I2P dev meeting, October 16, 2002

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

Quick recap

  • Present:

al-jabr, Chocolate, firegod, geully, Grishnav, hobbs, mason, mids, nemesis, nop, philocs, Robert, UserX, youkai,

Volledige IRC Log

--- Log opened Tue Oct 15 23:31:29 2002
23:31 < logger> test
23:32 < mason> sorry, that test did not work
23:32 < mason> :)
23:32 -!- mode/#iip-dev [+o mids] by Trent
23:32 <@mids> Tue Oct 15 21:32:19 UTC 2002
23:32 <@mids> meeting starts in 1:30 hours
--- Day changed Wed Oct 16 2002
00:44 < geully> hi all
00:50 <@mids> Public IIP meeting in 10 minutes here
00:50 < Robert> Hello all.
00:51 <@mids> shhh
00:51 <@mids> not yet
00:51 <@mids> 9 more minutes
00:51 < Grishnav> lol
00:51 < al-jabr> Tue Oct 15 22:51:23 UTC 2002
00:51  * Robert zips his lip.
00:51 < al-jabr> lalala
00:53 -!- geully is now known as Geully
01:00 <@mids> Tue Oct 15 23:00:02 UTC 2002
01:00 <@mids> welcome to the n-th public IIP meeting
01:00 <@mids> logfiles are on http://mids.student.utwente.nl/~mids/iip/
01:00 < nop> hehe
01:00 <@mids> oh, 15th
01:00 < nop> 15th
01:00 < nop> yes
01:00 <@mids> agenda for today:
01:01 <@mids> - new IIP developer
01:01 <@mids> - IIP logo contest
01:01 <@mids> - bug fixes
01:01 <@mids> - question rounds
01:01 <@mids> ,
01:01 <@mids> .
01:01 < nop> ok
01:02 < nop> welcome back all
01:02 < nop> to another round of meetings ;)
01:02 < nop> for all that work in a corporate office
01:02 < nop> you have permission to sleep
01:02 < nop> ok
01:02 < nop> new IIP developer
01:02 -!- mode/#iip-dev [+o nop] by mids
01:02 <@nop> and is a talented and quick learning C programmer
01:02 -!- mode/#iip-dev [+o UserX] by mids
01:02 <@nop> and has already added some patches and some grunt work that was needed to the code
01:03 <@mids> hurray!
01:03 <@nop> we are glad to have him
01:03 <@nop> and we feel that he will be an essential part of the team
01:03  * al-jabr claps
01:03 <@nop> ok
01:03 <@nop> next on list
01:03 <@nop> IIP logo contest
01:03 <@nop> any graphix designers etc
01:03 <@mids> graphix? you mean graphics?
01:04 < Grishnav> No, he means graphix :P
01:04 < hobbs> nop: maybe. Me or my mom. She's good, and she got a tablet recently. :)
01:04 <@nop> who would like to come up with a cool slogan and/or logo for invisiblenet, and IIP (yes I mean graphics) for t-shirts can submit their entries to iip@invisiblenet.net
01:04 <@nop> the winner
01:04 <@nop> will win a free t-shirt
01:04 <@nop> black or white
01:04 <@nop> of his choice
01:04 <@nop> or her choice
01:04 <@mids> woohoo!
01:04 <@nop> and 10.00 DRAN
01:04 < hobbs> nice.
01:05 <@nop> this can definitely include slogans as well
01:05 <@nop> so there could be two winners
01:05 <@nop> if one comes up with logo
01:05 -!- mode/#iip-dev [+o Chocolate] by mids
01:05 <@nop> and one comes up with an awesome slogan
01:05 -!- mode/#iip-dev [+o Chocolate] by Trent
01:05 <@nop> but submit to iip@invisiblenet.net
01:05 <@nop> and they will be reviewed
01:05 <@nop> I hope that if you're not a graphics guy, that you can tell a friend
01:05 <@nop> and maybe split the profits
01:05 <@nop> ;)
01:06 <@nop> because we would like to have cool shirts 
01:06 <@nop> for the e-store
01:06 <@nop> and in general
01:06 <@nop> as well
01:06 <@nop> for bumper stickers etc
01:06 <@nop> maybe a mascot would be good too
01:06 <@mids> :)
01:06 <@nop> either way
01:06 <@nop> do what you can
01:06 <@nop> submit them
01:06 <@nop> and we'll decide at the end of the month
01:06 < philocs> our only mascot is satan
01:06 <@nop> well
01:06 <@nop> that's taken
01:06 <@nop> BSD
01:06 <@nop> ;)
01:07 < philocs> we could make a scarier satan
01:07 < hobbs> that's a DAEMON!
