00:00:59 <ais523> wow, the rules of baseball have the most repetitive URL I've ever seen: http://mlb.mlb.com/mlb/official_info/official_rules/official_rules.jsp
00:01:26 <zzo38> O, yes it is very repetitive
00:02:36 <pikhq> But is it official? Or MLB?
00:02:41 <pikhq> I need to know for certain.
00:08:51 <Koen_> interestingly google tells me mlb may stand for "major league baseball" or "minor league baseball", so your url is repetitive but not unambiguous.
00:15:19 <Koen_> also those appear to be the 2014 official rules, so i'm not sure they work in 2015. and by providing us with this url you committed a copyright infringement: "No part of the Official Baseball Rules may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording or any information storage and retrieval system now known or to be invented, without permission in writing from the Office
00:15:19 <Koen_> Commissioner of Baseball."
00:15:32 -!- L8D has joined.
00:16:25 <zzo38> Is the URL part of the rules too?
00:16:32 -!- AndoDaan has joined.
00:17:20 <ais523> Koen_: I'm not convinced that the URL itself is part of the rules
00:17:20 <Koen_> I guess not but they are a means to transmit them
00:17:26 <ais523> even if it is, you can't copyright single words or short phrases
00:17:32 -!- Patashu_ has joined.
00:17:32 -!- Patashu has quit (Disconnected by services).
00:21:16 <Koen_> ais523: my point is that it is weird to forbid to reproduce or transmit electronically a pdf that is freely available
00:21:34 <ais523> I suspect that they don't want people sharing the PDF via any method other than their website
00:22:23 <Taneb> Although you can transmit via a medium invented at the time of writing but unknown
00:22:31 <Taneb> eg Apache Wave probably
00:31:26 -!- Patashu_ has quit (Remote host closed the connection).
00:31:48 -!- Patashu has joined.
00:34:39 <zzo38> No they mean you aren't allowed to print out and photocopy it.
00:36:12 -!- Patashu_ has joined.
00:36:12 -!- Patashu has quit (Disconnected by services).
00:36:53 <zzo38> Although, writing your own document which is describing the same stuff might be OK.
00:37:30 <zzo38> So that way you could make a document using a different file format such as .txt
00:37:54 <int-e> Koen_: it's funny, you did actually reproduce a part of the PDF.
00:38:20 <zzo38> Yes but that is the copyright notice; they copy that everywhere including in other books too
00:38:29 <int-e> (arguably not a copyrightable part, but more so than a URL)
00:39:14 <ais523> int-e: Debian's copyright parser complained that my summary of the copyright status didn't explain the copyright of the summary of the copyright status
00:39:39 -!- Patashu_ has quit (Remote host closed the connection).
00:39:49 -!- Patashu has joined.
00:39:55 <ais523> so I improvised this: http://trac.nethack4.org/browser/copyright#L5
00:40:13 <Koen_> int-e: it is part of the pdf but is it part of the Official Baseball Rules? also the url isn't part of the rules but it is a means of transmitting them
00:43:02 <int-e> Koen_: the transmission of the rules is performed by the browser once you click on a link on the page that is referred to by the given URL
00:43:32 <int-e> So it's like giving somebody the address of a book shop that sells exactly one book.
00:43:51 <Koen_> well, that's transmitting from the server to my computer, but it was already on ais's computer
00:44:06 <Koen_> and it would never have been on mine if he hadn't given the url
00:44:46 <Koen_> so, even though the pdf did not physically travel from ais's computer to mine, it was transmitted
00:44:56 <Koen_> just like a wave doesn't need matter to travel
00:45:32 <int-e> Koen_: the transmission was done by the mlb.mlb.com server; presumably the owners have the required written permission to perform that transmission.
00:45:47 -!- L8D has quit (Ping timeout: 272 seconds).
00:46:08 <int-e> Koen_: If they do not, it's them that violate copyright, not ais523 or you.
00:46:22 <Koen_> my point is, this irc channel took part in the transmission
00:46:41 <Koen_> then i guess this settles that
00:46:41 -!- L8D has joined.
00:47:01 -!- adu has joined.
00:48:49 <fizzie> I cannot get this Internet configured right. :/
00:49:08 <int-e> fungot: how do you do?
00:49:10 <Taneb> fungot, you're alive!!!
