Wednesday, November 12, 2014< ^ >
Room Configuration
Room Occupants

[00:00:21] <Dave Thaler> Pete Resnick at mic
[00:00:57] <Brian Trammell> actually jana, if you could mic that: the factoring of this terminology, if not the words, support _everything_ we're talking about here.
[00:01:25] <Brian Trammell> you can ask for a component, you can ask for a composition, you can ask for a protocol instance
[00:01:49] <Dave Thaler> Joe Hildebrand at mic
[00:01:52] <Brian Trammell> (indeed this discussion is a very nice test of this factoring)
[00:02:10] <> Brian: What should I mic?
[00:02:28] <> (sorry lost context while standing in line)
[00:02:32] <Dave Thaler> Andrew McGregor at mic
[00:02:34] <Brian Trammell> the factoring of this terminology, if not the words, support _everything_ we're talking about here. you can ask for a component, you can ask for a composition, you can ask for a protocol instance
[00:02:55] <Gorry Fairhurst> - i think so too
[00:02:58] <Brian Trammell> what it doesn't support describing yet is getting info up from the lower layer
[00:03:48] <Gorry Fairhurst> - I think we're going to need some lists next, which will be equal fun
[00:03:54] <Dave Thaler> virtual Brian Trammell at mic
[00:04:29] <Brian Trammell> yep yep
[00:04:30] <Dave Thaler> Kevin Fall at mic
[00:04:42] <Brian Trammell> i'm looking for appropriate swiss german words in another tab
[00:04:46] <> (I look a lot like you too, Brian)
[00:05:26] <Brian Trammell> :)
[00:06:12] <Gorry Fairhurst> (timw to move the discussion on to the next slides guys...)
[00:07:40] <Dave Thaler> slides:
[00:08:13] <> slide 2: Document Goal
[00:08:24] <> slide 3: Document organization
[00:09:25] <> slide 4: Relationship between transport protocols and service component ...
[00:10:27] <Gorry Fairhurst> Mic: I think the draft mentions the things that Joel said
[00:10:33] <Brian Trammell> mic: gorry and I literally threw this together in the nine seconds before the deadline
[00:10:48] <Brian Trammell> we expect this list to get much bigger
[00:11:30] <Gorry Fairhurst> +1 (but have the people with comments read it?)
[00:11:40] <hildjj> Brian: understood.  i just wanted to remind folks.  the earlier you can add those things, the easier it will be to pull in outside expertise into the overall work stream
[00:11:43] <Brian Trammell> this draft is about structure: do we think this represents a good way to describe transport (aspects/components) etc.
[00:12:14] <Brian Trammell> joe: yes, of course.
[00:12:18] <MarieJose Montpetit> Is it about structure or features?
[00:12:40] <Dave Thaler> Jana Iyengar at mic
[00:12:54] <Dave Thaler> Aaron replying that charter limits this WG to only include IETF protocols
[00:12:56] <Brian Trammell> structure *of the document*
[00:13:01] <MarieJose Montpetit> ok
[00:13:13] <Brian Trammell> once we decide we like this structure
[00:13:16] <Brian Trammell> then we fill it in
[00:13:20] <hildjj> I'll note that I disagreed with that part of the charter. :)
[00:13:46] <Dave Thaler> where in the charter is it (I believe it's there but can't find it right now)
[00:14:12] <Dave Thaler> Andrew Mcgregor at mic
[00:15:07] <hildjj> Dave: "Define a set of Transport Services, identifying the
services provided by existing IETF protocols and congestion
control mechanisms."
[00:15:23] <Brian Trammell> CoAP is absolutely a transport.
[00:15:33] <Dave Thaler> absolutely
[00:15:52] <Brian Trammell> i have recorded it as well
[00:15:55] <Brian Trammell> offsite backup
[00:16:01] <> :-)
[00:16:10] <Dave Thaler> @Joe: that statement says to *prioritize* them, not that others are out of scope
[00:16:13] <Gorry Fairhurst> Gorry -> MIc: These are just examples of what we need to fill in
[00:16:38] <> Gorry — can you give more context?
[00:16:44] <Brian Trammell> +bignum joe
[00:17:04] <Brian Trammell> (jana, are you in queue?)
[00:17:13] <> I can be
[00:17:17] <Brian Trammell> ah
[00:17:52] <Gorry Fairhurst> Gorry: What we want is a plkan to make one of these sections concrete with expertise.
