I2P dev meeting, February 18, 2003 @ 23:01 UTC

(Courtesy of the wayback machine http://www.archive.org/)

Quick recap

  • Present:

aum, Barney, bpb, hezekiah, jeremiah, LeerokOnKnoppix, lonelynerd, mids, namless, nop, PsionX, _Trent, WindowsHater,

Volledige IRC Log

--- Log opened Tue Feb 18 23:54:43 2003
23:54 -!- Topic for #iip-dev: IIP Meeting - logfiles: http://mids.student.utwente.nl/~mids/iip/
23:54 [Users #iip-dev]
23:54 [ aum   ] [ bpb      ] [ jeremiah       ] [ logger] 
23:54 [ Barney] [ codeshark] [ LeerokOnKnoppix] [ mids  ] 
23:54 -!- Irssi: #iip-dev: Total of 8 nicks [0 ops, 0 halfops, 0 voices, 8 normal]
23:54 -!- Irssi: Join to #iip-dev was synced in 2 secs
23:55  * bpb points at the logging bot then jumps up and down like a japanese animated school girl
23:55 -!- mode/#iip-dev [+o mids] by Trent
23:55 -!- mode/#iip-dev [+v logger] by mids
23:57 -!- LeerokOnKnoppix is now known as Trent
23:57 -!- Trent is now known as nickthief76011
23:57 <@mids> :p
23:57 -!- nickthief76011 is now known as LeerokOnKnoppix
23:57 < lonelynerd> wwwhat?
23:58 <@mids> Trent protects the nick
23:58 < lonelynerd> oh, nothing
23:58 < LeerokOnKnoppix> How does it do that?
23:58 -!- LeerokOnKnoppix is now known as _Trent
23:58 < _Trent> I am Trent, hear me roar!
23:58 < hezekiah> LOL
23:58 < _Trent> Get out, all unverified users!
23:59 -!- mode/#iip-dev [+o hezekiah] by Trent
23:59 < _Trent> What about me?
23:59 <@mids> hey hezekiah :)
23:59 -!- _Trent is now known as LeerokOnKnoppix
23:59 <@hezekiah> Hi, mids. :)
23:59 <@hezekiah> I think I'm getting the hang of some of this IRC thing! :)
23:59 <@mids> kool
--- Day changed Wed Feb 19 2003
00:00 <@hezekiah> So have any bugs defied the great IIP dev team and risen their ugly heads since your last mail to the iip-dev mailinglist?
00:01 < LeerokOnKnoppix> It disconnects every now and then.
00:01 <@hezekiah> I don't seem to experiance that bug. :(
00:01 < lonelynerd> LeerokOnKnoppix, it's probably your connection. i haven't had any problems
00:01 < LeerokOnKnoppix> Hmm.
00:01 <@hezekiah> mids: Where are nop and UserX?
00:01 <@mids> Tue Feb 18 23:01:51 UTC 2003
00:01 <@mids> user was here but he pinged out
00:01 < LeerokOnKnoppix> _Trent kicked them out.
00:02 <@mids> userx
00:02 <@mids> just icq-ed nop
00:02 < LeerokOnKnoppix> Is he coming?
00:02 <@mids> dunno
00:02 < bpb> lonelynerd: it's always happened though...  and it's not just him
00:03 < lonelynerd> bpb, still, it's not a problem with iip but unreliable connections
00:04 <@mids> nop is comming
00:04 < nop> hi
00:04 < lonelynerd> hi
00:04 < nop> ok
00:04 < nop> welcome
00:04 < LeerokOnKnoppix> Hello nop.
00:04 <@hezekiah> Hi, nop! :)
00:04 < nop> Leerok likes Knoppix?
00:04 < LeerokOnKnoppix> So far.
00:04 < nop> kewl
00:04 < nop> ok
00:04 < bpb> lonelynerd: it's a problem with the actual network then
00:04 < LeerokOnKnoppix> But I want to install it to the hard drive.