00:49:10 <fungot> Taneb: maybe two, actually, not so complicated you constantly get lost... but i'm sure i'll manage to sort that out :)
00:49:21 <fungot> Available: agora alice c64 ct darwin discworld enron europarl ff7 fisher fungot homestuck ic irc* iwcs jargon lovecraft nethack oots pa qwantz sms speeches ss wp youtube
00:49:27 <fizzie> I have a ugly workaround so that I could get fungot up, but that's all.
00:49:27 <fungot> fizzie: what other kinds of sequences such as make-vector and make-string. later on
00:51:07 <fungot> olsner: that's one of the google image search, even
00:51:23 <int-e> Koen_: What's your take on this: If I have a language model trained on the text of a book, does that make it a derived work of this book? If so, would it still qualify as fair use?
00:52:13 <fungot> Selected style: discworld (a subset of Terry Pratchett's Discworld books)
00:52:21 <fungot> int-e: " you've got me there.' she drummed her fingers on the table, chair,' said cohen.
00:53:13 <fungot> Selected style: irc (IRC logs of freenode/#esoteric, freenode/#scheme and ircnet/#douglasadams)
00:53:34 <fungot> olsner: but i was under the impression that it just compiles code to sql. " you don't have anything fnord right, then
00:54:18 <fungot> oerjan: there was also a friggin' syntax replacement for c would be inconvenient
00:54:29 <fizzie> The Internet is VDSL2 PTM, with PPPoE over the Ethernet frames that go in PTM. It works when I put the bundled router-modem into "routed" mode (so that it negotiates the PPPoE link), and I've even partially got the 8-IP subnet the ISP assigned kinda-sorta working (the router-modem had a "public subnet" mode, which makes it do some sort of proxyarp trickery, but that means all the public IP ...
00:54:35 <fizzie> ... addresses need to be directly in the Ethernet segment formed by the LAN ports of the router-modem, and with distinct MACs), but it's ugly and wastes a public IP for the router-modem.
00:54:39 <fungot> olsner: doesn't chez scheme, and boolean contexts, such as t, notably the fact that i'm using is probably written in c.
00:55:04 <Koen_> int-e: not sure what a language model trained is
00:55:15 <fizzie> Logic says it *should* work just fine if I toggle the router-modem to "bridged" mode, and have my regular Linux box speak PPPoE into the Ethernet interface, but it just doesn't go.
00:55:55 <fizzie> (I've ran the manual "pppoe-discovery" tool and tcpdumped it on the Linux box, and the PPPoE "PADI" packets are going out, but no answers are coming in.)
00:56:27 <fizzie> And the router-modem only has this custom CLI and no way of getting a proper shell (so I could go and see how they've configured it), which is always so infuriating.
00:56:45 <mitchs> all hail fungot, master of fnords
00:56:45 <fungot> mitchs: the argument in the function itself
00:58:42 -!- Patashu_ has joined.
00:58:42 -!- Patashu has quit (Disconnected by services).
01:00:31 <fizzie> int-e: If you have an unpruned fixed-length n-gram model of sufficient length that all the counts for the highest-n grams are 1, you can reconstruct the entire book, which is maybe kind of dubious copyrightically. (Of course you normally wouldn't. Especially since fungot doesn't do backoff, so you'd only get verbatim quotes.)
01:00:31 <fungot> fizzie: and -s restrict that scope
01:15:21 -!- lifthrasiir has joined.
01:20:57 -!- Patashu has joined.
01:23:47 -!- trn has quit (Ping timeout: 272 seconds).
01:23:55 -!- trn has joined.
01:24:21 <zzo38> Sometimes in a 6502 code you might not want to use CMP for comparison so you might use EOR instead, such as if you do not want the carry flag to be affected.
01:24:23 -!- Patashu_ has quit (Ping timeout: 256 seconds).
01:32:51 -!- Koen_ has quit (Quit: The struct held his beloved integer in his strong, protecting arms, his eyes like sapphire orbs staring into her own. "W-will you... Will you union me?").
01:38:53 -!- copumpkin has joined.
01:45:01 -!- Patashu has quit (Disconnected by services).
01:45:01 -!- Patashu_ has joined.
01:45:21 -!- contrapumpkin has joined.
01:46:17 -!- contrapumpkin has quit (Client Quit).
01:46:35 -!- dianne has quit (Ping timeout: 272 seconds).
01:47:07 -!- dianne has joined.
01:47:51 -!- Phantom_Hoover has quit (Quit: Leaving).
