with alsa when opening or closing the dsp from idle | 08:35 |
CAPTCHA_REQUIRED | but not on my workstation | 08:35 |
CAPTCHA_REQUIRED | which has a slightly higher end one | 08:35 |
CAPTCHA_REQUIRED | there's a workaround you can enable in mpv when using playlists that keeps the dsp open at a constant sample rate and resamples each track to fit | 08:36 |
lts | You may have better luck with pipewire or just plain alsa, if you don't need a sound server. Pulseaudio is still buggy, and probably will remain so until replaced by pipewire | 08:38 |
CAPTCHA_REQUIRED | if you do need a soundserver use jack2 instead | 08:39 |
CAPTCHA_REQUIRED | lts, what is pipewire? | 08:39 |
lts | Ah, Afdal went with the netsplit. | 08:39 |
lts | https://pipewire.org/ | 08:39 |
CAPTCHA_REQUIRED | do you use it? | 08:40 |
lts | I do, not on devuans though since I don't run it on desktops | 08:40 |
lts | meaning: I don't run devuan on desktops | 08:41 |
CAPTCHA_REQUIRED | i thought it was still alpha/betaware | 08:41 |
lts | It's a quite mature beta | 08:41 |
CAPTCHA_REQUIRED | and just an abstraction layer for translating pulse stuff into alsa | 08:41 |
CAPTCHA_REQUIRED | or jack | 08:41 |
lts | A lot less buggy than pulseaudio :-) YMMV | 08:42 |
CAPTCHA_REQUIRED | it's a freedesktop project so it can't be that good | 08:43 |
CAPTCHA_REQUIRED | >flatpak | 08:43 |
CAPTCHA_REQUIRED | >wayland | 08:43 |
CAPTCHA_REQUIRED | call me skeptical | 08:43 |
lts | Beats pulseaudio | 08:45 |
lts | :-) | 08:45 |
CAPTCHA_REQUIRED | I guess? I've still never been stuck in the position where I actually needed pulseaudio | 08:46 |
CAPTCHA_REQUIRED | except for like | 08:46 |
CAPTCHA_REQUIRED | a proprietary videogame | 08:46 |
CAPTCHA_REQUIRED | which i returned because audio wouldn't work without pulse | 08:46 |
humpelstilzchen[ | CAPTCHA_REQUIRED: you could have used https://github.com/i-rinat/apulse | 08:50 |
CAPTCHA_REQUIRED | i've tried that before but it didn't work | 08:51 |
CAPTCHA_REQUIRED | probably using the wrong sound card? | 08:51 |
CAPTCHA_REQUIRED | idk | 08:51 |
CAPTCHA_REQUIRED | thanks though | 08:52 |
Afdal | I actually don't know many comfy OFTC channels | 08:59 |
Afdal | I hang around #openttd sometimes | 08:59 |
hyrcanus | looks like one of the best developed foss games | 09:00 |
CAPTCHA_REQUIRED | there was a resource pack for that game microsoft censored | 09:01 |
Afdal | Yeah it's been around for a while | 09:01 |
CAPTCHA_REQUIRED | it's on archive.org though | 09:01 |
Afdal | It's based on an existing game though | 09:01 |
Afdal | I'd say Wesnoth, SuperTuxKart, Mindustry are way more impressive overall | 09:01 |
CAPTCHA_REQUIRED | xonotic | 09:01 |
CAPTCHA_REQUIRED | don't forget about xonotic | 09:01 |
Afdal | I wish I liked Quake deathmatch enough to enjoy Xonotic :'( | 09:02 |
hyrcanus | whats mindustry | 09:03 |
CAPTCHA_REQUIRED | seconded ^ | 09:03 |
Afdal | It's like... | 09:03 |
Afdal | Factorio if the Factorio developers actually followed through on their developer roadmap ambitions | 09:03 |
hyrcanus | that sounds pretty cool | 09:03 |
Afdal | Imagine an automation game where you make all your units | 09:04 |
Afdal | a cross between automation and RTS | 09:04 |
Afdal | Also it's arguably the most successful commercial libre game of all time now | 09:04 |
Afdal | proving that you can actually success with libre code and commercial sales | 09:05 |
CAPTCHA_REQUIRED | idk what either of those are | 09:05 |
Afdal | either of what | 09:05 |
CAPTCHA_REQUIRED | mindustry or factoryio | 09:05 |
Afdal | Mindustry is basically a game about building impressive factories to produce units to attack your enemies and base defenses | 09:06 |
hyrcanus | with an android and ios client too? | 09:06 |
Afdal | It's coded in Java so I think it can run on anything | 09:07 |
Afdal | but as to whether you would actually want to play it without a keyboard and mouse... | 09:07 |
hyrcanus | i prefer portables with keyboards | 09:09 |
Afdal | OpenTTD is another game that's been ported to phones | 09:09 |
Afdal | which is uh... LOL | 09:09 |
hyrcanus | how much cpu does mindustry typically use | 09:10 |
hyrcanus | what's low-end hardware for it | 09:10 |
CAPTCHA_REQUIRED | it's java, so the question should probably be how much ram | 09:10 |