[00:17:53] rachel huang joins the room
[00:17:59] <Brian Trammell> security is definitely an aspect
[00:18:04] <Gorry Fairhurst> +1
[00:18:22] <Brian Trammell> scroll up
[00:18:23] <Brian Trammell> jana
[00:18:43] <Brian Trammell> please underline we're asking the structure question here
[00:19:29] <> (sorry if I'm not relaying accurately)
[00:19:38] <Brian Trammell> np it must have scrolled up
[00:19:48] <Brian Trammell> sorry i'm sitting at a desk in a suburb of zurich :)
[00:19:58] <> Yeah, I lose context when I'm standing in line and thinking of other things.
[00:20:27] <Gorry Fairhurst> Thanks Jana - it's a fairly random set of thoughts on the whole charter... you do well)
[00:20:28] <Brian Trammell> next time i'll ask for the Scary Disembodied Voice from the Sky power.
[00:20:36] <> :-)
[00:21:06] <Brian Trammell> +bignum joe again
[00:21:43] <> 'bignum' used without initialization
[00:22:02] <Brian Trammell> import silly_constants
[00:22:37] <> Kevin Fall contributed for UDP
[00:22:37] <hildjj> I live in a 16bit world.  bignum=65535 is perfectly adequate.
[00:22:49] <> *contributed -> volunteered
[00:23:01] <Brian Trammell> RTP yes
[00:23:15] <Dave Thaler> Kevin Fall from floor: "open mouth, insert work"
[00:23:28] <csp> I can help Varun with the RTP text
[00:23:29] <Brian Trammell> mic: I can move to markdown over github no problem
[00:23:33] <Dave Thaler> Varun Singh at mic
[00:23:39] <hildjj> Kevin: nice one.
[00:23:55] <Dave Thaler> Kevin Fall at mic
[00:24:30] <Gorry Fairhurst> I think we need to get ONE or TWO sections good for the next IETF --- I think we should not write lots of sections in parallel
[00:24:39] <Brian Trammell> yes absolutely
[00:25:04] John Leslie joins the room
[00:25:10] <Gorry Fairhurst> [[Kevin's on this - if we all start with different language we go crazy in final edit]]]
[00:25:24] <> Varun at mic
[00:25:38] <> Joe Hildebrand earlier
[00:26:32] <> slide: Terminology slide from
[00:27:10] <csp> Some of the protocols suggested for this draft offer functionality that’s much broader than “services used between two endpoints” – how wide is this draft aiming to become? (e.g., RTP is fundamentally multiparty)
[00:27:36] <Gorry Fairhurst> gorry->CSP :: Yes that worries me RTP does a lot.
[00:28:04] <Brian Trammell> +1 aaron
[00:28:07] <hildjj> we need to capture the multicast/broadcast service components.
[00:28:14] <Gorry Fairhurst> I think we need to start simply with the core transports we understand (as aaron said)
[00:29:07] <csp> Gorry: so do many of the others, of course
[00:29:09] <MarieJose Montpetit> Agree
[00:29:20] <Gorry Fairhurst> Oh yes
[00:29:33] <hildjj> start with TCP and UDP.
[00:30:06] <Brian Trammell> Dave: TCP?
[00:30:35] <csp> Just doing TCP and UDP initially might lead to surprises later; might be better to start with two that are more different?
[00:30:41] <Gorry Fairhurst> Gorry: I think we would like to see ideas for the matrix
[00:30:42] <Brian Trammell> csp point:
[00:30:59] <hildjj> fine.  start with TCP and something much simpler than SCTP
[00:31:01] <Dave Thaler> Jana Iyengar at mic
[00:31:02] <Brian Trammell> so UDP and RTP are nicely different
[00:31:14] <Brian Trammell> but one of them has to be TCP
[00:31:39] <hildjj> i think tcp is much more interesting to do than UDP
[00:31:46] <Dave Thaler> I had a matrix of TCP, MPTCP, SCTP, QUIC, Minion, WebSockets
[00:31:46] <Brian Trammell> yep
[00:31:47] <csp> agree that TCP is fundamental
[00:31:48] <hildjj> (at the beginning)
[00:31:53] <Brian Trammell> who's on the hook for tcp?
[00:32:13] <Gorry Fairhurst> SCTP already has a volunteer that helps
[00:32:14] <Dave Thaler> (of course some of those change as protocols evolve...)
[00:32:50] <Dave Thaler> I can help with TCP
[00:33:10] <Dave Thaler> Karen Nielsen at mic
[00:35:03] <Dave Thaler> Pal-Erik Martinsen at mic, volunteering to cover STUN
[00:35:10] <Dave Thaler> Charles Eckel at mic
[00:36:17] <Brian Trammell> giant table: another reason to test the new RFC format with this.
