#archiveteam-bs 2015-03-28,Sat

↑back Search

Time Nickname Message
00:15 🔗 toad1 has joined #archiveteam-bs
00:58 🔗 mistym has quit IRC (Remote host closed the connection)
02:00 🔗 mistym has joined #archiveteam-bs
02:29 🔗 dashcloud has quit IRC (Read error: Operation timed out)
02:36 🔗 dashcloud has joined #archiveteam-bs
03:02 🔗 schbirid2 has joined #archiveteam-bs
03:03 🔗 schbirid has quit IRC (Read error: Operation timed out)
04:10 🔗 aaaaaaaaa has quit IRC (Leaving)
04:30 🔗 mistym has quit IRC (Remote host closed the connection)
04:57 🔗 mistym has joined #archiveteam-bs
05:28 🔗 Jonimus has quit IRC (Ping timeout: 370 seconds)
05:42 🔗 Jonimus has joined #archiveteam-bs
05:52 🔗 Jonimus has quit IRC (Ping timeout: 370 seconds)
06:04 🔗 Jonimus has joined #archiveteam-bs
06:14 🔗 godane i'm getting Wentworth Millitary Academy Yearbooks
06:15 🔗 godane i found them when trying to find old pri the world podcast episodes in google
08:02 🔗 vitzli has joined #archiveteam-bs
08:03 🔗 vitzli has quit IRC (Client Quit)
08:33 🔗 dashcloud has quit IRC (Read error: Operation timed out)
08:39 🔗 dashcloud has joined #archiveteam-bs
08:46 🔗 godane i'm starting to upload Westworth Military Academy Yearbooks: https://archive.org/details/Westworth_Military_Academy_Yearbook_1881
09:00 🔗 mistym has quit IRC (Remote host closed the connection)
09:18 🔗 godane also i'm grabbing ARG Netcasts that are in wayback machine
09:18 🔗 godane there site is down/gone
09:19 🔗 godane good news is most of episodes are in wayback
09:35 🔗 schbirid2 nice, wpull finished with 200kb left on the harddisk
09:39 🔗 godane what wore you downloading
09:43 🔗 Ctrl-S you need a bigger HDD
09:44 🔗 schbirid2 some forums
09:49 🔗 dashcloud has quit IRC (Read error: Operation timed out)
09:55 🔗 dashcloud has joined #archiveteam-bs
10:29 🔗 johtso has joined #archiveteam-bs
10:31 🔗 johtso schbirid2, yeah, parallel definitely wont understand the spaces in the paths
10:32 🔗 johtso not sure how to expand a glob and deal with the spaces..
10:32 🔗 johtso maybe I should be piping into parallel instead
10:32 🔗 schbirid2 hm, i never had that problem
10:32 🔗 schbirid2 maybe my script is bad?
10:33 🔗 schbirid2 i use "parallel ffmpeg -i {} -acodec libvorbis -aq 0 -y -vn {}.ogg ::: *.mp3" all the time with spaces in filenames, no issues
10:34 🔗 schbirid2 oh lol
10:34 🔗 schbirid2 johtso: file=$1
10:34 🔗 schbirid2 not the other way around
10:35 🔗 schbirid2 never trust me again ;)
10:35 🔗 schbirid2 also, of course "parallel bash script.sh {} ::: *.rar" or make it executable then "parallel ./script.sh {} ::: *.rar"
10:38 🔗 johtso schbirid2, doh!
10:38 🔗 johtso works beautifully now
10:38 🔗 schbirid2 :)
10:38 🔗 johtso now just to wait for the next back of downloads to finish and take it for a spin :)
10:39 🔗 johtso schbirid2, quick question, what kind of server would you use for doing quick archive ops, I'm currently doing this on a digitalocean instance
10:39 🔗 johtso free data transfer being a nice plus
10:39 🔗 schbirid2 no idea
10:39 🔗 schbirid2 i have a cheap and huge online.net one
10:40 🔗 schbirid2 via oneprovider.com
10:40 🔗 johtso very small hard drives being pretty annoying..
