code logs -> 2013 -> Wed, 18 Dec 2013< code.20131217.log - code.20131219.log >
--- Log opened Wed Dec 18 00:00:52 2013
00:39 dmiles_afk [dmiles@Nightstar-hdcn1r.or.comcast.net] has quit [Connection closed]
00:40 dmiles_afk [dmiles@Nightstar-hdcn1r.or.comcast.net] has joined #code
01:03 Derakon[AFK] is now known as Derakon
01:29 gnolam [lenin@Nightstar-be7b01.cust.bredbandsbolaget.se] has quit [[NS] Quit: Crashy time]
01:32 gnolam [lenin@Nightstar-be7b01.cust.bredbandsbolaget.se] has joined #code
01:32 mode/#code [+o gnolam] by ChanServ
02:28
< Shiz>
https://www.youtube.com/watch?v=qT90jZP58jM
02:28
< Shiz>
daily dose!
02:31 Turaiel is now known as Turaiel[Offline]
02:32
<&McMartin>
Man. The newest thecodelesscode hits a little close to home~
03:35 Turaiel[Offline] is now known as Turaiel
03:48 VirusJTG [VirusJTG@Nightstar-6i5vf7.sta.comporium.net] has quit [[NS] Quit: Program Shutting down]
04:53 Stalker [Z@Nightstar-484uip.cust.comxnet.dk] has quit [Ping timeout: 121 seconds]
05:07 Derakon is now known as Derakon[AFK]
05:21 RichyB [RichyB@Nightstar-c6u.vd5.170.83.IP] has quit [[NS] Quit: Gone.]
05:24 RichyB [RichyB@Nightstar-c6u.vd5.170.83.IP] has joined #code
05:29 Turaiel is now known as Turaiel[Offline]
05:52 Kindamoody[zZz] is now known as Kindamoody
06:05 himi [fow035@Nightstar-q9amk4.ffp.csiro.au] has quit [Connection closed]
06:14 celticminstrel [celticminst@Nightstar-gj43l1.dsl.bell.ca] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
06:21 ErikMesoy|sleep is now known as ErikMesoy
06:21
< jeroud>
McMartin: I like it when they do that.
06:22
< jeroud>
I've become pretty good at writing code that matches the style of the code around it.
06:49
<&McMartin>
Yeah, me too
06:49
<&McMartin>
I'm also in the process of adding new material so that I might have less material overall in the near future.
06:55 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has joined #code
07:06 You're now known as TheWatcher
07:48 Kindamoody is now known as Kindamoody|afk
07:53 himi [fow035@Nightstar-v37cpe.internode.on.net] has joined #code
07:53 mode/#code [+o himi] by ChanServ
08:10 You're now known as TheWatcher[afk]
08:17 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has quit [[NS] Quit: This computer has gone to sleep]
08:21 Kindamoody|afk is now known as Kindamoody
08:52 Kindamoody is now known as Kindamoody|out
09:44 You're now known as TheWatcher
10:03 AverageJoe [evil1@Nightstar-fb1kt4.ph.cox.net] has joined #code
11:16 AverageJoe [evil1@Nightstar-fb1kt4.ph.cox.net] has quit [[NS] Quit: Leaving]
11:18 Stalker [Z@Nightstar-484uip.cust.comxnet.dk] has joined #code
11:39
<@Tarinaky>
Wait... wut...
11:41
<@Tarinaky>
Oh.
12:29
<@Tarinaky>
Does anyone understand Gomry's Method?
12:30
<@Tarinaky>
For err, Linear Integer Programming Problems?
12:31
<@Tarinaky>
I... don't understand the Wikipedia page.
12:59
< RichyB>
So you want to solve a LIP problem. You solve the LP version first. The solution to the LP version might be all-integer-coefficients, in which case you're done.
13:00
< RichyB>
If it isn't, you add an extra constraint to the LP problem that will cut out the optimal LP solution and won't remove any points that aren't integers.
13:00
< RichyB>
http://www.ms.unimelb.edu.au/~moshe/620-362/gomory/ looks like a way easier explanation to follow
13:02
<@Tarinaky>
RichyB: I saw that. I had trouble following it :/
13:10
<@Tarinaky>
I hate exams >.<;
13:10
< RichyB>
Tarinaky, which side is problematic, the how or the why?
13:10
<@Tarinaky>
I'm... strugling to grok how you divine the Gomry's constraint.
13:10 * TheWatcher ...s, has one of Those Moments Of Clarity, finally sees how to do this properly
13:10
<@Tarinaky>
Also it doesn't help that I am having difficulty concentrating on anything due to the plague.
13:15 Vornicus [vorn@Nightstar-sn7kve.sd.cox.net] has quit [Connection closed]
13:15
<@Tarinaky>
Okay. I think I get it... maybe.
13:16
<@Tarinaky>
I need a shower, some asprin and a coffee.
13:27
< RichyB>
Tarinaky, okay, so the HOW of the Gomry's constraint is:
13:29
< RichyB>
Tarinaky, the simplex method gives you a table of equations like "1*x[1] + 0*x_0 + 2.25*slack[1] + -0.25*slack[2] = 2.25"
13:30
< RichyB>
take all of those numbers and split them into integers and positive fractions, like 1 1 + 0, 2.25 2 + 0.25, -0.25 -1 + 0.75
13:31
< RichyB>
1*x[1] + 0*x[1] + 0*x[2] + 0*x[2] + 2*slack[1] + 0.25*slack[1] + -1*slack[2] + 0.75*slack[2]
13:32
<@Tarinaky>
Take the integer part to the LHS
13:32
< RichyB>
gather all of the integer parts on the left hand side of the equation first
13:32
<@Tarinaky>
Cross it out and change the statement to >=
13:32
< RichyB>
yyeup
13:32
< RichyB>
exactly
13:32
<@Tarinaky>
Which leaves you with a greater-than constraint on fractions which 'snips off' the portion of the feasible region containing the non-integer optimal solution
13:32
< RichyB>
*nodnod*
13:33
<@Tarinaky>
I'm struggling to find anything on google for the 'Method of Fictitious Costs' wrt Transport Problems.
13:33
<@Tarinaky>
To verify that a least-cost solution is optimal.
13:33
< RichyB>
No clue.