[00:36:33] <Dave Thaler> Ken Calvert at mic
[00:36:41] <Gorry Fairhurst> Tabletext editor ;-)
[00:37:03] <Brian Trammell> mic: ken: yes, but i am allergic to the world ontology
[00:37:19] <Brian Trammell> s/world/word/
[00:37:42] <Brian Trammell> and the point is we look at the protocols first, then we list the service components
[00:37:58] <Dave Thaler> Brian Adamson at mic
[00:38:43] <Brian Trammell> yes
[00:38:45] <Gorry Fairhurst> gorry MIC ::: THis is the NEXT JOB TO MAKE THIS LIST OF THINGS TO FIND
[00:39:02] <Dave Thaler> Joe Hildebrand at mic
[00:39:32] <Dave Thaler> (virtual brian and virtual gorry)
[00:39:39] <Brian Trammell> can we get the doc structure slide back up?
[00:39:47] <Dave Thaler> Michael Ramalho at mic
[00:40:04] <resnick> Those of us with philosophy degrees hate the use of the term "ontology" in this context even more than the rest of you. ;-)
[00:40:23] <hildjj> resnick: can you come up with a set of terminology we could use instead?
[00:40:40] <hildjj> some sort of meta-ontology, perhaps?
[00:40:53] Frank Xia joins the room
[00:41:12] <hildjj> i'll note that for a WG meeting that was supposed to be over in 30m, we're almost 2H in.
[00:41:40] <Brian Trammell> well we've had a lot of pretty good (off-agenda) discussion
[00:42:03] <Brian Trammell> and i think since people are basically already working on this document, i don't anticipate a big fight over adoption
[00:42:31] <Dave Thaler> I can imagine the matrix (list of components) will get *really* big
[00:43:02] <hildjj> we don't have to format it as a table.  we can think of it as a set of tags.
[00:43:16] <Gorry Fairhurst> Indeed ::: Let's move on with the first document - not worry about the whole problem (it's getting late hear to worry about fixing the world tonight)
[00:43:26] <hildjj> FIX THE WORLD!
[00:43:32] <Dave Thaler> that's fine too (but I predict it will be a long list of tags)
[00:43:32] <Brian Trammell> dave: p51 of
[00:43:43] <Brian Trammell> (thesis of a student of Michael Welzl's)
[00:43:57] <Brian Trammell> *hand up*
[00:44:09] <Brian Trammell> *hums*
[00:44:15] <MarieJose Montpetit> I also agree
[00:44:16] <csp> hum
[00:44:34] <Dave Thaler> Brian: that's a good start
[00:44:34] leaves the room
[00:44:46] <Brian Trammell> okay we'll get a -00 into github soon
[00:44:51] <MarieJose Montpetit> Thanks
[00:44:57] MarieJose Montpetit leaves the room
[00:45:04] <Gorry Fairhurst> Some peopel should say thanks to Mirja --- for being where we could not be, and doing it all so abley!
[00:45:29] <Dave Thaler> Brian: some overlap with my table, easy to combine
[00:45:58] <Gorry Fairhurst> That's good
[00:46:39] resnick leaves the room
[00:46:44] <Brian Trammell> gorry: are you okay with .md and kramdown?
[00:47:02] John Leslie leaves the room
[00:47:42] <Gorry Fairhurst> Eh? what did you write?
[00:47:57] cheshire leaves the room
[00:48:34] PasiS leaves the room
[00:49:11] <Gorry Fairhurst> ... in other words not used this.
[00:49:23] Dave Thaler leaves the room
[00:49:36] csp leaves the room
[00:50:24] cheshire joins the room
[00:50:49] Gorry Fairhurst leaves the room
[00:54:58] Meetecho leaves the room
[00:56:58] cheshire leaves the room
[00:57:15] rachel huang leaves the room
[00:57:37] Brian Trammell leaves the room
[01:00:27] hildjj leaves the room
[01:07:34] Frank Xia leaves the room
[01:24:34] cheshire joins the room
[01:27:55] cheshire leaves the room
[01:29:45] cheshire joins the room
[01:38:23] PasiS joins the room
[01:38:48] PasiS leaves the room
[01:41:35] resnick joins the room
[01:41:37] resnick leaves the room
[01:58:25] cheshire leaves the room
[02:00:08] cheshire joins the room
[02:21:51] joins the room
[03:22:43] cheshire leaves the room
[03:25:26] leaves the room
[08:06:55] cheshire joins the room
[09:44:10] joins the room
[10:07:01] leaves the room
[18:38:13] cheshire leaves the room
[18:58:24] cheshire joins the room
[21:30:58] cheshire leaves the room
[23:03:44] cheshire joins the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!