10:40 🔗 schbirid2 20€ for 2tb hdd
10:40 🔗 johtso ah okay
10:41 🔗 johtso I quite liked the idea of being able to only pay for the instance while it was needed
10:56 🔗 johtso this data archiving lark could get rather addictive :)
11:33 🔗 dashcloud has quit IRC (Read error: Operation timed out)
11:44 🔗 dashcloud has joined #archiveteam-bs
12:18 🔗 BlueMaxim has quit IRC (Quit: Leaving)
12:55 🔗 schbirid2 has quit IRC (Read error: Operation timed out)
12:58 🔗 schbirid2 has joined #archiveteam-bs
13:09 🔗 dashcloud has quit IRC (Read error: Operation timed out)
13:16 🔗 dashcloud has joined #archiveteam-bs
13:25 🔗 primus104 has joined #archiveteam-bs
14:10 🔗 joepie91_ [11:39] <johtso> free data transfer being a nice plus
14:10 🔗 joepie91_ that is offered by nearly every VPS provider :)
14:10 🔗 johtso not AWS :P
14:10 🔗 joepie91_ hence 'nearly'
14:10 🔗 joepie91_ AWS is not a very good thing to compare to
14:10 🔗 joepie91_ it's overpriced, unreliable, high on vendor lock-in
14:10 🔗 johtso yeah
14:11 🔗 joepie91_ and the fee structure is ridiculous
14:11 🔗 johtso I'm seeing some very cheap VPS options out there..
14:11 🔗 joepie91_ johtso: http://vps-list.cryto.net/index.php?action=list (somewhat outdated)
14:11 🔗 joepie91_ and there's vpsboard and lowendtalk
14:12 🔗 joepie91_ and a few other VPS comparison things but I can't recommend those
14:12 🔗 joepie91_ as they use affiliate links
14:13 🔗 ersi lol aws
14:15 🔗 johtso this seems pretty cheap http://backupsy.com/
14:15 🔗 joepie91_ johtso: and don't forget that there's kimsufi ;)
14:16 🔗 joepie91_ johtso: http://www.kimsufi.com/en/
14:17 🔗 johtso now that's cheap..
14:17 🔗 joepie91_ and leaseweb: https://www.leaseweb.com/bare-metal-server/all-servers
14:18 🔗 joepie91_ and online.net, but I don't like their network: http://www.online.net/en/dedicated-server/dedicated-server-overview-perso
14:18 🔗 joepie91_ and hetzner: http://www.serverbidding.com/
14:18 🔗 joepie91_ 2 x 3TB for 30eur/mo right now, it seems
14:19 🔗 joepie91_ https://robot.your-server.de/order/market/398741/country/NL/culture/en_GB#398741
14:21 🔗 joepie91_ or 10.5TB for 57 eur/mo: https://robot.your-server.de/order/market/page/4/sortcol/hd/sorttype/down/hdsize/1000/limit/100#412236
14:21 🔗 joepie91_ er, might as well get 12TB for the same price then: https://robot.your-server.de/order/market/page/4/sortcol/hd/sorttype/down/hdsize/1000/limit/100#404248
14:21 🔗 joepie91_ so yeah :P
14:25 🔗 johtso thanks for the links!
14:25 🔗 johtso there's a lot out there..
14:26 🔗 johtso these seem tempting too.. especially the ssd-cached options https://www.ramnode.com/vps.php
14:31 🔗 johtso that's a lot of data waiting to be dealt with :P https://archive.org/catalog.php?identifier=MadrotterTreasureHuntBlogDownloads
14:37 🔗 primus104 has quit IRC (Leaving.)