00:04 < LeerokOnKnoppix> But it didn't work when I tried it.
00:04 < nop> well, welcome to the 31st meeting
00:04 < nop> of IIP-dev
00:04 < LeerokOnKnoppix> Cool.
00:04 < lonelynerd> bpb, well, it could handle unrealiable links, but it doesn't at the moment
00:04 < LeerokOnKnoppix> I'm just hanging around to let knowledge osmose into my head.
00:05 < nop> ok
00:05 < nop> shh
00:05 < nop> meeting started
00:05 < nop> on the agenda
00:05 -!- mode/#iip-dev [+o nop] by mids
00:05 <@nop> IIP 1.1 and pleasantries with that
00:05 <@nop> hezekiah: 
00:06 <@hezekiah> Yeah?
00:06 <@nop> hezekiah: all entropy stuff have we resolved? For instance the hhhhhh issue
00:06 <@hezekiah> I just told you about the hhhhhhhh isue.
00:06 <@hezekiah> We haven't done anything about it yet.
00:06 <@nop> ok
00:06 <@hezekiah> The seed.rnd issue is resolved by porting the --randomdialog option from development ..
00:06 <@nop> ok
00:07 <@nop> and can you explain what they do
00:07 <@nop> just for the record
00:07 <@hezekiah> ... and having IIP abort when seed.rnd doesn't exist.
00:07 <@hezekiah> Explain what --randomdialog and --ignoreemptyrandom do?
00:07 <@nop> yes
00:07 <@hezekiah> OK. :)
00:07 <@nop> for official purposes
00:07 <@nop> aka the meeting
00:07 <@hezekiah> Right. :)
00:08 <@hezekiah> When isproxy is run with --randomdialog, the user is asked for entropy (even is seed.rnd already exists).
00:09 <@hezekiah> When isproxy is run with --ignoreemptyrandom, and there isn't enough entropy (a condition that would cause isproxy to _normally_ abort), the program just keeps going. This is a security risk, so a suitablely scary warning label is added in the help description.
00:09 <@hezekiah> (Speaking of all these options, unless mids as already added them, they aren't in the isproxy.pod file yet.)
00:09 <@hezekiah> .
00:09 <@mids> I didnt
00:09 <@hezekiah> We should do that before release. :)
00:10 < lonelynerd> (shouldn't it be --random-dialog and --ignore-empty-random ? ;P)
00:10 <@hezekiah> lonelynerd: I code what UserX orders. :)
00:10 < lonelynerd> ok
00:10 <@hezekiah> nop: So, the seed.rnd problem is solved, but the hhhhhhhh problem isn't.
00:10 <@nop> right
00:11 <@nop> hmm, well, we could have a counter that won't allow the same keystroke more than three times in a row
00:11 <@hezekiah> (I don't even know if hhhhhhhh is a problem. I just mentioned it to you and UserX. You tell me!) :)
00:11 <@nop> real simple
00:11 <@nop> it's a slight problem
00:11 <@hezekiah> PGP some how knows how much entropy text is worth.
00:11 <@nop> well, we calculate the keyboard timings as well
00:12 <@nop> we could just force a wait till something beyond 3 same char in a row are pressed
00:12 <@hezekiah> Does GPG evaluate the entropy of a string of text? If they do, we could look at their code. :)
00:12 <@nop> I am not sure
00:12 <@nop> I advise looking at it
00:12 < lonelynerd> hezekiah, i think it just uses /dev/random, or?
00:12 <@nop> lonelynerd no
00:12 <@nop> it doesn't
00:13 < lonelynerd> ah
00:13 <@nop> we're talking about creation of entropy
00:13 <@nop> aside /dev/random
00:13 <@nop> like an Initialization
00:13 <@hezekiah> GPG does make the user pound on the keyboard. :)
00:13 <@mids> well no
00:13 <@mids> it reads from the entropy pool
00:13 <@mids> and suggests that you move the mouse + press keys
00:14 <@mids> but entropy pool can be filled with interrupts etc too
00:14 <@nop> yes
00:14 < lonelynerd> yep
00:14 <@hezekiah> Ah. I just remembered when GPG made me pound on the keyboard once. ;)
00:14 < lonelynerd> it was pgp perhaps?