13:34
<@TheWatcher>
... and, of course, the minute I work out exactly what needs doing, the fucking internet gateway in work falls over
13:51 Turaiel[Offline] is now known as Turaiel
14:32 You're now known as TheWatcher[afk]
14:35
<@froztbyte>
haha
14:35
<@froztbyte>
yes
14:35
<@froztbyte>
that happened to me over the weekend on this other thing I was working on
14:47
<@Tarinaky>
Aaaand another part of my laptop breaks.
14:58
<@Tarinaky>
And the t is dodgy :/
14:58
<@Tarinaky>
Another They Let Mate Math Met
14:58
<@Tarinaky>
Okay, nevermind.
15:15 jeroud [sid10043@Nightstar-a1k27g.irccloud.com] has quit [Connection closed]
15:25
< RichyB>
Does anyone here understand Debian/dpkg/apt? I've got a custom repo, which has a signed Release file. There is *also* an option somewhere or other to have debhelper sign the individual debs as I build them. Is there any actual point to doing the latter, or is it just meant for if I'm, say, gathering debs from other people to put into my repo?
15:27
<@froztbyte>
RichyB: maintaining your own distro/buildset, signing organizational packages, that kind of thing
15:28
<@froztbyte>
"yes, this build of $companySoftware was actually signed by $company and is legit"
15:28
<@froztbyte>
presuming you have all your keys and stuff loaded in the right places
15:31
< RichyB>
It's confusing me because apt is perfectly happy to install stuff with *just* the Release file signed
15:31
< RichyB>
which is of course perfectly sensible since that file contains SHA256 hashes of files which in turn contain SHA256 hashes of the actual deb files
15:51 Turaiel is now known as Turaiel[Offline]
16:17 Turaiel[Offline] is now known as Turaiel
16:28 jeroud [sid10043@Nightstar-a1k27g.irccloud.com] has joined #code
16:29 celticminstrel [celticminst@Nightstar-gj43l1.dsl.bell.ca] has joined #code
16:29 mode/#code [+o celticminstrel] by ChanServ
17:15
<@gnolam>
http://www.cs.tau.ac.il/~tromer/acoustic/
17:24 Turaiel is now known as Turaiel[Offline]
17:27 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has joined #code
17:44
<@froztbyte>
RichyB: because apt considers the repo as a trusted source
17:44
<@froztbyte>
RichyB: the per-file signing would be for a scenario where you release files without actually putting them in a repo
17:45
<@froztbyte>
you could probably find the full policy writeup
17:45
<@froztbyte>
I'm dead tired atm, think I'm getting sick :/
17:47 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has quit [Client closed the connection]
17:48 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has joined #code
17:50 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has quit [[NS] Quit: This computer has gone to sleep]
17:59
< RichyB>
froztbyte, yeah, the way that apt treats the repo as a trusted source makes perfect sense to me.
18:01
<@froztbyte>
well, not trusted
18:01
<@froztbyte>
you need to import the right keys for the trust
18:01
<@froztbyte>
but as a valid data channel, certainly
18:02
<@froztbyte>
that said, you could actually do apt over sftp:// if you felt like it :D
18:10 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has joined #code
18:11
< RichyB>
Okay, okay. The way that apt lets trustworthiness cascade through a series of hashes from the top-level Releases file all the way down to the debs, makes sense.
18:12
< RichyB>
This is one of the reasons why IMHO HTTPS is bullshit
18:12
< RichyB>
*HTTPS is bullshit for tamper-protection, not for secrecy
18:12
< RichyB>
protecting the transport from tampering is pointless when the endpoint could be tampered with
18:12
< RichyB>
protecting the *content* from tampering by signing it is much better
18:13
< RichyB>
bonus: if you don't care about secrecy, you can use cleartext transport and even things like transparent http proxies
18:13
< Syka>
er what
18:13
< Syka>
the content is protected from tampering
18:13
< RichyB>
No it isn't.
18:14
< Syka>
yes it is
18:14
< RichyB>
HTTPS does not protect content from being tampered with.
18:14
< RichyB>
HTTPS protects the *transport*, not the content. If the server is compromised and files modified there, HTTPS does not protect me from being served bad content.
18:15
< Syka>
...that's because that's not HTTPS' job?
18:15
< RichyB>
An awful lot of the www has mistaken that for what HTTPS' job is.
18:15
< Syka>
that's because the server is not the threat
18:15
<&ToxicFrog>
RichyB: er. Who?
18:16
< Syka>
transport layer security is defending against the threat of content being snooped on or modified in transport
18:16
< RichyB>
The server *is* a threat if you want the end-user to be able to trust an application that is running in her browser.
18:16
< RichyB>
Client-side JS apps cannot trust servers that they're retrieving data from.
18:17
< Syka>
...yes but, if the server is pwn'd, it's game over
18:17
< RichyB>
Poppycock. You can pwn an apt server and you won't take me over when I next run apt-get update.
18:17 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has quit [[NS] Quit: This computer has gone to sleep]
18:17
< Syka>
because there is no defence against that, because anything you put in, an attacker who has owned the box can spoof it
18:17
< RichyB>
ToxicFrog, also the assumption which is baked into web-browsers that content delivered via HTTP is suspect on a HTTPS page.
18:18
< Syka>
RichyB: I will if I have got the signing keys
18:18
< RichyB>
Syka, signing keys don't have to be present on the edge servers.
18:19
< RichyB>
ToxicFrog, which IMO reflects an assumption that anything coming in via HTTPS is "safe", ignoring the concept that evil people buy HTTPS certs too.
18:19
< RichyB>
Syka, e.g. a CDN like Akamai has edge servers with SSL certs with 400+ subjectAltNames on them. :)
18:20
< Syka>
that's because youre not trusting what the server gives you in an apt scenario
18:20
< RichyB>
That's supposed to be a *cache*. Instead, it's required to be a trusted object because the design is broken.
18:21
< RichyB>
If you compromise one of Akamai's edge servers (or any other CDN, they all have to do this for HTTPS-using customers), of which they have a lot globally distributed, right now you will get to compromise on the order of 400 sites.
18:26
< RichyB>
BTW, look up content-centric networking. The idea is really cool! It fits what people are actually using the internet for IRL much more closely than TCP does.