15:16 🔗 DopefishJ has joined #archiveteam-bs
15:19 🔗 pir^2 has joined #archiveteam-bs
15:19 🔗 serapeum has quit IRC (Read error: Connection reset by peer)
15:20 🔗 deathy_ has joined #archiveteam-bs
15:20 🔗 DFJustin has quit IRC (Ping timeout: 417 seconds)
15:20 🔗 torvik has quit IRC (Ping timeout: 417 seconds)
15:20 🔗 LittUp has quit IRC (Ping timeout: 417 seconds)
15:20 🔗 deathy has quit IRC (Ping timeout: 417 seconds)
15:20 🔗 deathy_ is now known as deathy
15:20 🔗 Zebranky has joined #archiveteam-bs
15:20 🔗 Zebranky_ has quit IRC (Read error: Connection reset by peer)
15:25 🔗 pir^2 has quit IRC (bye)
15:27 🔗 deathy has quit IRC (Ping timeout: 393 seconds)
15:27 🔗 johtso has quit IRC (Ping timeout: 393 seconds)
15:27 🔗 balrog has quit IRC (Ping timeout: 393 seconds)
15:27 🔗 marvinw has quit IRC (Ping timeout: 393 seconds)
15:27 🔗 johtso has joined #archiveteam-bs
15:27 🔗 deathy has joined #archiveteam-bs
15:30 🔗 torvik has joined #archiveteam-bs
15:38 🔗 serapeum has joined #archiveteam-bs
15:38 🔗 deathy has quit IRC (Ping timeout: 498 seconds)
15:38 🔗 jk[SVP] has quit IRC (Ping timeout: 498 seconds)
15:38 🔗 deathy has joined #archiveteam-bs
15:39 🔗 LittUp has joined #archiveteam-bs
15:40 🔗 marvinw has joined #archiveteam-bs
15:46 🔗 jk[SVP] has joined #archiveteam-bs
15:48 🔗 LittUp has quit IRC (Ping timeout: 488 seconds)
15:48 🔗 deathy has quit IRC (Ping timeout: 488 seconds)
15:48 🔗 Kazzy has quit IRC (Ping timeout: 488 seconds)
15:48 🔗 LittUp has joined #archiveteam-bs
15:48 🔗 deathy has joined #archiveteam-bs
15:48 🔗 aaaaaaaaa has joined #archiveteam-bs
15:49 🔗 Kazzy has joined #archiveteam-bs
15:51 🔗 balrog has joined #archiveteam-bs
15:51 🔗 swebb sets mode: +o balrog
16:02 🔗 Nertsy has quit IRC (Quit: Nertsy)
16:08 🔗 Nertsy has joined #archiveteam-bs
16:10 🔗 joepie91_ johtso: oh, still there?
16:10 🔗 joepie91_ (yes, ramnode is very very good, but not quite as bang-for-the-buck storage-wise as some other hosts, though certainly better than most)
16:10 🔗 johtso joepie91_, yep :)
16:10 🔗 joepie91_ johtso: another suggestion!
16:10 🔗 joepie91_ https://servercrate.com/storage-vps
16:11 🔗 joepie91_ 10gbit uplink, too
16:11 🔗 johtso wow, nice!
16:12 🔗 joepie91_ not quite the disk perf of ramnode, but the ramnode disk perf is hard to beat anyway :)
16:12 🔗 joepie91_ but certainly more than enough
16:13 🔗 joepie91_ 1073741824 bytes (1.1 GB) copied, 3.03315 s, 354 MB/s
16:13 🔗 joepie91_ (sequential)
16:30 🔗 johtso joepie91_, a big pipe isn't very useful when I'm only getting 20Mbps upload to IA :(
16:33 🔗 primus104 has joined #archiveteam-bs
17:08 🔗 joepie91_ johtso: hehe, you can parallelize
17:08 🔗 joepie91_ basically, parallelize away until you get a 522
17:09 🔗 joepie91_ (iirc)
17:09 🔗 johtso joepie91_, just run multiple instances of "ia upload ident --delete" on a single directory?
17:12 🔗 joepie91_ johtso: no
17:12 🔗 joepie91_ if ia offers a parallel option, use that
17:12 🔗 joepie91_ otherwise, one ia process per item
17:12 🔗 johtso oh yeah, that'll fail when it gets to a file that's deleted
17:12 🔗 johtso oh WHAT? nice!!