01:48:18 -!- copumpkin has quit (Ping timeout: 250 seconds).
01:51:40 -!- AndoDaan has quit (Quit: Going, going, gone.).
02:00:48 -!- ais523 has quit (Ping timeout: 256 seconds).
02:04:42 -!- Patashu has joined.
02:07:49 -!- Patashu_ has quit (Ping timeout: 245 seconds).
02:20:50 -!- ProofTechnique has joined.
02:23:44 -!- rodgort has quit (Ping timeout: 245 seconds).
02:26:42 -!- Patashu_ has joined.
02:26:42 -!- Patashu has quit (Disconnected by services).
02:28:47 <Sgeo> It's possible that I talked about it but never did it
02:29:00 <Sgeo> Thinking about it, I more distinctly remember thinking about Deadfish in Tcl
02:29:11 <zzo38> Do ARCFOUR with a (log2(256!)+16) bit key (this number of bits isn't a integer)
02:33:27 -!- rodgort has joined.
02:48:04 <oren> god damn it is the return type of these numpy functions documented ANYWHERE
02:48:18 -!- Patashu has joined.
02:50:24 <zzo38> I have never seen it so I don't know
02:51:35 -!- Patashu has quit (Remote host closed the connection).
02:51:45 -!- Patashu has joined.
02:51:47 -!- Patashu_ has quit (Ping timeout: 255 seconds).
02:54:09 <oren> Ohhhh so when I do A * B it tries to do it element wise?!?! of all the crap i have to deal with, matrices that don't multiply right
02:55:31 -!- arjanb has quit (Quit: .).
02:58:30 <zzo38> Maybe there is a separate function for matrix multiplication?
02:58:57 <oren> Yeah, I have to do np.dot(A,B) which gets pretty ugly
03:10:31 -!- Patashu has quit (Disconnected by services).
03:10:31 -!- Patashu_ has joined.
03:11:18 <oren> Also conflating dot product with matrix product is... iffy
03:13:23 <oren> really v dot u is v matrix product u transpose
03:13:47 <oren> er, the opposit
03:14:44 <oerjan> hm _surely_ we already had an ~ATH article?
03:15:59 <FreeFull> oren: Yeah, dot doesn't seem right for that
03:19:44 <oren> What we really need is a math library based not on matrices but on Tensors
03:20:01 <zzo38> Yes, that would help too, I believe you
03:21:52 <Sgeo> From Ashes is gorgeous
03:24:09 <oren> well even if my code is made ugly by this library, it works at least
03:25:11 <oren> which my C neural network code does not
03:25:37 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)).
03:28:09 <oren> ~ATH is conceptually interesting but its syntax is generic
03:30:06 <oren> Hmm... would it not be possible to automatically chack obituaries?
03:31:16 -!- Patashu has joined.
03:31:28 -!- ProofTechnique has quit (Ping timeout: 256 seconds).
03:32:33 <oren> Actually, deaths of people would be much easier to check for than animals or objects
03:33:14 <oren> Still very hard, but I bet google could do it
03:34:13 -!- Patashu has quit (Remote host closed the connection).
03:34:27 <oerjan> death of public decency
03:34:31 -!- Patashu has joined.
03:34:53 -!- Patashu_ has quit (Ping timeout: 272 seconds).
03:35:45 -!- ProofTechnique has joined.
03:37:54 <oren> alternatively, an implementation could just assume all named individuals are alive, and use actuarial tables
03:38:08 -!- dianne has quit (Ping timeout: 255 seconds).
03:40:10 -!- dianne has joined.
03:49:16 <zzo38> With the proper virtual tables in SQLite you could receive such information (unless it is already in SQLite format in which case you don't need virtual tables); I have written extension to read JSON and RDF Turtle documents as virtual tables, so if it is those formats you can access them in this way.
03:51:58 -!- ^v has joined.
03:53:06 -!- Patashu has quit (Disconnected by services).
03:53:06 -!- Patashu_ has joined.
03:56:34 -!- Patashu has joined.
04:00:03 -!- Patashu_ has quit (Ping timeout: 265 seconds).
04:02:13 <zzo38> But I don't know what format such actuarial tables would be in?
04:03:46 -!- ^v has quit (Read error: Connection reset by peer).
04:05:40 -!- hjulle has joined.
04:09:31 -!- ProofTechnique has quit (Ping timeout: 246 seconds).