00:14 <@hezekiah> Nope. It was GPG running under Mandrake Linux.
00:14 < Barney> yep
00:15 <@hezekiah> (It was a few years ago, before AES was released I remember.) Anyway, back on topic. :)
00:15 <@nop> I think we should look at the code
00:15 <@hezekiah> OK. :)
00:15 < lonelynerd> :)
00:15 <@nop> if it's not clear, I suggest limiting the characters in a row
00:15 <@nop> and waiting
00:15 <@nop> for new characters to be pressed
00:15 <@hezekiah> Are there any methods outline (maybe in whitepapers from acedemia) on evalutating the entropy worth of a string of text?
00:16 <@nop> yes there are, I will dig them up in a bit
00:16 <@hezekiah> Cool! :)
00:16 < lonelynerd> "ent - A pseudorandom number sequence test program"
00:16 <@nop> yes, chi squaring 
00:16 <@hezekiah> We could just implement one of those. (Believe it or not, it might be faster than digging through GPG code!) :)
00:16 <@nop> http://www.gnu.org/software/gnu-crypto/api/gnu/crypto/tool/Ent.html
00:17 <@nop> that's java
00:17 <@nop> but still
00:17 <@hezekiah> I can read Java. :)
00:17 <@hezekiah> I can translate too. :)
00:17 < lonelynerd> ent is in C, too
00:17 <@nop> yep
00:17 <@hezekiah> lonelynerd: Where can I find the C version?
00:18 < lonelynerd> apt-get inst.. whoops, ;P http://www.fourmilab.ch/random
00:18 < lonelynerd> Upstream Author(s): John Walker
00:21 <@nop> ok
00:21 <@nop> anyway
00:21 <@nop> I'm sure that needs to be taken care of b4 1.1
00:21 <@nop> this will give mids time to update isproxy.pod
00:21 <@hezekiah> lol
00:21 <@mids> pff
00:22 <@mids> say nop, did you work on the docs? :)
00:22 <@nop> ;)
00:22 <@nop> I am working on them
00:22 <@nop> and I have an editor too ;)
00:22 <@mids> well
00:22 <@mids> dont edit the html
00:22 <@nop> I know
00:22 <@nop> I'm just writing it in notepad
00:22 <@nop> ;)
00:22 <@mids> just send me the text changes, yeah thats fine
00:22 <@hezekiah> Yay! Real HTML coder! Uses a plain text editor! ;-)
00:23 <@nop> haha
00:23 <@nop> I ain't coding it
00:23 <@nop> just typing the doc
00:23 <@nop> ;)
00:23 <@mids> so
00:24 <@mids> will any more features slip in for 1.1?
00:24  * mids hopes not
00:24 <@mids> I tried to contact codeshark
00:24 <@nop> no
00:24 <@mids> but didnt get a reply yet
00:24 <@nop> it's not a feature
00:24 <@nop> it's a bug fix
00:24 <@nop> ;)
00:25 <@nop> anyway
00:25 <@nop> moving forward
00:25 <@mids> ok, I got other things to do
00:25 <@mids> if you need me shout loud
00:25 < lonelynerd> (gpg seems to use get_entropy_count(fd) to check how many bytes /dev/random has to offer)
00:26 <@nop> right
00:26 <@nop> aka checking the entropy pool size
00:26 < lonelynerd> which is some ioctl
00:26 < lonelynerd> dunno how that would work under windows
00:26 <@hezekiah> lonelynerd: It probably _doesn't_ work under windows. This a GNU project, remember? :)
00:27 < lonelynerd> great :)
00:27 <@hezekiah> What we really want (at least I think we do) is something that can evaluate how much entropy is in a buffer of text.