17:12 🔗 joepie91_ (per IA item, not per file)
17:12 🔗 joepie91_ johtso: you may want to look into GNU parallel
17:12 🔗 joepie91_ it's specifically for this kind of thing
17:13 🔗 joepie91_ takes a list of paths
17:13 🔗 joepie91_ and manages X processes concurrently
17:13 🔗 joepie91_ giving each process a chunk of the paths
17:13 🔗 johtso yeah, I've been playing around with gnu parallel today for validating the rars and zips
17:14 🔗 vitzli has joined #archiveteam-bs
17:14 🔗 johtso oh, "if ia offers a parallel option"
17:14 🔗 johtso missed the if :)
17:14 🔗 joepie91_ :P
17:14 🔗 joepie91_ it does for metadata
17:14 🔗 joepie91_ not sure if also for uploads
17:21 🔗 johtso joepie91_, so many options, struggling to find the right ones for concurrency and job size
17:22 🔗 johtso okay -P for job slots
17:30 🔗 mistym has joined #archiveteam-bs
17:33 🔗 johtso joepie91_, parallel -j 8 -N 10 --eta ia upload MadrotterTreasureHuntBlogDownloads {} --no-derive --delete --checksum ::: ./alpha/*
17:33 🔗 johtso awesome
17:35 🔗 primus104 has quit IRC (Leaving.)
17:48 🔗 johtso 503 Server Error: Slow Down :(
17:49 🔗 godane i get those sometimes
17:51 🔗 Sanqui https://i.imgur.com/removed.png
17:52 🔗 Sanqui does anybody know why could imgur remove an image in less than 24 hours??
17:53 🔗 schbirid2 uploader can delete
17:54 🔗 johtso wonder when IA will forgive me..
17:54 🔗 johtso take my damn data already :P
17:54 🔗 Sanqui I don't think he did, but I'll ask him
17:57 🔗 godane so i found some video of linux 2002 korea conference
17:58 🔗 godane i was briefly in kbs korea culture news for 2002-10-25
18:27 🔗 Jonimus has quit IRC (Ping timeout: 370 seconds)
18:32 🔗 vitzli has quit IRC (Quit: Leaving)
18:37 🔗 primus104 has joined #archiveteam-bs
18:58 🔗 DopefishJ is now known as DFJustin
19:06 🔗 Jonimus has joined #archiveteam-bs
19:23 🔗 joepie91_ johtso: 503s aren't on a per-client basis
19:23 🔗 joepie91_ they're system-wide
19:25 🔗 mistym_ has joined #archiveteam-bs
19:31 🔗 mistym has quit IRC (Read error: Operation timed out)
19:39 🔗 johtso joepie91_, ah, so I've just overloaded the s3 server?
19:42 🔗 johtso wow.. still getting a 503 2 hours later :(
19:47 🔗 SN4T14_ has joined #archiveteam-bs
19:51 🔗 kyan maybe once it lets you upload again, you could tar the files together so there aren't as many S3 requests being made. I think that would pl
19:51 🔗 kyan *would probably keep you from hitting the rate limiting.
19:55 🔗 SN4T14__ has quit IRC (Ping timeout: 512 seconds)
20:06 🔗 joepie91_ johtso: it just means the infra is a bit busy
20:06 🔗 joepie91_ kyan: iirc the explanation was that the 503 is based on the task queue
20:06 🔗 joepie91_ and it's largely derives that are responsible for the task queue
20:06 🔗 kyan Aah
20:06 🔗 joepie91_ so it's likely just unusually busy right now
20:06 🔗 kyan cool
20:07 🔗 johtso I am using --no-derive too
20:07 🔗 joepie91_ johtso: yeah, it's prob not your jobs :)
20:07 🔗 joepie91_ I'd wait a day or so and retry
20:07 🔗 joepie91_ or half a day
20:07 🔗 johtso I broke everything! :)
20:07 🔗 joepie91_ beat the timezones
20:07 🔗 joepie91_ :p
20:07 🔗 joepie91_ anyway
20:07 🔗 joepie91_ good news everyone!
20:07 🔗 joepie91_ .t http://arstechnica.com/business/2015/03/eu-announces-plans-to-banish-geo-blocking-modernize-copyright-law/
20:07 🔗 joepie91_ er
20:07 🔗 joepie91_ .title http://arstechnica.com/business/2015/03/eu-announces-plans-to-banish-geo-blocking-modernize-copyright-law/
20:07 🔗 botpie91 joepie91_: EU announces plans to banish geo-blocking, modernize copyright law | Ars Technica
20:08 🔗 johtso great news!
