IRC log of mediafrag on 2010-02-24

Timestamps are in UTC.

10:00:29 [RRSAgent]
RRSAgent has joined #mediafrag
10:00:29 [RRSAgent]
logging to
10:00:31 [trackbot]
RRSAgent, make logs public
10:00:31 [Zakim]
Zakim has joined #mediafrag
10:00:33 [trackbot]
Zakim, this will be IA_MFWG
10:00:33 [Zakim]
ok, trackbot, I see IA_MFWG()5:00AM already started
10:00:34 [trackbot]
Meeting: Media Fragments Working Group Teleconference
10:00:34 [trackbot]
Date: 24 February 2010
10:02:46 [raphael]
raphael has joined #mediafrag
10:04:27 [Zakim]
10:05:01 [Zakim]
10:05:11 [erik]
regrets: raphael
10:07:10 [erik]
present: yves, michael, davy, erik
10:07:49 [erik]
present+ silvia
10:08:49 [erik]
TOPIC: Admin
10:09:06 [jackjansen]
jackjansen has joined #mediafrag
10:09:11 [davy]
scribenick: erik
10:09:13 [erik]
10:09:47 [erik]
Accept the minutes at
10:10:04 [davy]
10:10:08 [erik]
10:10:20 [silvia]
10:10:29 [Yves]
10:10:31 [Zakim]
+ +31.20.592.aabb
10:10:44 [jackjansen]
zakim, aabb is me
10:10:44 [Zakim]
+jackjansen; got it
10:11:25 [Zakim]
10:12:03 [Zakim]
10:12:38 [davy]
Zakim, unmute erik
10:12:38 [Zakim]
Erik&Davy was not muted, davy
10:14:01 [Zakim]
10:15:45 [Zakim]
+ +329331aacc
10:17:18 [davy]
scribenick: davy
10:17:46 [davy]
Erik: charter extension: everything fine, but milestones need to be adjusted
10:18:27 [davy]
Action: raphael to adjust the milestones in the charter
10:18:27 [trackbot]
Sorry, couldn't find user - raphael
10:18:44 [raphael]
ah, thanks :-)
10:18:49 [davy]
10:18:57 [raphael]
ACTION: troncy to adjust the milestones in the charter
10:18:57 [trackbot]
Created ACTION-145 - Adjust the milestones in the charter [on Raphaƫl Troncy - due 2010-03-03].
10:19:25 [raphael]
... actually on the group front page since we cannot change the charter, but the charter refers to the group home page for the changes
10:19:38 [erik]
10:19:38 [trackbot]
ACTION-143 -- Silvia Pfeiffer to move 5.1.5 into a new section -- due 2010-02-24 -- OPEN
10:19:38 [trackbot]
10:19:53 [erik]
10:19:53 [trackbot]
ACTION-144 -- Silvia Pfeiffer to move the section 5.1.1 to the top -- due 2010-02-24 -- OPEN
10:19:53 [trackbot]
10:19:54 [silvia]
not done yet
10:20:01 [silvia]
neither that one
10:20:05 [silvia]
I promised it for F2F
10:20:13 [erik]
ok, actions still going on
10:20:36 [davy]
erik: status about changes in the header syntax?
10:21:02 [davy]
yves: made modifications in the wiki regarding philip's remarks
10:21:46 [davy]
... only do it in the spec if we switch back to ABNF
10:22:04 [davy]
jack: I'm fine with either ABNF or EBNF
10:22:23 [davy]
yves: I have tools for ABNF only
10:23:10 [davy]
jack: no reason why we switched to EBNF, I don't see an advantage to use EBNF
10:23:17 [silvia]
I don't mind either - if you prefer to switch back to ABNF, go for it
10:23:41 [silvia]
I think Philip found it easier to write, but that doesn't have to influence everyone else
10:23:56 [davy]
jack: if Yves prefers ABNF, then I also prefer ABNF
10:24:05 [davy]
yves: will continue in ABNF
10:24:25 [davy]
... also create a section where all the ABNF is located together
10:24:34 [davy]
... having a single view of the grammar
10:25:45 [davy]
erik: propose to use ABNF to represent the grammar
10:26:23 [jackjansen]
PROPOSAL: dtddadada??? Ok, +1 :-)
10:26:33 [jackjansen]
I know...
10:26:34 [davy]
PROPOSAL: use ABNF to represent the grammar, because of the existence of tools
10:26:50 [erik]
10:26:51 [mhausenblas]
10:26:54 [davy]
10:26:55 [jackjansen]
10:27:08 [davy]
RESOLUTION: ABNF will be used to represent the grammar
10:27:50 [davy]
yves: a-126 is still ongoing
10:28:06 [davy]
... should be able to do it today
10:28:27 [silvia]
while you're at it, can you look at writing ABNF for the HTTP headers, too?