18:30
<&ToxicFrog>
RichyB: it's not suspect because it's HTTPS, it's suspect because it can leak information about the page you requested over HTTPS and are thus presumably interested in not having sniffed by everyone.
18:30
<&ToxicFrog>
Er, because it's HTTP on an HTTPS page.
18:31
< RichyB>
This is true.
18:42
< Azash>
http://i.imgur.com/rIgtShO.png
18:55
<@froztbyte>
RichyB: dash likes CCN a lot
18:55
<@froztbyte>
RichyB: but I guess that's where you heard of it
18:55
<@froztbyte>
(I still need to go read properly about it someday)
18:56
<@froztbyte>
RichyB: can you expand upon what you mean with the akamai remark?
18:56
<@froztbyte>
like, do you say it in the sense of attack surface in terms of what's on akamai's global edge, or some other sense?
18:58
<@froztbyte>
<Syka> because there is no defence against that, because anything you put in, an attacker who has owned the box can spoof it
18:58
<@froztbyte>
there is, actually
18:59
<@froztbyte>
but it becomes non-trivial to start implementing
18:59
<@froztbyte>
having content signatures is one of the mechanisms for doing this
18:59
<@froztbyte>
(and this is why you can have debian mirrors without having access to the keys)
19:18 Kindamoody|out is now known as Kindamoody
19:46 Turaiel[Offline] is now known as Turaiel
19:47 Harlow [Harlow@Nightstar-cb5f6d54.cc.il.us] has joined #code
19:51 Harlow [Harlow@Nightstar-cb5f6d54.cc.il.us] has quit [[NS] Quit: This computer has gone to sleep]
20:04
< Azash>
http://security-world.blogspot.fi/2013/12/security-dsa-2821-1-gnupg-security.htm l
20:13
<@gnolam>
Azash: [18:15:50] <gnolam> http://www.cs.tau.ac.il/~tromer/acoustic/
20:13
< Azash>
Oh
20:13
< Azash>
Cheers
20:33 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has joined #code
20:36 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has quit [[NS] Quit: This computer has gone to sleep]
20:39 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has joined #code
20:39 Harlow [Harlow@Nightstar-2dbe3d64.il.comcast.net] has quit [[NS] Quit: BED]
20:41 * McMartin completely misreads the headline "Google recalls melting Chromebook 11 power adapters, will resume sales"
20:41
<&McMartin>
"Oh man. Remember when we melted those power adapters? Good times. Wanna buy one?"
20:41 * Alek snerks.
20:43 Orthia [orthianz@Nightstar-avg.1ee.224.119.IP] has quit [Ping timeout: 121 seconds]
20:48
< simon_>
what's the plural of UPS?
20:48
< simon_>
UPS'es?
20:49
<&McMartin>
That's how I'd say it
20:49
<&McMartin>
I'd write it without the apostrophe, probably.
20:51
< Syka>
I use UPSs
20:52
< simon_>
so UPSes. I know what I'd do in Danish :)
20:52
< simon_>
I was thinking of UPS'es or UPSes (and ABCs in case the last letter wasn't an S)
20:52
< simon_>
thanks!
20:52 Derakon [chriswei@Nightstar-4k2ccr.ca.comcast.net] has joined #code
20:52 mode/#code [+ao Derakon Derakon] by ChanServ
20:53
<&Derakon>
Continuing the saga of my boss requesting a printout of a complete directory listing of the repository...
20:53
<&Derakon>
He claimed that back in the old days when you made a directory, as part of that process you'd also write in a bit of text describing the purpose/contents of the directory (i.e. some kind of attached metadata).
20:53
<&Derakon>
Has anyone heard of a filesystem that worked that way?
20:54
<&Derakon>
I told him about readme files and he seemed to think this was a colossal step backwards.
20:54
<&McMartin>
He'd have to name one.
20:54
<&McMartin>
It's barely possible that CTSS, ITS, Multics, or TOPS-10 did this
20:55
<&McMartin>
In the mid-1960s, however, Bell Labs's experiments with a universal file abstraction bore some fruit and over the next fifty years it has gradually become entirely universal.
20:55
<&McMartin>
Locking in its hold around 1998 or so.
20:55
<&McMartin>
... maybe 2001, once OS X was released and resource forks went away forever.
20:56
<&Derakon>
Heh.
20:56
<&Derakon>
Yeah, I honestly really didn't see the point of what he was asking for...
20:56
<&Derakon>
But oh well.
20:57
< simon_>
sounds like he just wanted to tell you that he once knew computers, too. :)
20:58
<&Derakon>
That's the thing, all of his "practical" knowledge (i.e. knowledge-about-using) of computers is 20 years out of date.
20:58
<&McMartin>
Possibly fifty~
20:58
<&Derakon>
And he never sees the advantages of the new systems.
20:58
<&McMartin>
I was using computers with everything-is-a-file abstractions in the 1980s.
20:58
<&Derakon>
I don't think he really appreciates that I'm doing the job of at least 3-4 oldschool developers here, on my own~
21:00
<@celticminstrel>
I wouldn't say resource forks went away forever with MacOS 10.0...
21:02
< AnnoDomini>
http://www.techdirt.com/articles/20131210/08174425520/norway-to-digitize-all-nor wegian-books-allowing-domestic-ip-addresses-to-read-all-them-irrespective-copyri ght-status.shtml
21:04 redwire [redwire@Nightstar-ohmu6b.eastlink.ca] has joined #code
21:05 Kindamoody is now known as Kindamoody[zZz]
21:06
<@Alek>
o_o
21:08 dmiles_afk [dmiles@Nightstar-hdcn1r.or.comcast.net] has quit [Connection closed]
21:09 dmiles_afk [dmiles@Nightstar-hdcn1r.or.comcast.net] has joined #code
21:09 Orthia [orthianz@Nightstar-avg.1ee.224.119.IP] has joined #code
21:09 mode/#code [+o Orthia] by ChanServ
21:13 Orthia [orthianz@Nightstar-avg.1ee.224.119.IP] has quit [Ping timeout: 121 seconds]
21:14 * ToxicFrog eyes the shit out of bash and/or gnome
21:15
<&ToxicFrog>
So I just spent 20 minutes troubleshooting an issue where I could log in to tty1 but not X
21:15
<&ToxicFrog>
Apparently the issue is:
21:15
<&ToxicFrog>
- my .bashrc sources a file that defines functions with hyphens in the name
21:15
<&Derakon>
Wut
21:15
<&ToxicFrog>
- if the 'posix' shellopt is set ($POSIXLY_CORRECT), this is not permitted(?) and causes the shell to exit(?!?!?!)