00:27 < lonelynerd> but gpg supports also other means to get entropy
00:27 <@nop> yes
00:27 <@hezekiah> That way we can have the user keep pounding until we get that much _REAL_ entropy. :)
00:27 < lonelynerd> hezekiah, ok
00:27 <@nop> yes
00:27 <@nop> we should have 1 bits per 3 characters
00:28 <@nop> technically that's considered secure entropy
00:28 < lonelynerd> ok, but under linux it could just read /dev/random
00:28 < lonelynerd> let the windows users type ;)
00:28 <@hezekiah> Well, isn't "randomentropyneeded" the number of bytes of entropy still needed?
00:28 <@nop> yes
00:29 < LeerokOnKnoppix> MD5sum of the registry.
00:29 <@hezekiah> lonelynerd: UserX, nop, and I have talked about ways to get IIP to use /dev/urandom under controlled circumstances.
00:29 < lonelynerd> hezekiah, hmm
00:29 < lonelynerd> hezekiah, urandom is just like random, but it doesn't block when there isn't enough entropy?
00:31 <@hezekiah> I think so. Is that right, nop?
00:31 <@nop> yes, with urandom you can assign the amount of entropy in advanced
00:31 < lonelynerd> so urandom might or might not be secure
00:32 <@nop> urandom just uses random
00:32 < lonelynerd> yep
00:33 <@hezekiah> lonelynerd: That's the issue. :) We considering compiling a list of OS's (and their versions) with secure instances of /dev/urandom, and having IIP use it when it's secure; when it's not secure, IIP would stick with the current methods. It's a thought. :)
00:33 <@nop> right, but not for 1.1
00:33 <@nop> ;)
00:33 <@hezekiah> Of course!
00:33 < lonelynerd> hezekiah, wouldn't that be "secure versions of random" and not urandom?
00:33 <@hezekiah> This is all 1.2 stuff I'm talking about. (I am usually always talking 1.2 unless otherwise stated!)
00:33 < lonelynerd> ok
00:33 < lonelynerd> sounds great
00:34 <@nop> lonelynerd nothing wrong with urandom if random is considered secure
00:34 < lonelynerd> well if you check that there is enough entropy
00:36 <@hezekiah> So, nop? How exactly do we handle this hhhhhhhh problem?
00:36 <@nop> well
00:36 <@nop> for 1.1.0 I would limit the amount of repeated characters
00:36 <@nop> in a row that is
00:36 <@nop> and then leave it for in depth discussion in 1.2
00:37 <@hezekiah> OK. But what happens when someone does pattern like "hHhHhHhHhHhHhH"?
00:37 < lonelynerd> :)
00:37 <@nop> we need to create a counter that spaces it out
00:37 <@hezekiah> (Let's assume the user is an idiot. The software is only as secure as the user!) :)
00:37 <@nop> so so many bytes can not contain this letter
00:37 <@nop> since last used
00:37 < jeremiah> hello
00:37 <@nop> hi
00:37 <@hezekiah> jeremiah: hi. :)
00:37 < jeremiah> hey hezekiah
00:38 <@nop> for instance [h][H][xonH[xonH[ etc
00:38 <@hezekiah> So, 'h' can't be repeated more than _m_ times in _x_ bytes?
00:38 <@nop> yes
00:38 < jeremiah> are we talking about buffer overflows?
00:38 <@nop> it will be considered rejected entropy
00:38 <@hezekiah> jeremiah: Nope! We're talking about entropy! :)
00:38 <@nop> but still log the keyboard timings
00:38 <@nop> just ignore it as a string
00:39 < jeremiah> random number generators are supposed to make sure there isn't a correlation with the numbers anyways
00:39 < jeremiah> you can do that with chi-square
00:39 < jeremiah> right?
00:40 <@nop> well yarrow helps mix
00:41 <@hezekiah> You said we "log the keyboard timings". What's that mean?