01:07 <@nop> hehe
01:07 <@Chocolate> black
01:07 <@nop> ok
01:07 <@nop> next
01:07 <@nop> bugfixes
01:07 <@Chocolate> tshirt must be black
01:07 <@nop> ok
01:07 <@nop> yeah
01:07 <@nop> all artist must make inversed drawings
01:07 <@nop> so that it caters to black or white backgrounds
01:07 <@nop> and you can use color :)
01:07 <@nop> oh
01:07 <@nop> and the winner
01:08 < nemesis> http://www.stk.com/products/50_beta/about50.cfm
01:08 <@nop> will of course get full credit by having his logo on the t-shirt
01:08 < nemesis> nice
01:08  * al-jabr fears that this T-shirt may be hopelessly dorky
01:08 <@mids> al-jabr: make a better one
01:08 <@mids> okay...
01:08 <@mids> next poing?
01:09 <@mids> point :)
01:09 <@mids> beeing: bugfixes
01:09 <@mids> UserX fixed even more bugs then reported
01:09 <@mids> there are still a few (possible) bugs out there..
01:09 <@mids> if you found some that aren't mentioned
01:09 <@mids> please tell us
01:10 < al-jabr> I personally couldn't replicate the terminal bug, unless that was something in CVS
01:10 <@mids> without bugreports we cant fix
01:10 <@nop> neither could i
01:10 <@mids> al-jabr: I have had it in the past; but couldn't repeat
01:10 <@mids> I'll ask Jeekay for more details
01:10 < philocs> where do I find list of outstanding bugs?
01:11 <@nop> well everyone is encouraged to use the sourceforge bug tracker
01:11 <@nop> but most people don't
01:11 <@nop> ;)
01:11 < philocs> bug tracker is good
01:11 <@nop> we should probably link to that on our main site
01:11 <@mids> http://www.sourceforge.net/projects/invisibleip/
01:11 <@nop> for a bug submital
01:11 < firegod> too bad it doesnt have an IRC frontend (:
01:11 <@mids> most bugs are mailed to the iip-dev mailinglist though
01:11 < philocs> ok I just subscribed yesterday
01:11 <@mids> cool
01:12 < nemesis> cause the logo, whate resolution? and dpi ?
01:13 <@nop> any one knowing graphix have a suggestion for resolution and dpi?
01:13 < firegod> start big
01:13 < firegod> it can be resized
01:13 <@nop> ok
01:13 <@nop> kewl
01:13 < firegod> down if needed
01:13 < nemesis> -e
01:13 < firegod> it is much more difficult going the other way (:
01:13 < nemesis> hehe
01:13 < nemesis> firegod
01:13 < nemesis> something
01:14 < hobbs> nop: would you be interested in having it in a vector graphics format, if that just happens to be how it's done?
01:14 < firegod> always good to have high res masters
01:14 < nemesis> 10 megapixels
01:14 <@nop> svg?
01:14 < nemesis> 72dpi
01:14 < hobbs> (not that I even own a vector program, but somebody might care)
01:14 < nemesis> or 1000 ?
01:14 < nemesis> ;)
01:14 < nemesis> very dificult
01:14 < nemesis> +f
01:15 < firegod> sure, if they are creative..
01:15 < firegod> but svg isnt widly used just yet
01:15 < al-jabr> Question:  I'm patching IIP to use /dev/random.  Would you be interested in incorporating this?  I'm doing it #ifdef linux for until I or someone configurifies the source.