04:15:02 -!- hjulle has quit (Ping timeout: 265 seconds).
04:17:14 -!- Patashu has quit (Disconnected by services).
04:17:14 -!- Patashu_ has joined.
04:17:40 <oerjan> some actual format, i should assume.
04:21:07 <oerjan> ah and just as someone wiped ~ATH from wikipedia's esolang page.
04:28:02 <zzo38> Why do you like to define pi (or tau) in terms of geometry instead of in the other way?
04:28:25 <FreeFull> zzo38: What other way? Infinite series?
04:29:07 <zzo38> Well, there are several possible other ways, but what I was thinking of is e to power of i times tau makes 1
04:29:27 <zzo38> Being the smallest positive solution
04:34:22 -!- L8D has quit (Ping timeout: 265 seconds).
04:35:14 <FreeFull> The principal value of ln(1) is 0
04:35:18 <oren> hmm, wait why doesn't that work
04:36:57 -!- Patashu has joined.
04:38:06 <oren> Greek letters being fullwidth is so inconvenient when the terminal assumes they are half
04:38:08 <zzo38> Zero isn't positive anyways
04:38:38 <zzo38> oren: That's the problem with using Unicode in this way.
04:40:27 -!- Patashu_ has quit (Ping timeout: 256 seconds).
04:42:10 -!- L8D has joined.
04:42:43 <zzo38> For a terminal screen such encoding would seem to work better when wide character are 2-bytes long and narrow character are 1-bytes long.
04:43:28 <oren> Hmm... what happens if I switch the encoding to Shift-JIS?
04:43:48 <pikhq> Sigh, the correct thing in general is wcswidth.
04:44:10 <pikhq> (tells you the width of a wchar_t string in terminal cells)
04:46:00 <oren> Hey that actually works!
04:46:35 <oren> I just switch terminal settings to Shift-JIS and bang, correct spacing
04:47:09 <pikhq> Yeah, your terminal and your everything else need to agree on what the charset is for things to work.
04:47:21 <pikhq> (BTW Greek letters are half width normally. :))
04:48:17 <oren> In Shift-JIS they made all two byte characters fullwidth and all one byte characters half-width... I guess for simplicity
04:48:39 <zzo38> oren: Yes, that's a good idea when doing fix-pitch text.
04:49:25 <zzo38> Unicode is good for using to convert between encodings, that you can store them in tables (you can even add new codes if converting characters that aren't in Unicode), but for general purpose works badly for several reasons really and is pretty stupid as far as I am concerned.
04:49:55 <pikhq> zzo38: Unicode is not as complex as you might think.
04:50:14 <pikhq> Though most things that implement it suck at it.
04:51:49 <zzo38> It is not even complete
04:52:46 <zzo38> Complex scripts and ligatures and text direction and so on should belong in the font metrics file, and program that uses it should only read the font metrics file to determine such thing.
04:55:10 <zzo38> The program should only look at, it is trying to render character code 499488291 therefore you have to look up 499488291 in the font metrics file to determine the proper text direction, kerning, ligatures, default line breaking setting, extending into margin, pointer into font glyph file, etc
04:58:09 -!- Patashu_ has joined.
04:58:09 -!- Patashu has quit (Disconnected by services).
05:01:13 -!- Patashu_ has quit (Remote host closed the connection).
05:01:18 -!- Patashu has joined.
05:10:48 <oren> zzo38: most problems with unicode come from its insisting that you can convert characters from any charset X to unicode and back and not lose anything
05:11:39 <zzo38> oren: That is certainly some of it; as I said, it is even incomplete
05:11:47 <oren> (where anything can mean differences only visible to computer and not onscreen)
05:15:05 <oren> So, unicode is bugward compatible to the bugs of like 30 charsets, combines
05:15:25 <oren> which is why it is so damn buggy
05:15:36 -!- ^v^v has joined.
05:15:57 <zzo38> Either way rendering properties should belong in the metrics file rather than in the program
05:16:20 <zzo38> (Which also has the advantage in case you need to define your own characters)
05:20:06 <oren> I'm not sure under what conditions I would be defining my own characters... I guess conlangers would be ahppy?
05:20:20 <zzo38> Yes, that's one use
05:20:51 -!- Patashu has quit (Disconnected by services).
05:20:51 -!- Patashu_ has joined.
05:24:00 -!- Patashu_ has quit (Remote host closed the connection).