10:28:36 [Yves]
10:28:38 [Yves]
that's the plan
10:28:43 [davy]
erik: a-135 is still ongoing
10:28:57 [davy]
jack: a-137 is also ongoing
10:29:17 [silvia]
see note in 5.2.2 and 5.2.3
10:29:21 [davy]
10:29:21 [trackbot]
ISSUE-5 -- Handling spatial cropping requires information at client-side -- OPEN
10:29:21 [trackbot]
10:30:20 [davy]
erik: we must asap reach LC
10:30:33 [davy]
... if this is only with the temporal dimension, so be it
10:30:57 [jackjansen]
silvia, can you re-raise in AOB?
10:31:04 [davy]
10:31:36 [mhausenblas]
10:31:36 [trackbot]
ACTION-118 -- Michael Hausenblas to come up with individual 'normal' test cases (+/- 20) -- due 2009-10-07 -- OPEN
10:31:36 [trackbot]
10:31:39 [raphael]
I second silvia question, re: quotes in the syntax
10:31:49 [raphael]
Silvia, please, write your comments not in /me
10:32:07 [silvia]
re: ABNF - are we dropping the quotes on the fragment values?
10:32:34 [silvia]
just FYI - the "s" in npt dimension has already been dropped in spec
10:32:52 [jackjansen]
silvia, we first continue with 4, then get to 3.5
10:33:18 [davy]
michael: a-118 continues untill next week
10:33:32 [mhausenblas]
10:33:32 [trackbot]
ACTION-108 -- Michael Hausenblas to add the missing test cases in corrib -- due 2009-11-16 -- OPEN
10:33:32 [trackbot]
10:33:39 [mhausenblas]
close ACTION-108
10:33:39 [trackbot]
ACTION-108 Add the missing test cases in corrib closed
10:33:53 [mhausenblas]
10:34:10 [mhausenblas]
10:34:40 [davy]
michael: we can review them one by one
10:34:43 [silvia]
indeed - I was going to ask if people had had the time to check out those test cases for the time dimension
10:35:26 [davy]
jack: are the test cases already in a state that we can see that all the boundary cases can be seen
10:35:49 [silvia]
can we focus on time dimension for now and check there that we have all the boundary cases?
10:35:58 [erik]
michael: focus is on temporal dimension only
10:36:15 [davy]
michael: not totally sure, there might be at least a couple of test cases that are not yet covered
10:36:35 [davy]
jack: I will look at it
10:37:02 [silvia]
in my list I abstracted away from concrete values into t=a,b
10:37:21 [silvia]
then I looked at the values that a and b can take compared to each other and compared to start, end of file
10:37:34 [silvia]
in this way, I tried capturing all the possibilities
10:38:06 [silvia]
these are all no errors:
10:38:07 [silvia]
* Non-errors: b>=a
10:38:08 [silvia]
t=a,b -> 200: set playback to a-b
10:38:08 [silvia]
t=a, -> 200: set playback to a-end
10:38:08 [silvia]
t=,b -> 200: set playback to start-b
10:38:08 [silvia]
t=a,b with a=b -> 200: set playback to a
10:38:09 [silvia]
t=a,b with a<start -> 200: set playback to start-b
10:38:11 [silvia]
t=a,b with a>end -> 200: set playback to end
10:38:13 [silvia]
t=a,b with b>end -> 200: set playback to a-end
10:38:15 [silvia]
t=a,b with b<start -> 200: set playback to start
10:38:59 [davy]
michael: in the F2F, we should make a final decision regarding the test cases of temporal fragments
10:39:25 [davy]
jack: I missed b > end
10:39:56 [mhausenblas]
Michael: will add t=a,b with b>end and t=a,b with b<start
10:40:13 [davy]
jack: also miss b<a
10:40:34 [jackjansen]
silvia, both
10:40:41 [silvia]
it's in the list above
10:40:50 [mhausenblas]
we're looking at corrib ATM, but jackjansen referred to in general
10:41:10 [silvia]
b>end is, the other is in an error list
10:41:27 [silvia]
* non-existent
10:41:27 [silvia]
example: t=b,a with b>a (e.g. t=20,10)
10:41:27 [silvia]
-> 200: ignore fragment
10:42:10 [davy]
michael: it's easy to add/edit test cases in corrib
10:42:17 [davy]
... everyone can do it
10:42:28 [silvia]
and I also have this one that we haven't had before:
10:42:29 [silvia]
* Validity errors: value cannot be parsed for dimension
10:42:29 [silvia]
example: t=asdf
10:42:29 [silvia]
-> 200: ignore fragment (retrieve full resource, but with this adapted URL)
10:42:29 [silvia]
note: browsers already do this
10:43:16 [silvia]
further, I think we need to change TC0002
10:43:17 [silvia]
* TC0002: empty time segment - npt (see also above)
10:43:18 [silvia]
examples: t=0,0 ; t=3,3
10:43:18 [silvia]
-> 200: retrieve whatever the browser needs to set up playback and
10:43:18 [silvia]
set the playback position to the offset
10:43:18 [silvia]
NOTE: this is different to what we currently have at
10:43:19 [silvia]
10:43:29 [mhausenblas]
silivia, you're aware of that you can edit TC in corrib yourself?