00:42 < jeremiah> hezekiah: times between when you hit keys
00:42 < lonelynerd> that's difficult if input is buffered
00:42 <@nop> just do the character thing, it will be fine
00:42 <@nop> yarrow does the rest
00:42 <@hezekiah> nop: Are we getting the entropy from the keys being hit or the timing between when the keys get hit?
00:42 <@nop> both
00:42 <@hezekiah> Oh. OK. :)
00:42 <@hezekiah> I'll just write up some code that prohibits 'char' repeating more than 'm' times in 'x' bytes. :)
00:43 <@nop> yes
00:43 <@nop> thnx
00:43 <@hezekiah> Then for development, we brainstorm. ;-)
00:43 <@nop> yes
00:44 < LeerokOnKnoppix> How 'bout you get entropy from all possible inputs like the microphone, internet, processor speed, temperature, memory usage and speed, and other things?
00:44 < lonelynerd> LeerokOnKnoppix, calm down :P
00:44 <@hezekiah> I refuse to collect entropy from my internet connection; if someone can sniff it, then the quality of my entropy is lowwered.
00:45 <@nop> haha
00:45 < lonelynerd> yeah, and the kernel already does a nice job in gathering entropy. applications shouldn't do it
00:46 <@hezekiah> lonelynerd: That works fine as long as IIP is run on an OS with good kernel random number support. There are OS's that exist that _don't_ have good support!
00:46 < PsionX> like windows
00:46 <@hezekiah> lol
00:46 < lonelynerd> hezekiah, yep
00:46 -!- PsionX is now known as WindowsHater
00:46 < WindowsHater> lol
00:46 < lonelynerd> hezekiah, just use some #ifdef there
00:47 -!- WindowsHater is now known as PsionX
00:47 <@nop> this was discussed earlier
00:47 <@nop> and we will look at it on 1.2
00:47 <@nop> moving on
00:47 < lonelynerd> ok ok
00:47 <@nop> we'll end up repeating ourselves
00:47 <@nop> a dozen times
00:47 <@nop> before this meeting is over
00:47 <@hezekiah> Yeah. :)
00:47 <@hezekiah> Next item! :)
00:49 <@hezekiah> Uh, nop? Do we have a next item?
00:49 <@hezekiah> Perhaps: "What's on the list until we can release IIP 1.1"
00:50 <@hezekiah> Currently, I have:
00:50 <@hezekiah> 1.) mids adds docs for --randomdialog and --ignoreemptyrandom to isproxy.pod
00:50 <@hezekiah> 2.) nop finishes working on his HTML docs
00:50 <@nop> yeah
00:50 <@hezekiah> 3.) We correct the 'hhhhhhhh' entropy problem by prohibiting a certain number of repeated characters.
00:51 <@hezekiah> Anything else?
00:51 <@nop> right
00:51 <@nop> nope
00:51 <@nop> not that i know of
00:51 <@hezekiah> Well, hopefully all that will be done by next week! :)
00:51 <@hezekiah> (Or earlier!)
00:51 <@nop> yes
00:54 <@hezekiah> So what's next on the agenda?
00:55 <@nop> umm
00:55 <@nop> questions,
00:55 <@hezekiah> lol!
00:55 < LeerokOnKnoppix> Not much, it appears.
00:55 < LeerokOnKnoppix> Will there be socks?
00:55  * PsionX looks at his feet
00:55 < LeerokOnKnoppix> No, I mean the proxy thing.
00:55 <@nop> not in 1.1.0
00:56 < LeerokOnKnoppix> Hmm.
00:57 < LeerokOnKnoppix> Will there be IP spoofing?
00:57 < lonelynerd> w-what?
00:57 <@nop> umm, not in this lifetime
00:57 < LeerokOnKnoppix> Never mind.
00:57 < LeerokOnKnoppix> I'm just thinking up random questions.