21:16
<&ToxicFrog>
- posix is unset by default, unless you invoke bash as /bin/sh, but
21:16
<&ToxicFrog>
- gnome-session starts bash with POSIXLY_CORRECT set, which cases it to die when it hits my .bashrc and kick me back to xdm
21:17 * ToxicFrog gets around this for now by having his bashrc set +o posix at the start and restore it, if set, later
21:17
<&ToxicFrog>
(you can call functions with hyphens in them just fine in posix mode, you just can't define new ones)
21:19
<&ToxicFrog>
Derakon: which part are you whutting at?
21:20
<&Derakon>
The fact that hyphens are disallowed.
21:21
<&Derakon>
Or maybe that they're being used.
21:21
<&Derakon>
Or maybe the fact that this causes the shell to (silently?) exit.
21:21
<&ToxicFrog>
Not silently, it logs an error in .xsession-errors.
21:21
<&Derakon>
Instead of printing a warning and then ignoring the definition or something.
21:21
<&Derakon>
Well, that's something at least.
21:21
<&ToxicFrog>
But yeah, I'd expect it to do the usual bash thing of continuing blindly on.
21:22
<&ToxicFrog>
Also there's no indication in the error why that function name is disallowed, and of course running it from tty1 succeeds.
21:22
<&ToxicFrog>
I figured it out by logging the output of 'env' from both tty1 and X and then diffing the result.
21:23
<&ToxicFrog>
As for why it exits...
21:23
<&ToxicFrog>
"Function names must be valid shell names. That is, they may not contain characters other than letters, digits, and underscores, and may not start with a digit. Declaring a function with an invalid name causes a fatal syntax error in non-interactive shells."
21:23
<&ToxicFrog>
"Non-interactive shells exit if there is a syntax error in a script read with the . or source builtins, or in a string processed by the eval builtin."
21:23
<&ToxicFrog>
Those are two of the rules activated only in posix mode.
21:24
<&Derakon>
And the shell is considered non-interactive because it's just running your X session?
21:24
<&ToxicFrog>
Yes.
21:25
<&ToxicFrog>
An interactive shell, roughly speaking, is one started with -i, or one started with no arguments and connected to a terminal.
21:25
<&ToxicFrog>
This is neither.
21:28
<&McMartin>
I assume the hyphen bit is because otherwise it's mistaken for subtraction
21:28
<&McMartin>
I'm only used to hyphens working in LISPs, to be honest
21:28 Vornicus [vorn@Nightstar-sn7kve.sd.cox.net] has joined #code
21:28 mode/#code [+qo Vornicus Vornicus] by ChanServ
21:31 himi [fow035@Nightstar-v37cpe.internode.on.net] has quit [Ping timeout: 121 seconds]
21:33 Turaiel is now known as Turaiel[Offline]
21:34
< Shiz>
http://www.cs.tau.ac.il/~tromer/acoustic/
21:35
<&McMartin>
Pretty solid result, IMO
21:35
<@gnolam>
And that's the third time. :)
21:35
<&Derakon>
Has that article been getting linked here a lot?
21:35
<&ToxicFrog>
McMartin: the thing is, subtraction is only respected in bash inside $(()), like other mathematical operators
21:36
<&McMartin>
Derakon: Four times now if you count repeats.
21:36
<&ToxicFrog>
And it is fine (and, indeed, common) to have commands with hyphens in them backed by actual files on disk.
21:36
<&Derakon>
Righto.
21:36
< Shiz>
lol
21:36
<&ToxicFrog>
E.g. git-anything-at-all
21:36
<&McMartin>
But yes
21:37
<&McMartin>
"Oh hey guys we can crack 4096-bit GPG with a side-channel attack" is a pretty major result~
21:37 * Derakon vanishes for a bit.
21:39
<@celticminstrel>
I don't use hyphens in my git commands...
21:39
<&McMartin>
git whatever tends to forward to a command named git-whatever
21:39
<&McMartin>
https://groups.google.com/forum/#!msg/golang-dev/3qDvo2aXmMw/xgF7kA9CdngJ
21:39
< Ogredude>
ok that's frightening
21:39
<&McMartin>
"Congratulations, you've found three bugs: one in Go, one in the Plan 9 kernel, and one in the file server being used for /tmp."
21:41 * AnnoDomini does not unnerstan.
21:42
<@gnolam>
One of the better CS labs I've had was cracking DES keys with an oscilloscope by measuring power consumption.
21:42
<&McMartin>
Sounds like that works on GPG too.
21:43
< Ogredude>
just goes to show, security is an illusion
21:44 ErikMesoy is now known as ErikMesoy|sleep
21:44
<@gnolam>
Nah. It goes to show that implementing cryptography in practice is freakin' hard.
21:45
<&McMartin>
Security is as much an illusion as truth is. Don't make the same mistake postmodernists did. =P
21:45
<&ToxicFrog>
celticminstrel: 'git foo' under the hood is almost always invoking /usr/bin/git-foo
21:45
< ErikMesoy|sleep>
Abstractions (like computers) leak.
21:45
<@Tarinaky>
Isn't the abstraction in Cryptography 'math'?
21:46
<&McMartin>
And that abstracts over physical devices.
21:46
<&McMartin>
Which is where this attack lives~
21:46
<@Tarinaky>
I just like the idea of math leaking.
21:46
< ErikMesoy|sleep>
The abstraction is "to math".
21:46
< Syka>
eww, the math is leaking lambdas everywhere
21:46
< ErikMesoy|sleep>
IMO.
21:46
< Shiz>
'get these prime numbers off me D:'
21:47
< ErikMesoy|sleep>
I'm using "leaky" to indicate that the transition from physical device to math is not seamless.
21:47 * Syka gets the mop
21:47
< Shiz>
anyway, math is only a tool for crypto
21:47
< Shiz>
crypto is not an abstraction over math
21:47
<@Tarinaky>
Crypto /is/ math.
21:47
<@Tarinaky>
Except when physically implemented, then you've just got a bunch of idiots reusing OTPs
21:48
<@celticminstrel>
OTPs?