20:08 🔗 Kazzy this is good
20:09 🔗 johtso first step europe, next step the world
20:09 🔗 joepie91_ first step modernizing, next step abolishment ;)
20:09 🔗 johtso true
20:10 🔗 joepie91_ one comment in the comments thread that I particularly like
20:10 🔗 joepie91_ "Either than MEP is bought and paid for, or they might actually be dumber than some of the cleverer rocks."
20:20 🔗 espes__ has joined #archiveteam-bs
20:20 🔗 aaaaaaaaa If it is anything like the US, it will get introduced and the amended to the point where you'd wish they'd left it alone.
20:21 🔗 aaaaaaaaa Of course, Europe doesn't seem quite as bought-and-sold as the US is; but that might just be an outsider not seeing the sausage made.
20:26 🔗 antomatic Presumably this will be as successful as the 1989 'Television Without Frontiers' directive, which supposedly promised free movement and reception across europe, etc
20:26 🔗 antomatic Industry said "no" and the European Parliament spent the next 20 years issuing reports saying how successful it was
20:34 🔗 Start has quit IRC (Read error: Connection reset by peer)
20:35 🔗 Start has joined #archiveteam-bs
20:42 🔗 DFJustin has quit IRC (Quit: IMHOSTFU)
20:42 🔗 DFJustin has joined #archiveteam-bs
20:42 🔗 swebb sets mode: +o DFJustin
20:57 🔗 BlueMaxim has joined #archiveteam-bs
21:49 🔗 joepie91_ man, wtf irc
21:49 🔗 joepie91_ "user" (ie. the equivalent of 'ident' in rfc2812) is 1 to unlimited characters and can contain tab, bell, and various other nonprintables
21:49 🔗 joepie91_ yes, the bit after the exclamation mark
22:17 🔗 schbirid2 joepie91_: i got CDs for you \o/
22:19 🔗 Rotab haha, what
22:31 🔗 joepie91_ schbirid2: :D
22:31 🔗 joepie91_ for the IRC-curious
22:31 🔗 joepie91_ user = 1*( %x01-09 / %x0B-0C / %x0E-1F / %x21-3F / %x41-FF )
22:31 🔗 joepie91_ grab an extended ASCII table
22:31 🔗 joepie91_ and see exactly how insane that is
22:32 🔗 joepie91_ schbirid2: still have my address btw?
22:32 🔗 schbirid2 i remember almost BEEP getting thrown out of class by our professor BEEP when i wrote a random ascii print loop that BEEPed a lot
22:32 🔗 schbirid2 should have, will confirm before i send them
22:32 🔗 schbirid2 dont expect them too soon
22:32 🔗 schbirid2 but they are now at my place
22:33 🔗 schbirid2 gonna check which ones i want to keep for sentimental value and i will ask the kultcds.com guy first if he needs specific ones
22:33 🔗 joepie91_ (fun fact: some xchat builds will STILL beep when you send them a bell char - appending it to random messages in a large IRC channel is a great source of amusement)
22:34 🔗 joepie91_ schbirid2: no hurry :P
22:34 🔗 joepie91_ also, doesn't sending it to kultcds mean no images
22:36 🔗 schbirid2 wait, might that be the cause of the mysterious random neeps i get every once a month?
22:36 🔗 schbirid2 i hate those
22:36 🔗 joepie91_ hehe
22:36 🔗 schbirid2 kultcds has a limit on how recent discs are i think
22:36 🔗 schbirid2 older ones are online
22:36 🔗 dan_ oh dear, good job irc specs
22:36 🔗 joepie91_ schbirid2: computer bild spiele shows 0 images
22:36 🔗 joepie91_ only PDFs
22:36 🔗 joepie91_ as far as I can tell
22:37 🔗 joepie91_ dan_: best part is that this is the newer one
22:37 🔗 joepie91_ that supersedes the old one which DID have a limit
22:37 🔗 joepie91_ lol
22:37 🔗 schbirid2 ooh, incubation full version
22:37 🔗 schbirid2 such a great game
22:37 🔗 joepie91_ (old RFC relied on ident RFC limitations - the entire notion of ident has been abandoned in 2812)
22:37 🔗 dan_ ha, nice
22:37 🔗 schbirid2 oooh biing, such memories
22:37 🔗 joepie91_ the joys of writing an IRCd implementation...