01:15 < firegod> so standard raster formats would be more usable atm
01:15 < firegod> al-jabr: finish the patch and submit it to the mailing list
01:15 <@nop> al-jar
01:15 < al-jabr> okay
01:15 < hobbs> al-jabr: I'd suggest making it #ifdef SOME_FLAG_THAT_CAN_GO_IN_MAKEFILE
01:15 < al-jabr> yeah
01:16 < al-jabr> will do
01:16 < hobbs> (and have a well-commented DFLAGS line in Makefile)
01:16 <@nop> yarrow is a very good prng
01:16 <@nop> it's known to be secure
01:16 <@nop> and we have done a test with our randomness via chi-square
01:16 < al-jabr> nop: I believe yarrow would be redundard when we have /dev/urandom
01:16 <@nop> and it got 25% which is good 
01:16 <@nop> yes, but yarrow is portable
01:16 <@nop> and known to be stronger
01:16 <@mids> al-jabr: the problem is that not all operating systems have a good implementation for /dev/random
01:16 < firegod> not at all
01:17 <@nop> I would rather rely on what a cryptography expert developed
01:17 <@nop> then the /dev/random on the machines
01:17 < hobbs> true. A -DUSE_DEV_RANDOM might end up being useful, or it might just hurt a lot of people who don't know what they're doing.
01:17 <@mids> otoh, giving the more modular future of IIP, maybe several alternatives could be an option
01:17 < hobbs> and not the best odds on the first. :)
01:17 <@nop> yes
01:17 <@nop> and we do plan to add more entropy in the future
01:18 < al-jabr> Well, linux /dev/random and /dev/urandom are some of the most scrutinized crypto out there... I'm mostly thinking of that because it's a very good entropy pool that's out there on very many machines running IIP
01:18 <@nop> to increase this
01:18 < firegod> general question: with iip2 are we going to have more feedback from the proxy?
01:18 < al-jabr> you wouldn't have to go querying the user for entropy.
01:18 <@nop> yes firegod
01:18 <@nop> well you usually don't
01:18 <@nop> but it's definitely added plus
01:18 <@nop> if there isn't enough
01:18 <@nop> it will query
01:19 <@nop> and we will probably look into adding a form of /dev/random entropy very soon
01:19 < hobbs> does linux /dev/random support O_NONBLOCK ?
01:19 <@nop> because we intend on really strengthening the pool
01:19 <@nop> I'm sure it does hobbs
01:19 <@nop> /dev/random let's you select your pool size
01:19 < hobbs> nop: yeah, but there's a softlimit, and a hardlimit in the kernel, and the hardlimit isn't that big.
01:19 <@nop> al-jabr it would be best to hold off
01:19 < al-jabr> nop: personally I'd trust linux more, which uses SHA1 and uses all kinds of hardware sources of entropy, than a newbie who might just go entering 'aaaaaaaaaa...'  but anyway it's only an option
01:20 <@nop> al-jabr
01:20 < al-jabr> ok
01:20 <@nop> thats not all the entropy
01:20 <@nop> there is more
01:20 <@nop> there are network timings, and dh calculation timings as well
01:20 < al-jabr> but it only has access to user-mode entropy
01:20 <@nop> and we plan to add more
01:20 < al-jabr> why reinvent the wheel.   i recommend using /dev/random and for those who don't have it, EGD.
01:20 <@mids> nop: would it harm to give al-jabr a try, and maybe use it as plugin for entropy?
01:20 < al-jabr> since the GPG and linux people are doing it
01:21 <@mids> nop: alww
01:21 < al-jabr> why don't we concentrate on doing what we do best?
01:21 <@nop> that's fine
01:21 <@mids> nop: always good to have alternatives around
01:21 <@nop> if you want to submit a patch
01:21 <@nop> please do
01:21 <@nop> I'm not against it
01:21 <@nop> and we definitely want to add more entropy
01:21 < philocs> is the darwin /dev/random good? is it the same one in linux or openbsd?
