[01:51] *** killsushi has joined #archiveteam-bs [03:11] *** RichardG has quit IRC (Read error: Operation timed out) [03:16] *** qw3rty111 has joined #archiveteam-bs [03:22] *** qw3rty119 has quit IRC (Read error: Operation timed out) [03:28] *** nyany_ has joined #archiveteam-bs [03:37] *** odemgi has joined #archiveteam-bs [03:43] *** odemgi_ has quit IRC (Read error: Operation timed out) [05:15] *** phuzion has quit IRC (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) [06:32] *** Flashfire has quit IRC (Remote host closed the connection) [06:32] *** kiska has quit IRC (Remote host closed the connection) [06:33] *** kiska has joined #archiveteam-bs [06:33] *** Fusl sets mode: +o kiska [06:33] *** Fusl_ sets mode: +o kiska [06:33] *** Flashfire has joined #archiveteam-bs [07:03] *** antomati_ has joined #archiveteam-bs [07:04] *** antomatic has quit IRC (Read error: Operation timed out) [07:49] *** tomaspark has joined #archiveteam-bs [08:19] *** odemgi_ has joined #archiveteam-bs [08:22] *** odemgi has quit IRC (Read error: Connection reset by peer) [09:23] *** killsushi has quit IRC (Quit: Leaving) [09:56] *** Smiley has quit IRC (Read error: Operation timed out) [10:02] *** Smiley has joined #archiveteam-bs [10:06] *** d5f4a3622 has quit IRC (Read error: Connection reset by peer) [10:45] *** Dragnog2 has joined #archiveteam-bs [11:05] *** yuitimoth has quit IRC (Quit: Comme รง'est mignon) [11:08] *** yuitimoth has joined #archiveteam-bs [11:23] *** antomati_ is now known as antomatic [12:05] https://gitlab.com/eythian/warc-uploader just in case anyone finds it useful, I made a thing to make uploading WARCs to IA a bit easier for me. [12:39] https://gist.github.com/ivan/079530350ac94851d581b55b1d372440 [12:58] *** systwi_ has joined #archiveteam-bs [13:00] *** systwi_ has quit IRC (Read error: Connection reset by peer) [13:01] *** systwi_ has joined #archiveteam-bs [13:03] *** systwi has quit IRC (Read error: Operation timed out) [13:12] I always find using envvars for that kind of thing to not be very nice. [13:21] *** BlueMax has quit IRC (Read error: Connection reset by peer) [13:51] im not really sure why the rest of the world starting importing that antipattern from docker [14:28] *** systwi has joined #archiveteam-bs [14:35] *** systwi_ has quit IRC (Ping timeout: 600 seconds) [14:58] it's easier to implement parsers for. os.environ.get is easier than having to import getopt and using that to define arg options [15:00] and you can just as easily write daemonize wrappers around such things, (either implicitly or explicitly) inherit the environment variables to the subprocess and/or modify it in the process of doing so [15:02] *** Atom has quit IRC (Quit: Atom) [15:02] *** godane has joined #archiveteam-bs [15:06] oh and security. envs dont show up in the process list while args do. [15:09] some unix variants reveal the env to all other users too, though unlikely to be one that most use regularly. [15:10] (here I did it by having the script query values if needed and dropping them in a config file, which suited my needs.) [15:10] *query secret details [15:15] *** Atom has joined #archiveteam-bs [15:23] "some unix variants" those are configured incorrectly then, this is what should be happening: [15:23] fusl@linux-0lcz:/proc/1> cat environ [15:23] cat: environ: Permission denied [15:25] well, that's linux, which is fine. An older OpenBSD version (5.8?) allowed them to read, though it's since been fixed. It wasn't configuration, it was just what the OS permitted. [15:32] that being said, using environment variables isn't an "antipattern" as abstract mentioned. in fact, the linux kernel itself mostly uses environment variables to pass settings to the init daemon (parts of /proc/cmdline get mapped to /proc/1/environ) to get the job done. if this would be an antipattern, then linux in itself would be an antipattern [15:32] Fusl, thats cool because its like your environment [15:32] its