21:48
< ErikMesoy|sleep>
One Time Pads.
21:48
<@Tarinaky>
One Time Pads
21:48
<@celticminstrel>
Ah right.
21:48
< ErikMesoy|sleep>
As the name suggests, they should not be reused. :P
21:48
< Shiz>
crypto is about hiding information
21:49
< Shiz>
math can help in hiding by information with several (so far) nontrivialy solvable issues like the factoring problem
21:49
<@Tarinaky>
OTP is, theoretically, uncrackable... There are known cases of cheating.
21:50
<@Tarinaky>
Shiz: I'm pretty certain even if N=NP or Quantum Computers were scalable tomorrow it'd still count as non-trivial.
21:50
< Shiz>
point is that math is a tool used by crypto
21:51
<@Tarinaky>
What is Information though?
21:51
<@Tarinaky>
The only definition I know is a Math/Stats one...
21:51
< Syka>
i think enigma was part of said cheating
21:51
< Syka>
they built a machine that could decrypt enigma, not one that could really break it, iirc
21:51
<@Tarinaky>
I don't know if you can really talk about Information in a non-maths/intuitive/physical way.
21:52
< Syka>
and they relied on HUMINT to get the daily codes
21:52
<@Tarinaky>
Syka: I don't think Enigma was strictly a OTP.
21:52
<~Vornicus>
no, it was a daily pad.
21:52
<@Tarinaky>
I was thinking of later, during the cold war.
21:52
<@celticminstrel>
Cheating?
21:52
<@Tarinaky>
Things like supplying identical OTP pads to the other side.
21:52
<&McMartin>
IIRC, Enigma had PFS but then they broke *one* code
21:52
<~Vornicus>
If you had enough machines and a way to examine their output you could probably tell which pad they were using
21:53
<@Tarinaky>
Adding deliberate biases to RNGs
21:53
<&McMartin>
And from that you could get the broadcasts that were the next set of rotor settings
21:53
<~Vornicus>
pfs?
21:53
<&McMartin>
Perfect Forward Secrecy
21:54
< Syka>
yeah
21:54
< Syka>
enigma was a OTP
21:54
< Syka>
it came down to being able to decode it
21:54
< Syka>
in force
21:54
<&McMartin>
Hrm
21:54
< Syka>
which is what Bletchley did
21:54
<@Tarinaky>
You can't brute force OTP
21:54
< Syka>
they, IIRC, would feed it shitloads of codes
21:54
<&McMartin>
I'm not sure I accept it's truly a OTP in the normal sense, because the rotors existed.
21:54
<@Tarinaky>
That's the point. That's why it's uncrackable.
21:54
< Syka>
Tarinaky: you can if you have faulty operators!
21:55
< Syka>
Tarinaky: they relied on the Germans being shit at it
21:55
<@Tarinaky>
Syka: That's not brute force.
21:55
<~Vornicus>
This isn't otp
21:55
<~Vornicus>
this is a daily pad.
21:55
<&McMartin>
http://en.wikipedia.org/wiki/Enigma_machine
21:55
<&McMartin>
Enigma was a rotor cipher
21:55
< Syka>
from the wikipedia: "Though Enigma had some cryptographic weaknesses, in practice it was German procedural flaws, operator mistakes, laziness, failure to systematically introduce changes in encipherment procedures, and Allied capture of key tables and hardware that, during the war, enabled Allied cryptologists to succeed."
21:55
<~Vornicus>
Every message from every machine is encrypted with the same key, all day
21:55
<@gnolam>
Syka: key != OTP.
21:55
< Syka>
Vornicus: daily key, then
21:56
<@Tarinaky>
It also helped that Nazi intelligence was being run out of London.
21:56
< Syka>
the point being that you use the key in one instance (the day)
21:56
< Shiz>
maybe they just listened to its mechanical signals
21:57
< Shiz>
:)
21:57
<@Tarinaky>
By which I mean made up of nothing except double-agents.
21:58
<&McMartin>
How did they sort that out? Were German spy records declassified?
21:58
<@gnolam>
Syka: still nothing to do with OTP.
21:58
<@gnolam>
+has
21:59
<@Tarinaky>
The British records are declassified. Every agent sent to England was either a double agent, had a handler who was a double agent or was otherwise 'fed' information.
22:00
<&McMartin>
Tarinaky: Right, but in order to get the conclusion "Double Cross caught them all", you need the list of every agent sent to England, not the list of everyone they caught.
22:01
<@gnolam>
There are two important properties to an OTP - the first is the one-time use, the second is len(key) >= len(message).
22:02
<@Tarinaky>
McMartin: I imagine that would have been an easy feat to accomplish at the end of the war.
22:02
<@Tarinaky>
They did lose.
22:03
<@Tarinaky>
Shopping list of things to acquire at all of Berlin: Military Intelligence, Rocket Scientists...
22:03
<@Tarinaky>
*fall
22:04
<&Derakon>
Just because that knowledge existed at one time doesn't mean it wasn't destroyed.
22:04
<@Tarinaky>
True enough.
22:04
<&Derakon>
"List of spies" is among the items I'd want to shred/burn when my enemy's armies are marching into my city.
22:04
<@Tarinaky>
Fair point, fair point.
22:05
< Syka>
'list of spies' is really something you should have a spy get for you!
22:05
<@gnolam>
"Location of the secret moon base" is first on the list though.~
22:05
< Syka>
MOON NAZIS
22:05
<@Tarinaky>
("What is it that has put America at the forfront of the space race? Why, good old American know-how, as provided by good old Americans like Dr Werner von Braun.")
22:06
<&McMartin>
Ah, Project Paperclip
22:06
<@gnolam>
"Our Germans are better than their Germans"
22:07
< AnnoDomini>
Not to mention that USA was heavily German by the time the war started.
22:08
<&McMartin>
And by "the war" here, we mean the American Civil War
22:08
< AnnoDomini>
Any war involving the US!
22:08 * McMartin has heard that 19th-century German primary schools are one of the reasons for the old American stereotype of the Polish as fools
22:08
<&McMartin>
AnnoDomini: Mmm, not really war of 1812
22:09
<&McMartin>
IIRC the giant migration wave was mid-19th century
22:09
< AnnoDomini>
Yeah, yeah, I'm joking with that one.