01:21 <@nop> so please submit it to iip-dev when you've added it
01:21 < firegod> thats what mailing lists are for, people can digest it better
01:22 < al-jabr> okay, will do.
01:22 <@nop> thnx
01:22 <@nop> is that all?
01:22 <@nop> no more questions?
01:22 <@mids> hehe
01:22 <@nop> or suggestions
01:22 <@nop> or complaints
01:22 < nemesis> hm..
01:22 < philocs> I have a dumb newbie question ...
01:22 <@nop> sure
01:22 < firegod> well. release dates?
01:22 < nemesis> cache in the nodes
01:22 <@mids> sjoet
01:22 <@nop> oh oh on
01:22 <@nop> that wasn't on the list
01:22 <@nop> but
01:23 <@nop> we are at this time working on a short term todo list
01:23 <@nop> that will be publicized
01:23 < philocs> if someone hacks a relay to log, does that mean they can see the trafic for private channels that go through it?
01:23 <@nop> no
01:23 <@mids> philocs: all traffic is encrypted node-node and end-end
01:23 < philocs> ok, so you can only get the cleartext at the server, right?
01:23 < firegod> but not contextually withing IRC
01:23 <@nop> right
01:24 < firegod> right
01:24 < firegod> and the client
01:24 <@mids> philocs: correct
01:24 <@nop> yes
01:24 < philocs> good
01:24 < firegod> how far are you from encrypted channels?
01:24 < hobbs> and the client -- well, can only see stuff that's actually sent to it.
01:24 <@mids> firegod: nop is working on a roadmap and syncing it with the developers (if I understood well)
01:24 < nemesis> add an multicast option for filetransfers, when one user, will send the same file to some multiple clients
01:24 < hobbs> which means, if you don't even know that a channel exists, then you can't (intentionally or accidentally) snoop it.
01:24 <@nop> firegod it will be done when we decentralize
01:24 <@nop> which is our next goal
01:25 <@nop> after 1.1 stable
01:25 < nemesis> that the nodes between the nodes cache it
01:25 < firegod> okay. roadmap.
01:25 < hobbs> nemesis: actually.... that's worth thinking about -- talk to chocolate. :)
01:25 < philocs> is there an advantage to having "channel key encryption" before decentralization?
01:25 <@mids> nemesis: well, filetransfer isnt implemented in IIP itself anyway
01:25 < nemesis> lol
01:25 < firegod> hobbs: well, knowing about a channel is easy
01:25 <@mids> nemesis: it CAN do multicast, just send it to a channel :)
01:25 < hobbs> nemesis: it should be possible to add a hack to fileserv to have it use a channel, and then anyone who wants to receive just joins. :)
01:25 < hobbs> firegod: oh, is it?
01:25 < nemesis> what can you do with an anonymous network
01:26 < nemesis> when you can share code?
01:26 < nemesis> whats about some c code?
01:26 < firegod> multicast is a problem due to not spectacular widespread support..
01:26 <@mids> philocs: yes, I'd think so... less trust needed on the server
01:26 < nemesis> when the complet internet are banned for open source?
01:26 < hobbs> firegod: not multicast IP, just "multicast" :)
01:26 < firegod> hobbs: re fileserv channel: that gives you encrypted channels btw (:
01:26 < nemesis> how you can share this information?
01:26 < hobbs> firegod: oh, how's that?
01:27 < nemesis> <hobbs> nemesis: it should be possible to add a hack to fileserv to have it use a channel, and then anyone who wants to receive just joins. :)
01:27 < firegod> hobbs: sure, if you join IIP at all it is simple to /list the channels
01:27 < nemesis> not a hack
01:27 < philocs> I might start thinking about some 'channel key encryption'. it doesn't seem like it would be terribly complicated thing to me, just keep private keys in some directory maybe
01:27 < nemesis> built in
01:27 < nemesis> and an "server node" option
01:27 < nemesis> to allow that
01:27 < nemesis> or not
01:27 <@mids> philocs: you could implement it client side...