05:24:12 -!- Patashu has joined.
05:26:29 -!- Frooxius has quit (Read error: Connection reset by peer).
05:33:40 -!- Frooxius has joined.
05:42:20 -!- Patashu has quit (Disconnected by services).
05:42:20 -!- Patashu_ has joined.
05:56:21 <zzo38> 8/k7/P2b2P1/KP1Pn2P/4R3/8/6np/8 White to play and win.
06:04:21 -!- Patashu has joined.
06:07:27 -!- Patashu_ has quit (Ping timeout: 250 seconds).
06:07:42 -!- Patashu has quit (Remote host closed the connection).
06:07:54 -!- Patashu has joined.
06:11:08 -!- ProofTechnique has joined.
06:26:06 -!- Patashu_ has joined.
06:26:07 -!- Patashu has quit (Disconnected by services).
06:49:52 -!- Patashu has joined.
06:51:09 -!- Patashu_ has quit (Ping timeout: 265 seconds).
07:03:26 -!- FreeFull has quit (Ping timeout: 256 seconds).
07:07:41 -!- dianne has quit (Quit: Lost terminal).
07:08:21 -!- dianne has joined.
07:09:08 -!- Patashu_ has joined.
07:09:08 -!- Patashu has quit (Disconnected by services).
07:12:19 -!- Patashu_ has quit (Remote host closed the connection).
07:12:56 -!- Patashu has joined.
07:31:03 -!- Patashu has quit (Disconnected by services).
07:31:03 -!- Patashu_ has joined.
07:41:06 -!- adu has quit (Quit: adu).
07:46:57 -!- oerjan has quit (Quit: leaving).
07:53:12 -!- Patashu has joined.
07:56:07 -!- Patashu_ has quit (Ping timeout: 256 seconds).
08:15:03 -!- Patashu has quit (Disconnected by services).
08:15:03 -!- Patashu_ has joined.
08:36:52 -!- Patashu has joined.
08:40:33 -!- Patashu_ has quit (Ping timeout: 264 seconds).
08:58:26 -!- Patashu has quit (Disconnected by services).
08:58:26 -!- Patashu_ has joined.
09:10:50 -!- theTower has joined.
09:11:14 -!- theTower has left.
09:20:42 -!- Patashu has joined.
09:21:42 -!- ^v^v has quit (Read error: Connection reset by peer).
09:22:06 -!- ^v^v has joined.
09:24:09 -!- Patashu_ has quit (Ping timeout: 264 seconds).
09:28:50 -!- Patashu has quit (Quit: Soundcloud (Famitracker Chiptunes): http://www.soundcloud.com/patashu MSN: Patashu@hotmail.com , AIM: Patashu0 , YIM: patashu2 , Skype: patashu0 .).
09:48:00 -!- roasted42 has joined.
10:02:12 -!- roasted42 has quit (Quit: leaving).
10:18:12 -!- L8D has quit (Ping timeout: 255 seconds).
10:23:40 -!- Phantom_Hoover has joined.
10:33:06 -!- TieSleep has changed nick to TieSoul.
10:44:25 -!- dianne has quit (Ping timeout: 256 seconds).
10:46:09 -!- dianne has joined.
11:14:43 -!- L8D has joined.
11:16:06 -!- Koen_ has joined.
11:19:33 -!- L8D has quit (Ping timeout: 264 seconds).
11:35:56 -!- Fleur has joined.
12:46:03 -!- dianne has quit (Quit: Lost terminal).
12:46:37 -!- dianne has joined.
12:49:03 -!- arjanb has joined.
13:03:59 -!- L8D has joined.
13:05:30 -!- SopaXorzTaker has joined.
13:08:42 -!- L8D has quit (Ping timeout: 272 seconds).
13:28:29 -!- boily has joined.
13:53:58 -!- FreeFull has joined.
14:04:41 -!- GeekDude has joined.
14:15:07 -!- adu has joined.
14:46:10 -!- philippos has joined.
14:46:19 -!- philippos has left ("Leaving").
14:52:54 -!- SopaXorzTaker has quit (Remote host closed the connection).
14:58:34 -!- SopaXorzTaker has joined.
15:25:25 -!- SopaXorzTaker has quit (Remote host closed the connection).
15:29:25 -!- SopaXorzTaker has joined.
15:34:07 -!- bb010g has joined.