22:09
< AnnoDomini>
AFAIK, USA was a little bit Dutch when they rebelled.
22:09
<&Derakon>
USA was a little bit everything.
22:09
< AnnoDomini>
Yes.
22:10
<&McMartin>
There was clearly enough Germanic influence that in throwing off the Foul English German made a showing in a race to become the national language.
22:10
<&McMartin>
(It didn't because, I mean, come on)
22:10
< AnnoDomini>
IIRC, that was whether to bother with German as a national language.
22:10
< AnnoDomini>
USA does not have any official national languages to this date.
22:11
<&McMartin>
And rightly so
22:11
<&McMartin>
(We'd need at least three)
22:11
< AnnoDomini>
And? There are places like that.
22:11
<&Derakon>
And one of them being Spanish would give about 25% of our population aneurisms.
22:11
<~Vornicus>
I'm in one of the largest hispanic communities in the US and the voting station signs are in english, spanish, and tagalog
22:12
< AnnoDomini>
Norway has two official languages, both Norwegian. (:P)
22:13
<@Tarinaky>
Language is politicised far too much.
22:13
< AnnoDomini>
Everything is politicized too much.
22:13
<@Namegduf>
Official languages are how you can tell if people are suffering from crippling insecurity.
22:14
<@Namegduf>
One of the ways, anyway.
22:14
<@Namegduf>
It is a pointless hat.
22:14
<&McMartin>
A lot of voting materials around here are in about 12 languages.
22:14
< Syka>
um
22:14
< Syka>
official languages are for legal canon
22:14
< Syka>
but whatever~
22:14
<@Namegduf>
Pretty sure they're not necessary for that.
22:14
< Syka>
they're not, but that's what they're for
22:15
<@Namegduf>
They really aren't "for" that in any particular sense.
22:15
<@Tarinaky>
Well, they're necessary if you want to codify a set of languages in which government services will be offered.
22:15
<&McMartin>
Er
22:15
<&McMartin>
Sure they are
22:15
<&McMartin>
For what Tarinaky said~
22:15
<@Namegduf>
They're "for" making the people crying for them who are, you will note, not lawyers, shut up
22:15
<@Tarinaky>
Which is fine and all except for the implication that if you don't speak any of those languages you're hosed.
22:15
< AnnoDomini>
You ARE hosed if you don't.
22:15
<&McMartin>
As opposed to the unofficial case where if you know some language you *might or might not* be hosed and can't tell until you check
22:16
<&McMartin>
(More formally: Having an official set of working languages is not a sign of xenophobia)
22:16
< AnnoDomini>
(Yes.)
22:16
<@Namegduf>
It doesn't *imply*.
22:16
< Syka>
from the .za constitution
22:16
< Syka>
"The national government and provincial governments may use any particular official languages for the purposes of government"
22:16
<@Namegduf>
It is a *sign*, however, because there's a correlation and that's how signs work
22:17
<&McMartin>
Syka: ... oof. OK, that's a new one on me.
22:17
< Syka>
Namegduf: so you're saying that .za is xenophobic because they have official languages?
22:17
<&McMartin>
I'm more used to "the government has to make *all* the official languages available"
22:17
< AnnoDomini>
Namegduf: You're suggesting that almost every civilized nation-state (excepting weird places like the US) is xenophobic.
22:17
< AnnoDomini>
Or aren't you?
22:17
< Syka>
even though the constitution specifies no less than 11
22:17
<@Namegduf>
http://thefreedictionary.com/sign
22:18
< Syka>
and that it is codified in the constitution that the official languages are there to promote the use of them
22:18
< Syka>
"Recognising the historically diminished use and status of the indigenous languages of our people, the state must take practical and positive measures to elevate the status and advance the use of these languages."
22:18
<&McMartin>
Syka: I was going tos ay "That appears to contradict your earlier quote" but I can't tell which part I misread
22:18
<&McMartin>
And my compile is done so I have to get back to work, so no more lightsaber chair fighting for me
22:18
< Syka>
here's the source: http://www.constitutionalcourt.org.za/site/constitution/english-web/ch1.html
22:19
< Syka>
first quote was 3(a), second was 2
22:19
<@Namegduf>
More formally, "is evidence of", as opposed to "is uncorrelated with" and "is evidence against".
22:19
<@Namegduf>
Which are the only three possibilities
22:19
<&McMartin>
Yeah, it's the "any particular" part that soudns like they get to blow off people who only speak *the other* official languages.
22:19
<&McMartin>
Which is deeply bizarre and doesn't sound like a correct reading
22:19
< Syka>
also, re: legal canon
22:20
<@Namegduf>
Yeah, "may use" doesn't seem useful here
22:20
< Syka>
it means that you can legally use those languages for dealing with the government
22:20
< Syka>
for example, in .za's case, they can use a native language for dealing with the govt
22:20
< Syka>
and the govt *have* to accept it
22:20
<&McMartin>
Yeah
22:20
<&McMartin>
Section 3a sounds like that is backwards
22:20
<&McMartin>
It should be "citizen can know any; government must know all"
22:20
<@Namegduf>
Yes.
22:20
<&McMartin>
Clearly from 2 this is the intent
22:20
< Syka>
let me quote it fully
22:20
< Syka>
"The national government and provincial governments may use any particular official languages for the purposes of government, taking into account usage, practicality, expense, regional circumstances and the balance of the needs and preferences of the population as a whole or in the province concerned; but the national government and each provincial government must use at least two official ...
22:21
< Syka>
... languages."
22:21
< Syka>
i should have put "..." at the end really
22:21
<&McMartin>
Ah, OK
22:21
<&McMartin>
Hrm
22:21
< Syka>
it's for their use
22:21
< Syka>
eg. they publish information
22:21
<&McMartin>
That is "how do you administrate a large area with no linguistic conformity"~
22:21
<@Namegduf>
Okay, so it does not in fact establish that you can talk to the government using any official language.
22:21
<@Namegduf>
Instead it establishes that if you know at least 11 you can do so.
22:22
<&McMartin>
If you know at least 10.
22:22
<@Namegduf>
Oh, yes, it specifies 11, not 12
22:22
< Syka>
wat?