01:27 < hobbs> nemesis: okay, I'm just behind the times. I haven't worked on fileserv for... months
01:27 < nemesis> and an option for the cache size for it
01:27 <@mids> philocs: look at the blowfish.pl scripts for irssi and xchat
01:27 < firegod> philocs: and perl plugins on clients
01:27 <@mids> s/blowfish/blowjob/
01:28 < philocs> mids: would it make sense to implement it in the client side of isproxy?
01:28 <@mids> nemesis: caching wouldnt make much sense when everything goes still through the central ircd
01:28 < philocs> that way it would work with all clients
01:28 < nemesis> <mids> nemesis: caching wouldnt make much sense when everything goes still through the central ircd
01:28 <@mids> philocs: maybe; but that would require the 'vircd'
01:28 < nemesis> i think there are planned to be an p2p network?
01:28 < nemesis> and then theres no central hub
01:28 <@mids> nemesis: for IIP 2
01:29 < nemesis> only some nodes
01:29 < nemesis> where cache the datas
01:29 <@mids> nemesis: but that is long term; first IIP 1.2
01:29 < philocs> nemesis: I think you want freenet maybe
01:29 < nemesis> no
01:29 < philocs> p2p file transfers with caching
01:29 < nemesis> only an option to share some public files
01:29 < nemesis> or larger text
01:29 < philocs> thats what freenet does
01:29 < firegod> any merging of namespace possible between freenet and iip?
01:29 < nemesis> that you don'*t copy it line for line in the channel /query
01:29 < hobbs> what sits on top of the IIPv2 network could be a lot of interesting things -- but that's a while off. :)
01:29 <@mids> nemesis: first we would need decentralized routing...
01:30 < nemesis> k
01:30 < firegod> every isproxy was a freenet node?
01:30 < nemesis> but don't forget it ;)
01:30 < philocs> I don't think it makes sense to cannabalize freenet ...
01:30 <@mids> nemesis: once we have that; ask again :)
01:30 < firegod> philocs: does it do the job?
01:30 < nemesis> lol
01:30 <@mids> philocs: giving recent freenet-shit; I'd say no, indeed it doesn't
01:30 < firegod> philocs: and I like 'incorporate' a bit better
01:30 < hobbs> it should be possible to write a mini-freenet on top of IIP... but it would be better to leave freenet at what it does, and take advantage of the high speed and "pushiness" of IIP to write even better things.
01:30 < nemesis> in how many years? *fg*
01:31 < firegod> alright (:
01:31 < firegod> people do want to exchange chunks of binary data thru their messaging clients, in this case IIP
01:31 < firegod> how will that be addressed?
01:31 < philocs> firegod: well, I think it does the job well, and it will only get better. yes I agree that it would be better to have iip implement the freenet protocol for freenet type things rather than make something incompatible
01:31 < hobbs> for example, IIPv2 should be able to support the niftiest "anonymail" anyone's ever seen (without a bot), unless I'm hallucinating. :)
01:31 < nemesis> hm..
01:32 < nemesis> hacker ethic
01:32 < nemesis> the slogon
01:32 < nemesis> for..
01:32 < nemesis> miiiids!!
01:32 <@mids> hobbs: IIPv2 will be so smart that it could do your math homework
01:32 < hobbs> that's good, 'cause I don't do mine often enough.
01:32 < philocs> speaking of which
01:33 <@UserX> firegod: the intention is to do a DCC emulation using Freenet as the transport for files
01:33 < Grishnav> Sorry if this has already been suggested, I've missed much of the conversation being in and out of the room, but how about some sort of API for IIP to create modules? After IIP gets completely distributed (with v2) you could have all sorts of interesting modules pop up... a file transfer mod, perhaps a freenet node mod if you only wanted one service running...
01:33 < firegod> UserX: that'll work (:
01:33 < philocs> UserX: I think that is the best solution
01:33 < hobbs> Grishnav: that's more or less the plan, as I understand it. And if it's not, we'll beat nop with halibut until it is.