15:48:09 -!- GeekDude has quit (Ping timeout: 256 seconds).
15:50:58 <HackEgo> [wiki] [[~ATH]] http://esolangs.org/w/index.php?diff=42126&oldid=42125 * GermanyBoy * (+86) added some categories
15:51:43 -!- GeekDude has joined.
16:05:19 -!- dianne has quit (Ping timeout: 252 seconds).
16:07:03 -!- dianne has joined.
16:27:38 -!- S1 has joined.
16:41:16 -!- L8D has joined.
16:43:54 -!- TieSoul has changed nick to TieAFK.
16:45:42 -!- L8D has quit (Ping timeout: 245 seconds).
16:52:46 -!- copumpkin has joined.
16:58:28 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
17:01:04 -!- Phantom_Hoover has joined.
17:01:09 -!- Phantom_Hoover has quit (Changing host).
17:01:09 -!- Phantom_Hoover has joined.
17:29:55 -!- dianne has quit (Quit: Lost terminal).
17:30:16 -!- Gregor has quit (Remote host closed the connection).
17:30:28 -!- dianne has joined.
17:31:37 -!- Gregor has joined.
17:37:11 -!- ^v^v has quit (Read error: Connection reset by peer).
17:37:36 -!- ^v^v has joined.
17:42:16 -!- adu has quit (Quit: adu).
17:43:23 -!- ais523 has joined.
17:47:16 -!- ais523 has quit (Read error: Connection reset by peer).
17:47:23 -!- callforjudgement has joined.
17:50:38 -!- callforjudgement has changed nick to ais523.
17:51:21 -!- Phantom_Hoover has quit (Quit: Leaving).
17:52:17 -!- AnotherTest has joined.
17:56:46 -!- S1 has quit (Quit: S1).
18:01:03 -!- dianne has quit (Quit: brb~).
18:02:59 -!- dianne has joined.
18:14:27 -!- dianne has quit (Quit: Lost terminal).
18:15:21 -!- dianne has joined.
18:20:51 -!- AnotherTest has quit (Remote host closed the connection).
18:22:20 -!- AnotherTest has joined.
18:45:44 -!- SopaXorzTaker has quit (Remote host closed the connection).
18:49:55 -!- boily has quit (Quit: VISCOSE CHICKEN).
18:57:01 -!- arjanb has quit (Ping timeout: 246 seconds).
19:02:53 -!- ais523 has quit.
19:03:01 -!- ais523 has joined.
19:08:52 -!- TieAFK has changed nick to TieSoul.
19:41:20 -!- AndoDaan has joined.
19:58:31 -!- L8D has joined.
20:02:42 -!- L8D has quit (Ping timeout: 244 seconds).
20:18:23 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)).
20:23:52 -!- callforjudgement has joined.
20:25:41 -!- ais523 has quit (Ping timeout: 246 seconds).
20:25:49 -!- callforjudgement has changed nick to ais523.
20:32:17 -!- AnotherTest has quit (Remote host closed the connection).
20:33:33 -!- AnotherTest has joined.
20:38:20 -!- callforjudgement has joined.
20:38:39 -!- ais523 has quit (Disconnected by services).
20:38:39 -!- callforjudgement has changed nick to ais523.
20:46:29 -!- TieSoul has changed nick to TieSleep.
20:52:19 <zzo38> How many esolangs as ARCFOUR been implemented in so far?
20:59:36 -!- ais523 has quit.
20:59:48 -!- ais523 has joined.
21:02:11 -!- L8D has joined.
21:26:59 -!- ais523 has quit (Read error: Connection reset by peer).
21:27:10 -!- ais523 has joined.
21:34:57 -!- CADD has quit (Remote host closed the connection).
21:44:03 -!- boily has joined.
21:45:23 -!- L8D has left.
21:54:02 -!- GeekDude has joined.
22:05:59 -!- MoALTz_ has joined.
22:08:25 -!- MoALTz has quit (Ping timeout: 240 seconds).
22:14:28 -!- AndoDaan has quit (Ping timeout: 244 seconds).
22:16:32 -!- ProofTechnique has quit (Ping timeout: 272 seconds).
22:25:16 -!- AnotherTest has quit (Remote host closed the connection).