22:22
<&McMartin>
However, I *strongly* suspect in practice that this is "Afrikaans works everywhere, and whatever the local language that is mutually unintelligible with the other 9 is"
22:22
<@Namegduf>
At least 10 is correct
22:23
< Syka>
I think it's a case of practicality
22:23
<@Namegduf>
Probably.
22:23
<@Namegduf>
My point is that here the official language guarantee there isn't actually buying you anything.
22:23
< Syka>
eg. a south govt doesnt need to speak a northern local language
22:23
<@Namegduf>
You're relying on non-guaranteed de-facto conditions anyway.
22:23
< Syka>
2(b): "Municipalities must take into account the language usage and preferences of their residents."
22:24
<&McMartin>
In practice, in the US, it's "English and Spanish works everywhere, and French, Filipino, and Cantonese will work depending on region"
22:24
<&McMartin>
(I assume official documents use Filipino rather than Tagalog specifically, but I do not actually know this)
22:24
<@Namegduf>
In the UK, English works and I have no idea what else would. Welsh should work in Wales.
22:24
<@Namegduf>
Courts hire translators.
22:25
<&McMartin>
Namegduf: Bear in mind that the US has more native Spanish speakers than Spain does.
22:25
<@Namegduf>
McMartin: You know as well as I do that that is a morally wrong sentence.
22:25
<@Namegduf>
McMartin: But yes the US does also have a reasonably high density of Spanish speakers.
22:25
<&McMartin>
... no, you're going to have to unpack that.
22:25
<@Namegduf>
McMartin: Density is what matters rather than count, and the US is *huge*.
22:26
<@Namegduf>
McMartin: *China* might have more <foo> speakers than wherever the language <foo> comes from.
22:26
< Syka>
US is 350M people, Spain is 50M people
22:26
<@Namegduf>
The US does however also have a fairly high density of Spanish speakers.
22:26
< Syka>
so 1/7th of the US needs to speak Spanish to have more than Spain does, assuming everyone in Spain speaks Spanish
22:27
<&McMartin>
And it's more like 1/5th.
22:27
<@Namegduf>
That's not that high, given it allowed for as a second language.
22:27 * AnnoDomini thinks that it's very practical and efficient to institute a single official language, teach it in public schools, and thereafter use it exclusively in governance. A lot of European states did that in the past centuries to get rid of the huge jumble of dialects their territories encompassed.
22:27
<@Namegduf>
Spanish is one of the big high school languages.
22:28
<@Namegduf>
It is, but in *practice* it tends to happen either:
22:28
<&McMartin>
AnnoDomini: This happened less in the New World, and I'm not sure why
22:28
<&McMartin>
And then Greek had it happen in reverse, which is even weirder.
22:28
<@Namegduf>
A) Because someone's pointless dialect is at risk of going extinct. See: Wales
22:28
<@Namegduf>
B) Because people are feeling the need to assert their nationalist pride.
22:28
<@Namegduf>
Politics does things for stupid reasons even if they're good things.
22:29
<&McMartin>
I would rephrase B as "because someone else's pointless dialect *isn't* at risk of going extinct"
22:29
<@Namegduf>
Haha.
22:29
< Syka>
man, I'd like to see them try and put an official language in au
22:29
<@Namegduf>
The calls in the US right now seem to mostly come from the "In America, we speak American!" crowd.
22:29
< Syka>
since there are literally 300 or so aboriginal languages
22:30
< Syka>
most of which mix, and most of which are completely useless
22:30
<@Namegduf>
Rather than from a reasonable assessment of the efficiency gains in establishing a single shared language.
22:30
<&McMartin>
Namegduf: Yes, because they've noticed that 20% thing and it scares the ever-living piss out of them.
22:30
< AnnoDomini>
If the locals want to have their own silly dialect, they can use their assigned part of the budget to teach it alongside the national language. It's sort of what Norway does.
22:30
<@Namegduf>
McMartin: Thus, "because they're feeling insecure"
22:30
< Syka>
and of which people get rather annoyed about the languages going extinct
22:30
<&McMartin>
Namegduf: That does appear relatively unique to the US, though.
22:30
< Syka>
it is interesting seeing someone speak it, and have to mix in English words for certain concepts
22:31
< Syka>
like, in the local one, there is apparently no word for "idea"
22:31
<@Reiv>
Syka: NZ has three national languages: English, Maori, and NZ Sign Language.
22:31
< Syka>
NZ Sign Language?
22:31
< Syka>
is that what you call rugby these days? :P
22:31
<@Namegduf>
McMartin: I haven't seen any other places with major campaigns to institute a national language that currently lack one.
22:31
<@Reiv>
It should be noted to Namegduf that it was neither going extinct nor a matter of nationalism. :p
22:32
< Syka>
also, semi related
22:32
< Syka>
Maori chants are scary as fuck
22:32
<&McMartin>
Reiv: Is NZSL related to British and/or American Sign Language or is it totally distinct?
22:32
<@Reiv>
However, NZ law enshrines that official languages may be used in official proceedings, interpreters must be provided, etc
22:32
<&McMartin>
(AIUI, ASL is closer to French than British signs)
22:32
< Syka>
seriously, the All Blacks doing it is terrifying
22:32
<@Reiv>
McMartin: I'm not sure, tbh
22:33 * Syka makes some joke about NZSL having an accent
22:33
< Syka>
(can sign language have an accent?)
22:34
<@Reiv>
"New Zealand Sign Language has its roots in British Sign Language (BSL), and may be technically considered a dialect of British, Australian and New Zealand Sign Language (BANZSL). There are 62.5% similarities found in British Sign Language and NZSL, compared with 33% of NZSL signs found in American Sign Language.[2]"
22:34
<@Reiv>
I guess it does have an accent after all!
22:34
< Syka>
haha
22:35
< Syka>
i am just imagining a NZSL speaker going to the US
22:35
< Syka>
and the americans going "what the fuck is this guy on about"
22:35
<&McMartin>
Yeah, ASL is Seriously Wildly Different
22:35
<&McMartin>
The grammar's all different too
22:36
<@Reiv>
It uses the same two-handed manual alphabet as BSL (British Sign Language) and Auslan (Australian Sign Language). It uses more lip-patterns in conjunction with hand and facial movement to cue signs than BSL, reflecting New Zealand's history of oralist education of deaf people. Its vocabulary includes Mori concepts such as marae and tangi, and signs for New Zealand placenames. (E.g. Rotorua - mudpools, Wellington - windy breeze, Auckland - Sky
22:36
< Syka>
similar to how Americans react if you give them anything said by an Australian/New Zealander
22:36
<@Reiv>
So we use more lip movements than the british stuff, so I guess that qualifies as 'accent'
22:36
<&McMartin>
Depends
22:36
< Syka>
"Why do you keep saying you can't?"