01:33 < Grishnav> lol
01:34 < firegod> UserX: but if IIPv2 is decenteralized, would this dcc emulation need freenet? you already can do point multipoint point transfers, you just need a session handshake for that kind of transfer
01:34 < firegod> albiet dcc
01:34 < nemesis> waaaaaaaaaah
01:34 < philocs> plus if every iip user was running some sort of freenet implementation, that would make freenet much better
01:34 < nemesis> ardvark
01:34 < nemesis> grrrrrr
01:34 < nemesis> where is he?
01:34 < nemesis> where can speak german?
01:34 < hobbs> also, it should be (more) convenient to have multiple IIPv2 networks, but I think that's a given. :)
01:34 < nemesis> or known only a little bit german
01:34 < firegod> philocs: thats what I'm saying (:
01:34 < nemesis> and have the english hacker ethic?
01:34 < firegod> whos working on IIPv2?
01:35 < philocs> I need to go study for my german test soon
01:35 < philocs> firegod: are you left handed or in oz or something?
01:35 <@mids> hm, ppl; I got to go; keep chatting here
01:35 <@mids> bbl
01:35 < nemesis> hrhr
01:35 < nemesis> mids!!!
01:35 < firegod> philocs: nope, just a freak
01:35 < nemesis> don't drunk to much ;p
01:35 < nemesis> *fg*
01:36 < firegod> mids is working on IIPv2 I'm sure, anyone else? UserX?
01:36 < nemesis> nop
01:36 <@UserX> firegod: in theory yes. but currently we want to keep IIP low bandwidth. freenet would me suited transfering large volumes of data (and better because it doesn't have a constraint of realtime routing that IIP needs)
01:36 < nemesis> i think
01:36 <@nop> yes
01:36 < nemesis> aaaaaah
01:36 < nemesis> nop
01:36 < philocs> I guess what is really needed is for someone to write a C implementation of freenet ...
01:36 < firegod> UserX: this is true.
01:37 < firegod> UserX: or at least an opt-in on that feature
01:37 <@UserX> firegod: yes i am working v2
01:37 < hobbs> philocs: I agreed with that pretty heavily a few months ago, but right now I'm happy to let java fred do its thing, and settle down, before anyone clones.
01:37 < hobbs> (now that it _works_, that is)
01:37 < firegod> UserX: how have you solved scaling issues for resource location? ie: how do you find nodes originating #channels?
01:37 < philocs> UserX: yes well thats a good reason to not make it easy for people to do 'dcc' and to encourage them to use freenet
01:38 < firegod> philocs: it should just be opt-in.. people wanting to abuse their bandwidth, can go right ahead.. those on modems dont get killed (:
01:38 < hobbs> UserX: would be nice to keep in mind, though, that freenet is good at pulling things, and iip is good at pushing things. :)
01:38 < philocs> hobbs: well I agree, I think the java version is fine but if we are going to basically package freenet with iip somehow then eventually (and probably when freenet hits 1.0?) we will want a c implementation
01:38 < firegod> philocs: those wanting freenet backed features, change a setting and BLAM it just works
01:38 <@UserX> firegod: haven't worked out highly scalable system yet
01:39 < firegod> hobbs: IIP is a great way of grouping freenet keys (:
01:39 < hobbs> philocs: that's some pretty long thinking. :)
01:39 < firegod> UserX: ah. If you havnt peaked at Circle, I encourage you to (:
01:39 < firegod> I know mids said he'd played with it
01:39 < philocs> hobbs: well freenet is getting more stable all the time 
01:40 < youkai> yeah, i would never run freenet as long as its only java
01:40 < firegod> theres a slogan for ya d-:
01:40 < firegod> "getting more stable every day"
01:40 < youkai> too bulky
01:41 < philocs> youkai: its not too bad
01:41 < youkai> plus i think its shitty to have os software that only compiles on a corp owned language
01:41 < Grishnav> I don't like Java anymore than the next guy, but I certainly am a freenet fan. I'll use the java one, but only until I hear about a C implementation. :)
01:42 < youkai> i mean if you guys were using the os non sun java i wouldent mind as much
01:42 < youkai> ah yes
01:42 < youkai> blackdown
01:42 < hobbs> youkai: freenet works fine on a few flavors of non-sun java.
