Experimental IRC log happs-2007-11-08

Available formats: content-negotiated html turtle (see SIOC for the vocabulary)

Back to channel and daily index: content-negotiated html turtle

These logs are provided as an experiment in indexing discussions using IRCHub.py, Irc2RDF.hs, and SIOC.

00:00:00<perspectivet>then build up from there.
00:00:10<Saizan_>yup
00:00:42<Saizan_>but we should handle the membership messages in some way
00:01:17<Saizan_>to provide the information about groups/members in a organic way, but still consistent wrt ordering of messages
00:03:24<perspectivet>that might be something we could add after the initial message channel stuff was working
00:03:59<perspectivet>kind of a pre-emptive parse and dispatch for known membership messages
00:05:06<Saizan_>yes, but it kind of interacts with join/leave calls
00:06:15<perspectivet>how do you mean?
00:06:17<Saizan_>did you find more info about the Extended Virtual Synchrony model that's cited in the first pages of the guide?
00:06:22<perspectivet>no
00:06:32<Saizan_>well you've to be in a group to leave it for example
00:07:10<perspectivet>Right, we would need some sort of Group type
00:09:35<perspectivet>And a MailBox type
00:10:38<Saizan_>i'd call it Connection, but yes :)
00:11:08<perspectivet>well it looks like you can have multiple mailboxes per group
00:12:01<Saizan_>it's more like more mailboxes/connections can be a member of the same group
00:12:25<Saizan_>but yeah it's a N:N relation
00:12:35<perspectivet>Right
00:15:31<perspectivet>ok, so we're still thinking 1-1 mapping between groups and Chan, right?
00:16:15<Saizan_>initially i was thinking one Chan per mailbox
00:16:56<Saizan_>do we have guarantees wrt ordering of messages between groups?
00:18:54<perspectivet>I think we only have messages between members of a group, no?
00:19:21<perspectivet>even in the group multicast you would still be sending to groups your mailbox # was joined to
00:19:33<Saizan_>yes
00:21:36<Saizan_>if i receive a message for group A and then one for group B, is the order significant or not? i.e. the total ordering guarantee is only between messages sent to the same group or global? probably the former but i can't find it in the guide
00:22:39<Saizan_>ah, i think the description of Safe implies that the guarantee is valid only in a gruop
00:22:44<perspectivet>It depends on the ordering level. The highest level AGREED_MESS should be delivered with total order
00:23:46<Saizan_>uhm :)
00:24:07<perspectivet>but I'm not clear on how this would interact with lower ordering levels...
00:24:17<perspectivet>kind of complicated
00:24:56<Saizan_>however, this should be the repo, it's still empty though.. http://happs.org/hspread
00:25:31<Saizan_>but since i'm in europe i'm going to sleep ^^
00:26:22<Saizan_>you can send patches to sanzhiyan -at- gmail dot com if you write something, i'll work on this when i wake up
00:27:18<perspectivet>sounds good. I probably won't pound out anything for a few days. I should get spread up and running and play with the system a little bit
00:28:56<Saizan_>ok, bye
00:29:00<perspectivet>night
19:33:43<shapr>ugh
19:37:09<bakert>nice pattern
19:37:16<bakert>what happened there?
19:40:40<Saizan>there == freenode?
19:41:14<bakert>well i have been on all along but shapr, darring, MarcWeber and perspectivet all just dropped off and on again about 20 times
19:41:42<bakert>at the same time as my colleagues in palo alto had power problems. but that's just me making patterns where ther eare nown
19:41:46<bakert>*none
20:09:05<perspectivet>yeah, there have been a lot of netsplits going on today
20:09:05<perspectivet>not sure why
22:20:53<shapr>Repository urls have just changed a bit.
22:30:58<perspectivet>ok
22:31:42<perspectivet>Saizan was going to set up a repo for the hspread stuff we're doing/thinking of doing
22:31:57<perspectivet>Saizan: did that happen?
22:34:34<Saizan>yes, http://happs.org/repositories/hspread
22:35:17<Saizan>and i'm going to polish a little what i've so i can push it
22:36:00<perspectivet>I looked into the recollections that caused me to diverge a bit in our discussion about Network.* and ByteStrings
22:36:05<perspectivet>Have you seen this:
22:36:34<perspectivet>http://hackage.haskell.org/cgi-bin/hackage-scripts/package/network-bytestring-0.1.1
22:37:06<perspectivet>on a related note wrt serialization and deserialization:
22:37:08<perspectivet>http://www.cse.unsw.edu.au/~dons/binary.html
22:38:17<Saizan>yeah, and derive to have automatic instances of the Binary class
22:38:40<Saizan>is network-bytestring tested/affidable?
22:40:29<perspectivet>not sure. It should be easy to slip in later on too
22:41:21<perspectivet>Could be something for me to do after you do your first commit
22:51:10<Saizan>i've avoided the problem assuming i had an Handle some way :)
22:52:48<perspectivet>network-bytestring might not be an improvement over the approach we talked about yesterday and I agree it's preferrable to avoid questionable libs for the time being
23:13:16<Saizan>ok, first patch applied, the code obviously need to be rearranged but just to give the idea of what i have
23:14:58<perspectivet>cool
23:15:02<perspectivet>checkin' it out
23:28:33<perspectivet>Saizan: Looks decent.
23:30:03<perspectivet>Saizan: are you working with 6.8.1?
23:32:15<Saizan>6.6.1 right now because i've to reinstall binary on 6.8.1
23:33:26<perspectivet>yeah, just looking at that problem now
23:42:23<perspectivet>the version on http://code.haskell.org/binary works for me under 6.8.1
23:44:27<perspectivet>is tRANS_MEMB_MESS == tRANSITION_MESS?
23:47:22<Saizan>yes, i forgot to compile before pushing the patch..
23:51:06<Saizan>repull
23:51:26<perspectivet>aye aye captain
23:54:36<perspectivet>cool. that compiles under 6.8.1 with binary-0.4.1
23:54:48<perspectivet>on vista

Back to channel and daily index: content-negotiated html turtle