00:58 < lonelynerd> :)
00:58 < LeerokOnKnoppix> Will there be any sort of file-transer?
00:58 < LeerokOnKnoppix> *transfer
00:58 <@nop> not in 1.1
00:58 < LeerokOnKnoppix> Will there be buddy icons?
00:58 <@hezekiah> LOL
00:59 < PsionX> will someone please shut him up?
00:59 < lonelynerd> LeerokOnKnoppix, you can already transfer files over iip but it's a bit slow
00:59 < LeerokOnKnoppix> I know.
01:01 < LeerokOnKnoppix> Will there be Reversi?
01:01 < PsionX> this guy wants everything and a bag of chips too
01:02 < LeerokOnKnoppix> Indeed.
01:02 <@nop> the best question
01:02 <@nop> will the human race be freed 
01:02 <@nop> from oppression
01:02 < LeerokOnKnoppix> With IIP? Certainly!
01:02 <@hezekiah> lol!
01:02 < PsionX> i got a question
01:02 < PsionX> ... can i kick hezy in the butt for no good reson?
01:02 <@nop> sure
01:02 < PsionX> lol
01:03 <@hezekiah> nop: So my work for now is to write up the repeating-character-entropy fix.
01:03 <@nop> yes
01:03 <@nop> oh and change the world in one sitting
01:04 <@hezekiah> lol
01:04  * aum pulls his face out of the other windows
01:04 < LeerokOnKnoppix> Peace in Iraq.
01:04 < aum> hi all
01:04 < LeerokOnKnoppix> Hello.
01:04 <@hezekiah> Oh, boy! I spent TOO LONG going over THAT one!
01:04 <@nop> hi aum
01:04 < PsionX> ...
01:04 < PsionX> ... i got spikeys
01:05 <@hezekiah> Question!
01:05 < PsionX> Answer
01:06 <@hezekiah> After IIP 1.1 is released, are we going to sit down and draw up a nice linear list of what needs to be done to decentralize the ircd server? (For example: does the "core" code need to be completed first?)
01:09 <@nop> yes we will do that
01:09 <@mids> s/1.1/1.1.0/
01:09 <@hezekiah> lol. OK, mids. :)
01:10 <@hezekiah> mids: Though it should probably be s/1\.1/1.1.0/ just for clarity's sake. ;-)
01:10 <@mids> oops, sorry
01:10 <@hezekiah> lol. :)
01:11 <@hezekiah> Also:
01:11 <@mids> maybe before decentralizing there should be another step
01:11 <@hezekiah> I've seen that there is the capability to make varios crypto cores.
01:11 <@mids> keeping a central irc, but letting the proxy be IRC aware
01:11 <@mids> so it can already handle private conversations
01:11 <@mids> etc
01:14 <@hezekiah> mids: i.e. it interprets /msg and does direct user-to-user encryption?
01:14 <@mids> yes
01:14 <@hezekiah> Well, I don't know how abstract things have to say, but I think it's a good idea until we can get the server decentralized! :)
01:16 <@hezekiah> About the crypto cores: if IIP has several crypto cores (SOCKS, SSL, and custom for example), then what happens when you have a connection between a user and several relays that has different encryption between each link?
01:16 <@mids> you can do what GPG does
01:16 <@mids> require a minimum
01:16 <@mids> and if both parties have better, use that
01:16 <@mids> also look hope SILC does it
01:17 <@mids> bye hezekiah
01:17 < PsionX> LOL dial up
01:17 < PsionX> lol on hezy
01:17 <@mids> ;)
01:17 <@mids> wb
01:17 < hezekiah> Let's say that SOCKS is weaker than SSL (I have no idea if it is.)
01:18 < namless> what about own privat/publick keys that the user can add to (client or to IIP proxy) so he can crypt what he talks to chertain chanels or private querys? (meaning users who downt have the right key on these special chanels or private queries down  understant other users and IIP proxy filters this crypted text away)
01:18 < hezekiah> (Sorry about being nocked off the internet folks; my brother probably tried to get on.)