22:38 🔗 dan_ there's a reason lots of implementors basically ignore the newer rfcs as invalid
22:38 🔗 joepie91_ hehe
22:38 🔗 joepie91_ IRC is a mess
22:38 🔗 joepie91_ (but it's OUR mess)
22:38 🔗 dan_ actually helping write a python ircd right now, lots of fun to go through
22:39 🔗 joepie91_ I tried a Python IRCd
22:39 🔗 joepie91_ gave up
22:39 🔗 joepie91_ now writing one in Node.js
22:39 🔗 joepie91_ going much better
22:39 🔗 schbirid2 screenfun 2/03
22:39 🔗 dan_ haha, fair enough
22:39 🔗 schbirid2 1/00
22:39 🔗 schbirid2 err cbs
22:40 🔗 joepie91_ dan_: https://github.com/joepie91/circd/blob/master/ircd.py
22:40 🔗 schbirid2 1/02
22:40 🔗 schbirid2 12/99
22:41 🔗 dan_ ah shiny, https://github.com/kaniini/mammon here
22:42 🔗 dan_ (part of that ircv3 crowd here, so we're using it to prototype new implementations and specs and such, lots of fun)
22:42 🔗 joepie91_ ha, kaniini writing a new one?
22:42 🔗 joepie91_ (say hi for me)
22:42 🔗 dan_ haha yep, will do
22:43 🔗 schbirid2 ok, this is mostly poop
22:43 🔗 schbirid2 but hey
22:43 🔗 schbirid2 early 2000s
22:43 🔗 schbirid2 to mid
22:43 🔗 joepie91_ schbirid2: archived poop is best poop
22:43 🔗 dan_ working with about three groups that absolutely hate each other
22:43 🔗 * schbirid2 hands joepie91_ a jar
22:43 🔗 joepie91_ dan_: sounds like IRC then
22:43 🔗 dan_ don't you love the irc community? :)
22:43 🔗 joepie91_ hehe
22:43 🔗 joepie91_ I'm surprised it's only three
22:43 🔗 joepie91_ I'd expected a different order of magnitude
22:44 🔗 joepie91_ say, 10 warring networks times 3 warring opers on each of them
22:44 🔗 dan_ haha, when I work with more groups I'm sure it'll grow
22:44 🔗 joepie91_ dan_: anyway, I've been wanting to implement ircv3
22:44 🔗 joepie91_ but the documentation really confuses me
22:45 🔗 joepie91_ it's built off 14x RFC
22:45 🔗 joepie91_ but it's unclear to me how one writes an implementation that supports clients targeted at 14x, 28x AND ircv3
22:46 🔗 joepie91_ presumably it is possible, but I have no idea how
22:46 🔗 joepie91_ kind of a barrier to adoption :)
22:47 🔗 joepie91_ dan_: so yeah, I'd really recommend writing a short document for implementors that basically comes down to "ignore these things from these RFCs, implement these things from these RFCs, then implement these IRCv3 RFCs in this order, and then your IRCd should work with all common clients"
22:47 🔗 dan_ fair enough, the ircv3 stuff is designed to be hidden behind the CAP command, so that they're only implemented when the client requests them
22:47 🔗 joepie91_ yes, but the 14x root causes issues
22:47 🔗 dan_ but that does make sense, especially with the 2x rfcs
22:48 🔗 joepie91_ the ancestry is not linear
22:48 🔗 joepie91_ which makes it hard to determine what supersedes what
22:48 🔗 joepie91_ and it seems like a step back to me
22:48 🔗 joepie91_ since everything's optional with CAP, that means you're effectively forced to implement a 14x IRCd
22:48 🔗 joepie91_ with optional IRCv3
22:49 🔗 joepie91_ and no apparent ability to do stuff according to 28x *at all*
22:49 🔗 joepie91_ and the lack of docs on this makes it hard for me to determine whether that impression is accurate or not
22:49 🔗 joepie91_ (hence just writing an 28x implementation for now and worrying about SSLv3 later, which is probably not the desirable option, but the only realistic option for an implementor right now who isn't intimately familiar with the edge cases of IRC)