22:40:02 <ais523> so, a problem I've come across
22:41:02 <ais523> I need to transfer files to and from a remote system, but my only connection to it is an ssh connection that goes through multiple levels of terminal multiplexers (screen, tmux and the like)
22:41:16 <ais523> it's a Linux system with no direct network connectivity, it has a working C compiler but not much else
22:42:28 <ais523> I've already figured out how to transfer files there (send over a uudecode impl, compile it, then copy-and-paste uucode)
22:42:57 <ais523> but going the other way is rather harder; terminal multiplexers don't intercept a stream of printable characters you're sending to the inside terminal, but they certainly mess with the stream of characters that comes back
22:44:05 <ais523> I guess one method that would work would be to give uucode on the way back a screenful at a time (with delays added), then run a vt100 parser on the result after receiving it and translate back into uucode that way
22:44:10 <ais523> delays would be needed to know where the frames were
22:44:34 <zzo38> Yes that is one way
22:44:46 <coppro> ais523: can you not just open a nc connection?
22:44:58 <zzo38> But why do you even have such a system?
22:45:09 <ais523> coppro: the only connection I have to inside this is through multiple layers of terminal multiplexers
22:45:14 <coppro> ais523: so open a new one
22:45:15 <ais523> basically it's a case of nested VMs
22:45:39 <coppro> or is the innermost one not internet-connected?
22:45:40 <ais523> I think one of them doesn't even have an emulated network card
22:45:45 <ais523> and is using an emulated serial terminal
22:46:01 <ais523> yeah, this seemed like a fun enough problem to ask #esoteric about
22:46:08 <ais523> the solution may well be "change the VM setup somehow"
22:46:22 <zzo38> Yes, if you can change the VM setup, then you probably should
22:46:29 <ais523> however, I'd like to find a solution under the current parameters just out of intellectual curiosity
22:47:23 <zzo38> Yes; I suppose just telling it to display and then parse the display would probably be the way if without changing the VM.
22:49:21 <ais523> but what sort of information do we put on the display? can it be done without delays? perhaps not because the terminal emulators might elide everything that's output, but what if they don't?
22:49:31 <ais523> I'm going to experiment
22:50:14 -!- ais523 has quit (Read error: Connection reset by peer).
22:50:26 -!- ais523 has joined.
22:53:59 <fizzie> ais523: I've used screen's output logging (C-a H, writes to "screenlog.n") for data transfer (with just "uuencode file" writing to standard output on the "sending" side) but you're certainly right that it's conceivable that some terminal-optimization step somewhere on the path might drop stuff.
22:54:16 <ais523> fizzie: right, and there's more than one layer of terminal emulator here
22:54:28 <ais523> I ran `seq 0 10000` and recorded the output
22:54:45 -!- irctc541 has joined.
22:55:03 <ais523> it sometimes repeats sections, with a bunch of "\1b[4C\r" in between
22:55:38 <ais523> and ESC [ 4 C is "cursor 4 columns right"
22:56:03 <ais523> and then at the end it does a "move cursor to top-left"
22:56:08 <ais523> so yeah, a lot of redundant VT-100 in here
22:56:43 <ais523> I haven't noticed any omitted numbers though
22:58:26 -!- irctc541 has quit (Client Quit).
22:58:56 <fizzie> A line-oriented framing format (and a suitably slow speed between lines so that you would "see" every line at least once) combined with a "dumb" observer that just looks for the line start/end markers sounds like it might have some chance of working. Though it's certainly not guaranteed. I remember using lynx-in-screen on a slow-ish serial terminal, and it did quite a lot of trickery for ...
22:59:02 <fizzie> ... redraw, including scrolling arbitrary-looking rectangles around to send as little over the wire as possible.
22:59:26 <fizzie> (I don't know where most of that behaviour came from.)
22:59:45 <ais523> my current plan is to express lines as [sequence number][uucode content] and then strip out all the VT-100 codes
23:00:28 <ais523> we should be able to detect missing sequence numbers, at least
23:02:54 <int-e> that's a funny problem.
23:03:09 <ais523> yep, I thought #esoteric would like it
23:03:13 <int-e> add a checksum to each line?
23:03:33 <ais523> could do, although just md5sum on the resulting file would likely be good enough
23:09:22 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)).
23:38:36 -!- hjulle has joined.
23:45:11 <oren> sounds like we need an implementation of "ip over tmuxxers"
23:45:35 -!- oerjan has joined.
23:54:50 <boily> oren: helloren. ip over tmuxxer?
23:56:29 <zzo38> Yes, make up such program