01:18 < namless> down -> dont (dam bad english and typos :(
01:19 <@mids> namless: like the existing blowfish etc encryption for irc clients, but then server side
01:19 <@mids> (psybnc has it too)
01:19 < namless> jep
01:20 < namless> the system vould be build into the IIP proxy so it could be transparent to the client...
01:20 <@mids> another thing that I'd like to see is support for multiple layers on the same 'mixnet'
01:20 < namless> vould-> could
01:20 <@mids> so you can do multiple protocols with the same relays
01:22 < hezekiah> (OK. This is getting annoying.)
01:22 <@mids> probably it is bedtime for you :)
01:23 < hezekiah> Nah. That's not for hours.
01:23 <@mids> so no divine intervention?
01:23 < hezekiah> lol
01:23 < hezekiah> So, mids. From what the logs said (Thank God for logs), you mentioned that we could do as GPG does: require a minimum and use better if we have it.
01:24 <@mids> yeah
01:24 <@mids> probably different systems can be compatible too
01:24 <@mids> like homebrew DH and SSL one
01:24 <@mids> with SSL just beeing faster
01:24 < hezekiah> But all the advantage of one connection type being securer than the other are lost when someone along the long uses something less secure.
01:24 -!- LeerokOnKnoppix is now known as LeerokShovelSno
01:26 < hezekiah> Oops. s/along the long/along the line/
01:26 < hezekiah> Well, we can worry about that another day.
01:26 <@mids> k
01:26 <@mids> btw, any idea how to do something like <br> in POD format?
01:26 < hezekiah> nop: Sometime we should also work on implementing RKA. :)
01:27 < hezekiah> mids: You know more about POD than I do because I know nothing about POD! :)
01:29 <@nop> well we have RKA working but not as good as we'd like
01:29 <@nop> the GMP stuff will change that
01:29 < hezekiah> mids: "E<escape>" -- a character escape
01:29 < hezekiah> nop: We do? When did that happen?
01:29 <@nop> it's been in
01:30 <@nop> every 52 blocks it changes keys
01:30 < hezekiah> Ah
01:30 < hezekiah> Do you think we might make the key changing a little more random?
01:30 < namless> "everyting should be random"...
01:31 < hezekiah> Like having a random number checked every 10 blocks and a 1 out of 3 chance of the key changing?
01:31 < hezekiah> ... or having an SHA sum of the last 3 plaintext messages checked and a 1 out of 3 chance of the key changing based on that?
01:35 < lonelynerd> hmm
01:35 <@mids> ok, info about --randomdialog and --ignoreemptyrandom added to POD
01:35 < namless> by the way how "similar" is the traffic IIP is making? is there any plans to make this traffic to look as much random it can be?    if the "transfer macanism" makes a static mark in the traffic then it can be detected as IIP traffic and can be started to be monitored and well there only the sky is the limit...
01:35 < hezekiah> And committed to HEAD and development?
01:36 <@mids> oops
01:36  * mids does to dev too
01:37 < hezekiah> nop? Are you there?
01:37 <@nop> sorry
01:37 <@nop> what
01:37 <@nop> dude, my brain is gone
01:37 <@nop> can we discuss all this later
01:37 < hezekiah> OK. :)
01:37 < hezekiah> Sorry. :)
01:38 <@mids> ok, lets close
01:38 < hezekiah> (If it makes you feel any better, my brain is going to be gone too after a Physics review and some calcus!) :)
01:38 <@mids> commited to development too
01:38  * hezekiah hands mids the "baff"er
01:38 < hezekiah> OK. Good! :0
01:38 <@mids> *baff*
01:38 < hezekiah> s/0$/)/
01:38 < hezekiah> Bye all! :)
01:38 <@mids> cya next week
01:40 < lonelynerd> bye
--- Log closed Wed Feb 19 01:40:48 2003