01:42 < hobbs> blackdown has sun behind it.
01:42 < youkai> you just cant win with java then :/
01:42 < firegod> so?
01:42 < youkai> i dont trust sun any more then i do microsoft
01:43 < firegod> java is not your friend (:
01:43 < Grishnav> Does anyone have a link to the souce download for Blackdown? (Their site is less than helpful)
01:43 < firegod> I encourage those who are disatisfied with java, to try phthon for their scripting needs (it is NOT java)
01:43 < youkai> yeah python is cool
01:44 < youkai> but i dident stop running m$ operating systems just so i could let another corp in the door (sun)
01:44 < hobbs> Grishnav: er. It's in "non-free" for a reason, isn't it?
01:44 < philocs> you are wanting me to write freenet in python? would a python module be distributed with iip?
01:44 < Grishnav> Ahh... I was under the impression is was free. my mistake.
01:45 < youkai> thats the only problem i have with freenet
01:45 < philocs> java is not evil, sun treats java differently than MS treats windows
01:45 < hobbs> Grishnav: no. If you ask sun, it's impossible to create a free java2 implementation, and they've done a good job of making it true.
01:45 < youkai> i mean java is a lot easier to code in because you dont have to worry about memory leaks and stuff as much
01:45 < Grishnav> rofl
01:45 < youkai> the garbage collector lets you be lazy
01:45 < philocs> hobbs: why is it impossible?
01:45 < Grishnav> [16:45] <youkai> i mean java is a lot easier to code in because you dont have to worry about memory leaks and stuff as much -- yeah, it's no wonder that all java apps are so goddamn memory hoggy!!
01:46 < youkai> yeah thats because they need the whole jre loaded in memory with the software
01:46 < hobbs> philocs: because if you write anything that's java2, and claims to be "java", then sun will destroy you. :)
01:46 < philocs> hobbs: yes but you can make java, just don't call it 'java'
01:46 < hobbs> er... without obtaining the appropriate license and signing the appropriate agreements first, that is. :)
01:46 < Grishnav> call it coffee
01:46 < philocs> kaffe
01:46 < Grishnav> hehe
01:46 < Grishnav> yeah
01:47 < Grishnav> I've played with Kaffe
01:47 < hobbs> philocs: true. But nobody's done it.
01:47 < Grishnav> not quite mature enough yet, but getting there
01:47 < philocs> hobbs: uh yes, the FSF has done it
01:47 < hobbs> philocs: oh?
01:47 < philocs> yes
01:47 < youkai> but seriously i think java is right up there with VB
01:47 < philocs> Kaffe
01:47 < hobbs> philocs: Kaffe is not java2.
01:47 < youkai> its for lazy programmers
01:47 < youkai> who dont mind being owned by a corp
01:47 < philocs> hobbs: but there is no reason it could not implement java2
01:47 < hobbs> philocs: except for the fact that it doesn't.
01:47 < philocs> plus gccj or whatever its called
01:48 < hobbs> er...
01:48 < youkai> the other thing is java2 is huge, and they have a gigantic team of programmers working on it all the time
01:48 < hobbs> yeah. gcj/gij are also nice.
01:48 < firegod> not to interupt, but java wars work out better in apropriatly named channels (:
01:48 < philocs> hobbs: but its not a legal issue, the java spec is an open standard, the java name is not
01:48 < youkai> gcj?
01:48 < philocs> youkai: gcc that compiles java code
01:48 < youkai> huh
01:49 < youkai> to binary or does it still need a jre
01:49 < philocs> binary I believe
01:49 < hobbs> philocs: that's a pretty heavy restriction, though.
01:49 < hobbs> You can't say: this is java, this is compatible with java, or this smells like java.
01:49 < philocs> hobbs: well I don't think so. You can make the claim that 'this software is not java, but you will probably find that it works the same'
01:50 < philocs> which most people would understand
01:50 < hobbs> probably.