hyrcanus | yeah that'd be no fun on a small screen | 09:11 |
Afdal | It seems to have pretty darn low requirements to me | 09:11 |
Afdal | it's a pretty simple game at a process level | 09:11 |
hyrcanus | the worst thing cpu wise is the www | 09:12 |
hyrcanus | you need the equivalent of a cray 1 to display some text and images nowadays | 09:12 |
CAPTCHA_REQUIRED | on a tablet maybe | 09:12 |
hyrcanus | and that's a crime | 09:12 |
CAPTCHA_REQUIRED | hyrcanus, https://www.nuegia.net/articles/open%20letter%20to%20webmasters.xhtml | 09:13 |
hyrcanus | yes ^ | 09:14 |
hyrcanus | didn't know about anybrowser.org | 09:15 |
hyrcanus | i'm joining up | 09:15 |
CAPTCHA_REQUIRED | thanks | 09:20 |
hyrcanus | CAPTCHA_REQUIRED: this glyph wasn't in the unicode block drawing set ▬ but incredibly useful | 09:33 |
hyrcanus | ▬▬█▬▬ swiss flag | 09:33 |
CAPTCHA_REQUIRED | what is that relavant to? | 09:35 |
Afdal | what is this torus@ascii.town thing :3 | 09:38 |
hyrcanus | ascii.town had this glyph in their helpscreen and i've been wanting it for years | 09:39 |
hyrcanus | it's not in the unicode block graphics section | 09:39 |
hyrcanus | only alternative i knew was ━ which is too narrow | 09:39 |
Afdal | I don't see it | 09:40 |
Afdal | All I see is BIG CHUNGUS | 09:40 |
CAPTCHA_REQUIRED | oh! | 09:40 |
CAPTCHA_REQUIRED | Afdal, ssh in and find out | 09:40 |
Afdal | I am :v | 09:40 |
Afdal | Where can I learn about more of these cool ssh servers | 09:43 |
Afdal | I don't even understand how this works | 09:43 |
Afdal | is this some kind of bash/dash/fish script | 09:43 |
hyrcanus | they link to the source | 09:44 |
CAPTCHA_REQUIRED | Afdal, no | 09:45 |
CAPTCHA_REQUIRED | it's just an interactive executable in place of a shell | 09:45 |
Afdal | a custom shell | 09:45 |
CAPTCHA_REQUIRED | in /etc/passwd | 09:45 |
Afdal | is the program itself a shell? | 09:45 |
CAPTCHA_REQUIRED | yes | 09:45 |
hyrcanus | you can set a user's executable to something besides /bin/bash | 09:48 |
hyrcanus | a roguelike is fun to dump them in after password entered | 09:48 |
hyrcanus | if i offered a free shell service, you'd get into your shell only after defeating a monster | 09:48 |
CAPTCHA_REQUIRED | all /sbin/login does is ass for a username/password and then set the process uid | 09:48 |
CAPTCHA_REQUIRED | then exec whatever /etc/passwd says to exec | 09:49 |
CAPTCHA_REQUIRED | hyrcanus, oh nice | 09:49 |
CAPTCHA_REQUIRED | interesting way to prevent abuse | 09:49 |
CAPTCHA_REQUIRED | do you have a screenshot of the monster? | 09:49 |
hyrcanus | i was thinking a simplified brogue | 09:50 |
hyrcanus | or a game of 'hunt' | 09:50 |
Afdal | how about make them get a netris high score | 09:52 |
hyrcanus | there we go | 09:53 |
Afdal | https://gitlab.com/tslocum/netris | 09:55 |
hyrcanus | ascii-invaders is rather fun. | 09:56 |
Afdal | is there ascii-galaga | 09:56 |
Afdal | oh lawd there is | 09:57 |
Afdal | GalagASCII | 09:57 |
hyrcanus | oo where's that | 09:57 |
Afdal | https://github.com/mpoegel/GalagASCII | 09:57 |
hyrcanus | very nice, no ncurses | 09:57 |
hyrcanus | how does 4152 bytes of Controller.cpp turn into 459kB of Controller.o | 10:01 |
Afdal | hmm is Devuan setup to run any SSD TRIM services automatically at installation? | 10:27 |
Afdal | I notice a popular TRIM tool runs on systemd | 10:28 |
hyrcanus | what does trim do for me | 10:28 |
hyrcanus | can you see two thick and two thin horizontal bars here? ▬ ━ ▬ ━ | 10:29 |
Afdal | yeah | 10:29 |
hyrcanus | ty | 10:29 |
Afdal | TRIM moves stuff around on your SSD to level out the amount of wear in different sectors to sustain long-term performance | 10:30 |
Afdal | I think *buntu is normally setup with some periodic trim command running automatically | 10:30 |
hyrcanus | i thought ssd firmware should handle that transparently | 10:31 |
sixwheeledbeast | ssd has wear levelling, trim works additionally to that. | 10:49 |
sixwheeledbeast | There is no moving it marks areas for cleanup. It's a systemd timer on ubuntu now but it was a cronjob in the past. | 10:58 |
DiffieHellman | Yes, actually, TRIM just lets the SSD know which sectors are unused (due to file deletions etc) which lets it use those sectors for wear levelling. | 10:59 |
DiffieHellman | s/unused/no longer used/ | 11:01 |
