*** povbot has joined #maemo-meeting | 09:44 | |
*** qwazix has quit IRC | 13:11 | |
*** qwazix has joined #maemo-meeting | 13:22 | |
*** qwazix has joined #maemo-meeting | 13:22 | |
*** jskarvad has joined #maemo-meeting | 15:41 | |
*** Pali has joined #maemo-meeting | 18:37 | |
*** jskarvad has quit IRC | 19:41 | |
*** Win7Mac has joined #maemo-meeting | 22:09 | |
Win7Mac | Hi! | 22:10 |
---|---|---|
Win7Mac | Meeting... anyone? | 22:11 |
juiceme | Win7Mac, hi! | 22:17 |
juiceme | how's things? | 22:17 |
Win7Mac | hi juice | 22:17 |
Win7Mac | ok I'd say | 22:17 |
Win7Mac | and you? | 22:18 |
juiceme | better, but still with the flu | 22:18 |
juiceme | coughing, no fever | 22:18 |
Win7Mac | get well soon then | 22:18 |
juiceme | hope so, it has been well over a week now with this | 22:19 |
Win7Mac | have you senn last weeks log? | 22:19 |
Win7Mac | http://mg.pov.lt/maemo-meeting-irclog/%23maemo-meeting.2016-02-09.log.html#t2016-02-09T22:47:10 | 22:19 |
juiceme | yes, did read it | 22:20 |
juiceme | I sent the mail peterleinchen had lost back to him | 22:20 |
Win7Mac | topicsgreat | 22:20 |
Win7Mac | *great | 22:21 |
juiceme | :) | 22:21 |
Win7Mac | i mean thanks for resending | 22:21 |
juiceme | yes, the checksums for files need to be available in a known-good location. | 22:22 |
Win7Mac | +1 | 22:23 |
reinob | hi there | 22:30 |
Win7Mac | hi reinob! | 22:31 |
juiceme | reinob, hi! | 22:31 |
Win7Mac | hope at least you are doing well | 22:31 |
reinob | yup, why not? :) | 22:32 |
reinob | peterleinchen around? | 22:33 |
juiceme | nope, not seen him yet | 22:33 |
*** peterleinchen has joined #maemo-meeting | 22:41 | |
*** peterleinchen is now known as Guest10434 | 22:41 | |
juiceme | peterleinchen, hi! | 22:44 |
*** peterlein has joined #maemo-meeting | 22:55 | |
peterlein | hi next try | 22:55 |
peterlein | someone still here? | 22:55 |
reinob | yes :) | 22:56 |
Win7Mac | hi peter | 22:56 |
peterlein | hello | 22:56 |
juiceme | hi | 22:56 |
peterlein | another hello | 22:56 |
peterlein | had some weird irc behaviour ... | 22:57 |
reinob | is your nick taken by somebody else? | 22:57 |
juiceme | happens, especially if you are on mobile | 22:57 |
peterlein | juiceme, thanks for te mail | 22:57 |
juiceme | np | 22:57 |
peterlein | reinob, seems so | 22:58 |
reinob | you don | 22:58 |
peterlein | juiceme, orly | 22:58 |
reinob | you don't use a bouncer, I guess | 22:58 |
juiceme | reinob, what happens to me if I get for example thrown off my nick by a netsplit or something, it will default to "Guestxxx" and I cannot gain ack my nick for maybe 30 minutes or so | 22:59 |
peterlein | used by myself prolly | 22:59 |
reinob | OK | 22:59 |
reinob | I guess I'm not so often here to experience a netsplit. I have znc running and it uses sasl to authenticate directly when joining | 23:00 |
juiceme | mmh, anybody done any AES-cipher implementation lately, prefereably in javascript? | 23:00 |
reinob | juiceme: no, I just looked in my pockets but only found 3DES ;-) | 23:00 |
juiceme | I am totally baffled by just why I cannot make it work... am trying to set up a link between nodejs backend and a client script running on a browser, pretty trivial stuff; comms thru websocket | 23:01 |
juiceme | first tried different codebases for client&server, didi not work. | 23:02 |
juiceme | now I have exactly the same code running on both ends and still I get only gibberish thru | 23:02 |
reinob | many years ago I made a conscious choice not to learn js or java or php. So far I stick to that decision :) | 23:02 |
juiceme | reinob, you need to keep up with times :p | 23:02 |
reinob | but K&R C still rocks :) | 23:03 |
juiceme | the nicety of nodejs is you can run the same shit in server & client :) | 23:03 |
juiceme | try that in c | 23:03 |
reinob | the question is why you'd run the same on the client and on the server.. :) | 23:04 |