01:50 < youkai> anyway, why rewrite java when you could just use c++
01:50 < youkai> its almost the same language
01:51 < philocs> arg, I would rather use java over c++
01:51 < philocs> but I'm not getting into that
01:51 < philocs> anyway, I forgot where this horrible diatribe started
01:51 < hobbs> youkai: not really. c++ doesn't force you to use OO crap when it's completely inappropriate, like java does. :)
01:51 < firegod> round and round we go, where we stop nobody knows
01:51 < firegod> philocs: exactly
01:51 < philocs> ok, so in isproxy, is there like a client side and a node side?
01:52 < firegod> philocs: you know how many times I've seen this exact same 'argument' ? (:
01:52 < youkai> hobbs: hah
01:52 < firegod> philocs: there are relays, and proxys and 'servers'
01:52 < firegod> as I see it
01:52 <@UserX> philocs: can you clarify your question?
01:52 < philocs> I mean, would it make sense to put channel key encryption in isproxy, the part that actually talks to the irc client on 6667?
01:52 < hobbs> philocs: sorta. there are nodes, and there are nodes. :)
01:52 < firegod> philocs: dont forget you have multiple clients for each isproxy
01:53 < hobbs> and nodes 1) talk to clients 2) talk to nodes 3) (one of them) talks to the server.
01:53 < philocs> firegod: really? I've never been able to see this behavior, actually maybe its just my configuration
01:53 < firegod> (:
01:54 < philocs> but anyway, does my question make sense?
01:54 < youkai> i just came here to beg you guys not to write the next ver of iip in java :D
01:54 < firegod> which question d-:
01:54 < firegod> youkai: i think thats a given
01:54 <@UserX> philocs: currently IIP 1.x is essentially a tunnel. having the client implement channel encryption would require a lot of work to do. and would become redunant when v2 gets done
01:54 < youkai> also if theres freenet people around, a c++ ver would be nice
01:55 < firegod> UserX: how about isproxy functioning as an http tunnel?
01:55 < firegod> UserX: IIPv2 as well?
01:55 < nemesis> <youkai> i just came here to beg you guys not to write the next ver of iip in java :D
01:55 < nemesis> noooooo
01:56 < philocs> I'm thinking that you could have it so that there ways like a 'keys/' directory and then you could have in that 'channel.key' or something and then just run blowfish or whatever on what goes in and out of that channel, understand?
01:56 < nemesis> native code are the best thing
01:56 < philocs> and fuck c++, I'll take java over c++ anyday
01:56 < philocs> but I also think that c is nice
01:57 <@UserX> firegod: 1.x could be used to tunnel to a single fixed HTTP server
01:57 < firegod> okay, enough language wars please?
01:57 < nemesis> m$ sponsored his .net campain, and will place his IL on the front
01:57 < youkai> k :D
01:57 < firegod> User: hrmm
01:57 < nemesis> you can controll the compiller
01:57 < philocs> youkai keeps brining it up, if he likes c++ so much, he should marry it 
01:57 < nemesis> thats the different
01:57 < firegod> oh jebus
01:57 < youkai> heh philocs: if you like java so much you should go work for sun
01:57 < nemesis> can't
01:58 < philocs> UserX: would that make sense or is it better to wait for next version to do that?
02:00 < youkai> UserX: thats a good idea
02:00 <@UserX> philocs: to do that with 1.x network would require giving nodes the intelligence to read and parse recompose IRC client messages/commands
02:01 < philocs> oh I see
02:01 < nemesis> <UserX> philocs: to do that with 1.x network would require giving nodes the intelligence to read and parse recompose IRC client messages/commands
02:01 < nemesis> xml ;)
02:01 <@UserX> it's possible but would take a fair amount of effort which i want to put into v2
02:01 < nemesis> very flexible
02:01 < philocs> I understand
02:02 < philocs> later
02:11 < logger> logging ended
--- Log closed Wed Oct 16 02:11:14 2002