shared [15:32] i think mostly docker just uses it because it used to be nails to get per-container config going [15:33] and now everyone thinks its a good idea [15:33] I usually find it quite unergonomic, that's my main issue. It's awkward to in some usecases. [15:33] well as said, its so much easier and less error prone to just passthrough an entire environment stack to a subprocess than having to parse, modify and rebuild arguments [15:33] it generally doesnt make sense to make application config part of your envuronment [15:34] "now everyone thinks its a good idea", i always though its a good idea because, as mentioned above, linux simply does it that way. just because *bsd doesnt it differently doesnt make it an antipattern or an incorrect usage of it [15:36] though, needs of a kernel and init system can be quite different to those of a small userspace tool, and different again to those of a large userspace application. [15:36] Not everything can be done that way easily though. You can have options appear multiple times in an arg list and parse that into a list, for example. With env vars, you need ugly tricks to achieve the same thing. [15:37] ^ dont get me wrong, argvs have their use cases, like the JAA described ones, but those are mostly dynamic things like supplying a list of urls to something, CLI tools, etc. [15:40] you also can end up with a weird setup where invoking the same command in different terminals can have different results due to different environments, which can be surprising (not a problem if you wrapper script them, but I suspect for simple cases often people just 'export AB=cd' into the shell.) [15:41] (extrapolating myself into many people, of course, is a risky call :) [15:43] export EDITOR=rm; alias vim=rm; there's always a way to fuck up your environments, unavoidable. i am more or less talking about large scale production systems and not development machines or anything like that. [15:45] oh yeah, the large scale systems I work with do it totally differently (a mix of database/distributed db values, puppet-delivered files to a special location, envvars to override certain things for dev purposes, etc.) [15:46] but that's a different case to a 200 line Perl script :) [15:46] 200 line perl scripts can also run in large scale production systems :) [15:47] I work for a company that has multi-millions of lines of perl running, and more than a few 200 line perl scripts gluing things together. I know this very well :) [17:25] I need a C# (c-sharp) programmer willing to do a dirty deed for me. [17:25] https://github.com/adammhaile/Elpis/issues/213 [17:34] *** Pixi has quit IRC (Quit: Pixi) [17:38] *** Pixi has joined #archiveteam-bs [18:04] Move that to -ot, or better yet, somewhere else entirely. [18:21] :'( ok [18:31] *** Mateon1 has quit IRC (Ping timeout: 360 seconds) [18:40] *** Mateon1 has joined #archiveteam-bs [18:43] *** ShellyRol has quit IRC (Ping timeout: 252 seconds) [18:43] *** ShellyRol has joined #archiveteam-bs [18:52] *** Mateon1 has quit IRC (Ping timeout: 745 seconds) [18:52] *** Mateon1 has joined #archiveteam-bs [19:46] *** Pixi` has joined #archiveteam-bs [19:59] *** Pixi has quit IRC (Read error: Operation timed out) [20:23] *** ShellyRol has quit IRC (Ping timeout: 252 seconds) [20:24] *** ShellyRol has joined #archiveteam-bs [20:32] *** d5f4a3622 has joined #archiveteam-bs [20:33] *** ShellyRol has quit IRC (Read error: Connection reset by peer) [20:34] *** ShellyRol has joined #archiveteam-bs [20:55] *** godane has quit IRC (Ping timeout: 246 seconds) [21:09] *** d5f4a3622 has quit IRC (Quit: WeeChat 2.5) [21:12] *** d5f4a3622 has joined #archiveteam-bs [21:26] *** killsushi has joined #archiveteam-bs [22:43] plug the audio out into your audio in [22:47] *** BlueMax has joined #archiveteam-bs [23:21] kpcyrd, do streaming radio services encode track and artist information the same way DFM does? :P [23:25] Again, -ot or wherever, but not here.