juiceme | but of course I know the limitatons; in functional languages you can get yourself in a state you'd never try to go in procedureal languages in the first place :) | 23:04 |
reinob | funny, functional languages should not have "state" :)))) | 23:05 |
juiceme | it is obvious, isn't it? if you run same code in both, your data structs work in both without conversins.. | 23:05 |
juiceme | heh, was referring to state-of-mind, more or less | 23:06 |
juiceme | cannot escape that even with js | 23:06 |
reinob | yup I know. Just kidding | 23:06 |
reinob | re. data structs: I'd always prefer some text/dialog-based exchange, ala telnet & co. | 23:07 |
juiceme | but yes, i have done c since the eighties, and now nodejs for about an year | 23:07 |
reinob | when I hear the word "serialization" and "marshalling" and such I get shivers | 23:07 |
juiceme | and of course you get excited by new toys | 23:07 |
reinob | or json! :) | 23:07 |
juiceme | json is pretty wonderful :p | 23:08 |
juiceme | serialization... well did come across that first time in a MSFT context... yes, *shivers* | 23:09 |
juiceme | but that's really 'cos they do all serialization in crazy-bloated-nonconformist-way in MSF classes | 23:10 |
juiceme | json is absolutely beautiful compared to that shit | 23:10 |
reinob | I may some day have look at nodejs and give it a try | 23:12 |
juiceme | hmm maybe I should try it other way around to see if it fails then too... now am sending a ciphered messahe from server to browser domain | 23:12 |
juiceme | you should, if you have free time to tinker with something... | 23:12 |
juiceme | see, the nicety is that there already are just about zillion npm modules you can use to do just about anything | 23:13 |
juiceme | so it is easy to implement small script that does a LOT of stuff | 23:14 |
reinob | I have a question re. referendium preparation (swearfilter and/or council extension) | 23:19 |
juiceme | yes | 23:19 |
reinob | can I put the stuff in the database without fear that some script will automatically start sending e--mails or doing stuff without my control? | 23:20 |
reinob | i.e. can I (we) trigger the process when I/we are sure that all the database bits are OK= | 23:21 |
juiceme | well, yes unless you trigger it yourself :) | 23:21 |
reinob | OK | 23:21 |
reinob | so no cron script messing around | 23:21 |
juiceme | more or less. | 23:21 |
juiceme | you need to know this about email sending; | 23:21 |
juiceme | there exists a cronscript that does it | 23:22 |
juiceme | and it does read the DB automatically. | 23:22 |
juiceme | just be careful you do not input stuff in *that* table | 23:22 |
juiceme | just a moment I check what was it again | 23:23 |
reinob | thx | 23:23 |
juiceme | it's the "outbound_email" table in maemoelections DB | 23:25 |
reinob | OK I put it in my notes | 23:25 |
reinob | thank you | 23:25 |
juiceme | any stuff put there will trigger the cronmailer | 23:25 |
juiceme | :) | 23:25 |
reinob | I will prepare all statements on a text file and send them to you for review and/or questions | 23:27 |
reinob | hopefully before next meeting (weekend or so). | 23:27 |
juiceme | thanks, that'll do | 23:27 |
juiceme | damn. | 23:28 |
reinob | the editing of maemo.org will take some time (i.e. until I have the time). | 23:28 |
reinob | what? | 23:29 |
juiceme | tested the other-way-around | 23:29 |
juiceme | still getting garbage | 23:29 |
juiceme | :( | 23:29 |
reinob | server-client | 23:29 |
reinob | OK | 23:29 |
juiceme | yes, and now from client-to-server | 23:29 |
juiceme | corrupted both ways | 23:29 |
juiceme | and I have exactly same key, printed it out | 23:30 |
reinob | I'm not sure I could help, and the latency is killing me.I can barely type without typos, so I'll probably leave soon | 23:30 |
juiceme | I have no idea what's happening here | 23:30 |
juiceme | righto, you on mobile too? | 23:30 |