22:50 🔗 dan_ I get what you mean, irc standards are a bloody mess
22:50 🔗 dan_ yeah, for a production ircd that makes sense
22:51 🔗 joepie91_ but I'd really like to implement IRCv3 stuff, so please write docs like I suggested above ;)
22:51 🔗 dan_ there are a few guys working on something along those lines, but I'll have a chat and see what sorta stuff in that direction we could do, since it'd be really useful
22:51 🔗 joepie91_ even if it's just "for experimental implementations"
22:51 🔗 joepie91_ alright, great :)
22:52 🔗 joepie91_ (they don't have to be hand-holding docs btw, as long as they give enough information to figure out my way)
22:53 🔗 joepie91_ (and of course I'm speaking from my personal POV here, but I suspect this would be similar for other implementors)
22:55 🔗 dan_ yeah, the guys I'm talking about were looking at basically taking RFC1x, and then updating it to actually match what production networks use these days, making a new 'descriptive' rfc that actually has everything people need to implement, etc
22:55 🔗 dan_ but I'll see what they're doing with it, try to make sure they get that perspective in there :)
22:55 🔗 joepie91_ right
22:56 🔗 joepie91_ that'd be a great thing
22:56 🔗 joepie91_ 28x seems like it was an effort at something similar
22:56 🔗 joepie91_ ... but then they cheaped out and said "fuck it, unlimited length idents" and more such fun...
22:57 🔗 joepie91_ and I am still amazed at how apparently nobody went "hmm, maybe bell characters in unique user identifiers are not the best of ideas"
22:57 🔗 joepie91_ while writing that spec
22:57 🔗 joepie91_ especially since bells are explicitly excluded elsewhere...
22:57 🔗 dan_ 28x was apparently "only actually implemented on IRCNet, and nowhere else"
22:57 🔗 joepie91_ lol
22:57 🔗 dan_ see also: safe channels, and almost everything but # channels
22:58 🔗 joepie91_ dan_: right now, would you recommend I implement 14x or 28x?
22:59 🔗 dan_ I'm not a big implementer, basically been a bot/services coder most of the time
22:59 🔗 dan_ from what I see, 14x seems more common as a baseline
23:00 🔗 dan_ and then going through, maybe seeing what other implementations take from 28x and implementing those
23:00 🔗 dan_ as well as ISUPPORT/etc that have come out since and aren't in any RFCs
23:01 🔗 joepie91_ ISUPPORT>\
23:01 🔗 joepie91_ ? *
23:03 🔗 dan_ ah: https://github.com/grawity/irc-docs/blob/master/client/RPL_ISUPPORT/draft-hardy-irc-isupport-00.txt
23:03 🔗 dan_ basically takes the 005 numeric, and you get it on connect on almost every network out there
23:04 🔗 dan_ gives the client lots of useful info on channel prefixes you support, nick lengths, casemapping
23:04 🔗 joepie91_ ahhhhhhhh
23:04 🔗 joepie91_ yep
23:04 🔗 joepie91_ definitely using that
23:05 🔗 dan_ aha yep
23:06 🔗 dan_ there's lots of stuff you can implement, but that should give you a decent baseline ircd right now
23:06 🔗 joepie91_ open to further suggestions ;)
23:06 🔗 joepie91_ (fwiw, the IRCd I'm implementing is meant to be highly modular)
23:06 🔗 joepie91_ (as in, there is almost no non-plugin code)
23:07 🔗 dan_ best thing I've found is just looking at another implementation or two, I always have the chary/hybrid source trees open while I'm coding, see how they handle things if I run into troubles
23:07 🔗 joepie91_ mmm
23:07 🔗 joepie91_ those are both C, no?