sixwheeledbeast | should simply be something to run fstrim from PATH | 11:01 |
hyrcanus | my ext4 ssd mount says the discard operation is not supported | 11:39 |
hyrcanus | it's on usb3 if that makes a diff | 11:40 |
ukine | there is also such thing as "hardware level" (chip on the SSD) "garbage collection" | 11:46 |
ukine | that i used to think may have been a useful feature... | 11:47 |
ukine | oh well :/ | 11:47 |
hyrcanus | well i do understand now that the ssd firmware needs to be told what's really free sectors | 11:56 |
ukine | 'hardware garbage collection' | 11:57 |
ukine | i don't know how it may or may not readily identify what fs you're using... | 11:57 |
ukine | that would apply obv | 11:58 |
ukine | possibly via trim | 11:58 |
ukine | that feature could have been an extension of trim | 11:58 |
ukine | *or is | 11:59 |
ukine | i'm just a little curious about it at the moment :| | 11:59 |
hyrcanus | hrmph well it seems ext4 isnt the problem | 13:02 |
gnarface | hyrcanus: no, it must have been an imposter. i've never followed any links from you in private before. please don't spam me again | 15:42 |
jla | for simple / local LXC use in devuan... is there something special / particular i need to know ? | 16:00 |
* jla reading ... https://wiki.debian.org/LXC | 16:02 | |
lts | Devuan is my go-to LXC base. Can't recall issues, and I'm just building a new environment so I think I would remember if encountering any in the last week or so. | 16:07 |
lts | No experience with devuan host, though | 16:07 |
jla | i'm talking more from the devuan host perspective.. | 16:08 |
jla | i mean, i see systemctl lxc services related... commands, here's why i have my doubts | 16:11 |
gnarface | jla: "service [whatever] stop/start" should still work. they inherited that wrapper's behavior from sysvinit and you can still install it here. the thing you need to worry primarily about being different is device permissions; you won't even have permission to play audio unless you add yourself to the "audio" group - systemd bypasses all device permission safeties for the local console users | 16:24 |
gnarface | well not just device permissions, filesystem permissions too but you won't have to mess with them unless you're changing something | 16:24 |
gnarface | you might still have to add yourself to the audio and video group manually | 16:24 |
gnarface | don't worry if you run into trouble with most simple stuff someone here can usually help you | 16:25 |
gnarface | there is tons of built-in documentation though | 16:25 |
e3d3 | Is the name Devuan choosen because of the meaning in Japanese, `Dev plan` according to Google translate ? | 16:56 |
brocashelm | the "vua" = veteran unix admins | 16:58 |
e3d3 | really ? so the Japanese meaning is a coincidence ? | 16:59 |
gnarface | e3d3: the story i heard was that it's phonetically spelled so it will sound like "dev one" to the internal monologues of the most different international languages | 16:59 |
brocashelm | https://www.devuan.org/os/team | 16:59 |
brocashelm | "devuan" honestly sounds more french or spanish to me | 17:00 |
gnarface | well the complaint is that actually spelling it "devone" would have only done that for english speakers at the cost of the french and spanish particularly | 17:01 |
e3d3 | aha, the international thing explains why it also fits in Japanese | 17:01 |
gnarface | and the italians i think? | 17:01 |
brocashelm | i think it's a great name/spelling | 17:01 |
gnarface | yea they really did try to make the name multi-purpose and something that would be familiar sounding in any locale | 17:01 |
brocashelm | although i also like/use refracta :) | 17:02 |
e3d3 | I think I still can pronounce it while I'm drunk. Maybe not so good | 17:02 |
e3d3 | anyway, thanks for the good explanation, and of course for wonderfull Devuan | 17:04 |
gnarface | eyalroz: did you ever get an answer to that? i'd have to install this to figure it out but it should have documentation really that tells you already and i was fairly sure it's a directory you can see and guess pretty easily in the package list that comes back when you run "dpkg -L iptables-legacy-save" | 17:08 |
gnarface | (or maybe one of its other related packages) | 17:09 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!