22:36 * Syka runs
22:36
<&McMartin>
Are we talking said by the Croc Hunter, or are we talking, like, Miranda Otto
22:36
< Syka>
(joke warning: may only be understood by Australians)
22:37
< Syka>
McMartin: miranda who
22:37
<&McMartin>
Australian actress. Played Eowyn in the LOTR movies.
22:37
< Syka>
oh
22:37
<@Reiv>
Syka: Yeah, even I don't get that one, and I'm picturing the aussie accent in the process
22:37
<&McMartin>
Has what I understand to be the "High Australian" accent that as an ignorant Yankee I cannot distinguish from Educated UK.
22:38
< Syka>
Reiv: say "caaaaarnt" in your head, but with a really strong Australian accent
22:38
< Syka>
(spoiler: the word begins with 'cu' :D"
22:38
< Syka>
)*
22:38
< Syka>
McMartin: ...High Australian?
22:39
< Syka>
that is something I have not heard of
22:39
< Syka>
she's from brisbane, huh
22:39
< Syka>
that's generally where you start getting Crocodile Hunter sounding from
22:39
< Syka>
eg. "Cairns" "Cayrrrns"
22:39
< Syka>
"Townsville" -> "Tawnsvil"
22:40
<&McMartin>
http://en.wikipedia.org/wiki/Variation_in_Australian_English
22:40
< Syka>
actually, it's less an accent, and more a "drawl"
22:40
<&McMartin>
Sorry, "cultivated Australian"
22:40
<&McMartin>
Strine is listed as "Broad"
22:41
<&McMartin>
... OK, I actually can't identify General Australian either as being particularly foreign sounding, if that's what Eric Bana and Hugh Jackman are speaking.
22:41
<@Namegduf>
Now I'm imagining high elven as being spoken like that
22:41
<@Reiv>
Syka: ... ah, I was totally misreading it then
22:41
<@Namegduf>
Just sort of a drawl version of a normal accent
22:42
< Syka>
http://en.wikipedia.org/wiki/Western_Australian_English haha what
22:43
<&McMartin>
Careful there :D
22:43
<&McMartin>
Last time I did that with California English ("Nobody does this") I got burned by doing those things while complaining about it
22:44
< Syka>
I have a special english
22:44
<&McMartin>
(In particular, "kin" and "king" getting different vowel sounds)
22:44
< Syka>
apparently I speak more eastern au english
22:44
< Syka>
since I say fear without moving my jaw
22:44
< Syka>
wait
22:44
< Syka>
no
22:44
< Syka>
nevermind, I do say it fia
23:05
<&ToxicFrog>
McMartin: this reminds me, symbol and I determined experimentally the other night that I use the same phoneme I use for 'python' (or a related phoneme with a different consonant) in some other words.
23:05
<&ToxicFrog>
I can't remember which now, though.
23:05 AnnoDomini [abudhabi@Nightstar-d00ld9.gw-out.freebnc.net] has quit [[NS] Quit: ZNC - http://znc.in]
23:09 himi [fow035@Nightstar-q9amk4.ffp.csiro.au] has joined #code
23:09 mode/#code [+o himi] by ChanServ
23:09 abudhabi- [Elite8361@Nightstar-rl3j3i.tl.libirc.so] has joined #code
23:11
<~Vornicus>
Myth, heath?
23:11 VirusJTG [VirusJTG@Nightstar-6i5vf7.sta.comporium.net] has joined #code
23:12
<&ToxicFrog>
Vornicus: the difference is in the py, not the th.
23:13
<&Derakon>
Uh, like "pie"?
23:15 abudhabi- is now known as AnnoDomini
23:15
<~Vornicus>
TF: both of those have the same th sound, and different vowel sounds...
23:16
<&ToxicFrog>
Derakon: no. Py has three phnoemes for me; "pie" (pyromania), "pee" (copy), and "py" (python)
23:17
<&ToxicFrog>
My partners inform me that the latter of those does not actually exist in english.
23:17
<&Derakon>
Yeah, I pronounce "python" like "pie thon".
23:17
<&ToxicFrog>
(this came up a while ago elsechannel when I commented to McM that it's interesting that pypy (pie-pie) and python (py-thon) have different phonemes. The response was "er what")
23:17
<&Derakon>
As in, the pie of the gender-nonspecific third-person entity over there.
23:18
< Syka>
python is pie-fon here
23:18
<&ToxicFrog>
(I assumed it was regional and commented on it to symbol when I got home)
23:18
< Syka>
and pypy AND pypi are both prounounced pie-pie for me
23:18
< Syka>
even though pypi is pie-pee or pippy
23:19
<~Vornicus>
PIPE
23:19
<~Vornicus>
19 o'clock and all is pipe.
23:19
<~Vornicus>
( http://www.youtube.com/watch?v=_CPiR6d0gm8 )
23:21
< Shiz>
I pronounce PyPI as Pie-Pee-I
23:22 Orthia [orthianz@Nightstar-avg.1ee.224.119.IP] has joined #code
23:22 mode/#code [+o Orthia] by ChanServ
23:23
<&ToxicFrog>
"pie-pee", I think. Or "pipey"
23:23
<&ToxicFrog>
not the same phoneme as 'python'
23:23
<&ToxicFrog>
I wish I could remember what the other word is now.
23:32
<@Reiv>
Toxicfrog: "Paithon"?
23:33
<@Reiv>
Here Python would have the same phenome as pyromania, ableit with perhaps a slightly harder stress on the vowel.
23:40 Derakon [chriswei@Nightstar-4k2ccr.ca.comcast.net] has quit [[NS] Quit: leaving]
23:55 Turaiel[Offline] is now known as Turaiel
--- Log closed Thu Dec 19 00:00:08 2013
code logs -> 2013 -> Wed, 18 Dec 2013< code.20131217.log - code.20131219.log >

[ Latest log file ]