23:08 🔗 dan_ yeah, based off the same codebase (hybrid), but chary is Atheme's 'newer' ircd, implements some things differently
23:08 🔗 joepie91_ my C/C++ code reading skills are not great :)
23:08 🔗 dan_ but any implementations that are decently stable should do fine, just picked those because they're popular and I work with them a lot :)
23:09 🔗 joepie91_ there's not much non-C IRCd code, unfortunately
23:10 🔗 joepie91_ and so far the only C/C++ codebase I've been able to read through without issues, has been the Node.js codebase
23:10 🔗 joepie91_ (not v8)
23:11 🔗 dan_ ah yeah, that's fair enough
23:27 🔗 joepie91_ ha! victory
23:28 🔗 joepie91_ dan_: https://gist.github.com/joepie91/ee2b8b3d19fd26b13658
23:28 🔗 joepie91_ auto-generating error code \o/
23:28 🔗 joepie91_ (by parsing the RFC)
23:29 🔗 dan_ ooh, nice
23:30 🔗 joepie91_ now I just need to give them fancy names
23:30 🔗 Rotab haha
23:31 🔗 * joepie91_ Achievement Unlocked: Generating code from RFC
23:33 🔗 raylee what are you doing guys?
23:33 🔗 raylee something coding ircds?
23:33 🔗 joepie91_ raylee: writing an IRCd!
23:33 🔗 joepie91_ :P
23:37 🔗 dan_ lots of fun
23:37 🔗 raylee in what language?
23:37 🔗 joepie91_ raylee: coffeescript + Node.js
23:37 🔗 dan_ raylee: python3 here
23:37 🔗 raylee heh
23:37 🔗 raylee if i could be bothered i'd think about writing an ircd in dlang
23:39 🔗 joepie91_ raylee: dlang is memory-safe?
23:40 🔗 raylee you can write it that way
23:40 🔗 raylee or you can write it interfacing with c/c++, using pointers, etc
23:40 🔗 joepie91_ raylee: is it the default? :p
23:40 🔗 raylee dlang being memory-safe is the default
23:41 🔗 joepie91_ okay, approved :D
23:41 🔗 raylee well
23:42 🔗 c_b has joined #archiveteam-bs
23:42 🔗 raylee there's a compiler switch at least, but if you're not using pointers then..
23:42 🔗 * joepie91_ raises eyebrow
23:42 🔗 joepie91_ raylee: if a clueless newbie were to start dlang
23:42 🔗 joepie91_ would his code be memory-safe?
23:43 🔗 antomatic has quit IRC (Read error: Operation timed out)
23:43 🔗 Sk1d has quit IRC (Read error: Operation timed out)
23:43 🔗 antomatic has joined #archiveteam-bs
23:43 🔗 primus has quit IRC (Read error: Operation timed out)
23:44 🔗 Sk1d has joined #archiveteam-bs
23:44 🔗 fenn has quit IRC (Read error: Operation timed out)
23:44 🔗 primus has joined #archiveteam-bs
23:44 🔗 fenn has joined #archiveteam-bs
23:46 🔗 w0rp has quit IRC (Read error: Operation timed out)
23:46 🔗 w0rp has joined #archiveteam-bs
23:48 🔗 brayden has quit IRC (Read error: Operation timed out)
23:49 🔗 joepie91_ dan_: what is the difference between ERR_UNKNOWNMODE and ERR_UMODEUNKNOWNFLAG
23:49 🔗 joepie91_ in 1459
23:52 🔗 dan_ ah, ERR_UMODEUNKNOWNFLAG looks like it's used when there's a nickname parameter with the MODE message
23:52 🔗 joepie91_ example case?
23:52 🔗 dan_ whereas ERR_UNKNOWNMODE just says it's an unknown mode character, so I guess otherwise that's used
23:53 🔗 serapeum has quit IRC (Ping timeout: 606 seconds)
23:55 🔗 dan_ ah, I see
23:56 🔗 dan_ when someone is trying to set modes on a user and the mode character is unknown, you use ERR_UMODEUNKNOWNFLAG
23:56 🔗 dan_ when setting modes on a channel and the character is unknown, ERR_UNKNOWNMODE
23:58 🔗 joepie91_ dan_: aha, so UnknownUserMode vs UnknownChannelMode would be sensible error names?
23:59 🔗 dan_ should be yeah, don't think you can set modes on anything but users/channels but I haven't scanned the rfc for that yet

irclogger-viewer