reinob | the router goes crazy when live TV also on (also via Internet). BW is fine but latency suffers. Some QoS/priorization crap | 23:33 |
reinob | happens every now and then | 23:33 |
juiceme | ah, okay | 23:34 |
juiceme | I knew there's a catch in these internets the TV companies sell to people :) | 23:34 |
reinob | with 16Mbit it worked perfectly. with 25Mbit latency suffers. Now the TV is off but still the router acts weird. Or maybe it's the provider. Pinging google takes 1000ms when (if) the packet actually comes back | 23:36 |
reinob | normally it's like 40ms | 23:36 |
juiceme | crazy latency, yes | 23:36 |
juiceme | this is just unbelivable, ran another test | 23:37 |
reinob | could be the WLAN but I'm not checking now. It's so random that I never can quite test it properly | 23:37 |
reinob | anyway | 23:37 |
juiceme | (btw nice thing with javascript, just edit&run, no compiling needed...) | 23:37 |
reinob | time to go to bed anyway | 23:37 |
reinob | :) | 23:38 |
juiceme | when I encrypt something in server, and then decrypt it --> beautiful | 23:38 |
juiceme | when I encrypt something in client, and then decrypt it --> beautiful | 23:38 |
juiceme | when I encrypt something in server, send it to client and decrypt it there --> garbage | 23:38 |
juiceme | when I encrypt something in client, send it to server and decrypt it there --> garbage | 23:39 |
reinob | try to send the data unencrypted? | 23:39 |
juiceme | it works | 23:39 |
reinob | maybe its some trivial byte swapping (endianness?) | 23:39 |
juiceme | I check the sent and received data is same | 23:39 |
juiceme | key is same | 23:39 |
reinob | the actual encrypted data is the same? | 23:39 |
juiceme | something like that, but the bloody code is same and it is js all the wayt thru... | 23:40 |
juiceme | yes | 23:40 |
reinob | but decrypting produces garbage? | 23:40 |
reinob | weird | 23:40 |
Win7Mac | NSA! | 23:40 |
juiceme | well not exactly | 23:40 |
juiceme | since every time you encryt the same data with same key, the ciphertext is different of course | 23:40 |
juiceme | that is why it is called "counter mode" | 23:41 |
reinob | but maybe that's it. try just one block, so you don't have to worry about CBC, counter or whatever | 23:41 |
juiceme | but yes, how I do it is I cipher a block of text, send the result to other end and decrypt it in both ends a the same time | 23:42 |
juiceme | local decrypt works, remote does not | 23:42 |
juiceme | hmm yes could do that | 23:42 |
reinob | yeah because you probably use a different counter! | 23:43 |
juiceme | there's also "code-book-mode", when every time same cleartext block produces same ciphertext block | 23:43 |
juiceme | no, the beauty of the CTR mode is that the counter is embedded in the ciphertext block :) | 23:44 |
reinob | but there must be some nonce/IV | 23:44 |
juiceme | I mean, if it wasn't so, how would you communicate between different parties | 23:44 |
reinob | is that also in the ciphertext? | 23:44 |
juiceme | and also, kind of cehcked that thing also; when I start the program again and input a CTR block from previus run, it will decrypt correctly :o | 23:45 |
juiceme | the counter? yes, embedded there | 23:45 |
juiceme | it's like; if you encrypt the same cleartext 10 different times, you get 10 different ciphertexts; however each of them decrypts back to the original cleartext with the same key :) | 23:47 |
juiceme | and I already tested it works correctly, just like that | 23:47 |
juiceme | only when I try to decrypt the same blocks in "the other system" it fails | 23:48 |
juiceme | :( | 23:48 |
juiceme | ahh, I guess I'll go to sleep on it now | 23:48 |
juiceme | too much headbanging and I just am not going forward with this now.... | 23:49 |
juiceme | so , gnite! | 23:49 |
Win7Mac | gnite | 23:50 |
*** Win7Mac has quit IRC | 23:50 | |
reinob | good night | 23:50 |
*** peterlein has quit IRC | 23:54 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!