10:43:37 [davy]
Action: Jack to identify and add in corrib any missing test cases for temporal fragments
10:43:37 [trackbot]
Created ACTION-146 - Identify and add in corrib any missing test cases for temporal fragments [on Jack Jansen - due 2010-03-03].
10:43:51 [silvia]
I thought we are discussing here whether we agree on those test cases and their results
10:45:13 [silvia]
the reason for my questions / list was that I wanted these in the spec as well as in corrib, so developers can refer to it
10:45:18 [mhausenblas]
silivia, I'll change the ownership re TC10-17 to you soon
10:46:03 [erik]
10:46:54 [erik]
michael: 'contributor' = 'one who raised the TC'
10:47:13 [silvia]
hmm, according to they are multi dimensional - wasn't really focused on that yet
10:47:33 [mhausenblas]
ACTION: mhausenb to add all MF WG members to corrib
10:47:33 [trackbot]
Created ACTION-147 - Add all MF WG members to corrib [on Michael Hausenblas - due 2010-03-03].
10:47:52 [mhausenblas]
ACTION: mhausenb to add a copy TC functionality
10:47:52 [trackbot]
Created ACTION-148 - Add a copy TC functionality [on Michael Hausenblas - due 2010-03-03].
10:49:15 [mhausenblas]
ACTION: mhausenb to come up with a fix for overview vs. edit single TC
10:49:15 [trackbot]
Created ACTION-149 - Come up with a fix for overview vs. edit single TC [on Michael Hausenblas - due 2010-03-03].
10:49:57 [RRSAgent]
I have made the request to generate davy
10:51:06 [davy]
jack: if we change the syntax, then someone needs to be responsible to modify the related test cases
10:52:26 [mhausenblas]
Michael: there is also an issue tracker for corrib in case you find a bug or want to request a feature
10:52:29 [mhausenblas]
10:52:37 [davy]
TOPIC: 3.1 Media Fragment URI syntax
10:52:56 [davy]
erik: do we keep the quotes around track names and ids?
10:53:53 [Yves]
10:53:58 [erik]
also see:
10:53:59 [Yves]
10:54:44 [davy]
jack: percent escapes of ? and & seems to work in the browser
10:55:08 [davy]
... which indicates that we don't need quotes (on condition that we have percent escaping)
10:55:43 [erik]
... reinvestigation is still needed!
10:55:49 [silvia]
which philip introduced into the spec
10:56:10 [jackjansen]
10:56:51 [silvia]
10:57:23 [erik]
jack: syntax of ? and # should be the same
10:57:31 [davy]
jack: we should first investigate how percent escaping works
10:58:53 [erik]
jack: what about client vs serverside ... should be the same at both sides
10:58:56 [mhausenblas]
10:59:02 [davy]
11:00:12 [mhausenblas]
11:00:42 [davy]
yves: before doing percent-encoding, the url is split up
11:01:21 [jackjansen]
If I compare section 2.1 to what I posted to the mailing list then it seems the behaviousr I saw is wrong.
11:02:06 [erik]
yves: URI-syntax is not clear on what should be returned on client or server .... therefore no ambiguity if we use quotes
11:02:08 [jackjansen]
2.4 states that even more clearly
11:02:16 [silvia]
according to, last section, only the names and values of a CGI query are encoded
11:02:24 [davy]
yves: quotes remove ambiguity
11:03:04 [davy]
jack: so youtube is not compliant to RFC3986?
11:03:10 [jackjansen]
This works:
11:03:13 [erik]
rrsagent, draft minutes
11:03:13 [RRSAgent]
I have made the request to generate erik
11:03:14 [jackjansen]
and it shouldn't
11:03:34 [mhausenblas]
my friend re URIs is
11:03:42 [jackjansen]
So, it seems that rfc3986 isn't followed by at least one major site
11:03:48 [mhausenblas]
lets you check stuff quite nicely
11:04:48 [silvia]
I think the "Under normal circumstances" part is important
11:05:01 [jackjansen]
11:05:12 [silvia]
I think in essence rfc3986 lets every server do what it wants
11:05:15 [jackjansen]
got it
11:05:26 [silvia]
link above -
11:05:46 [silvia]
but it suggests parsing before percent encoding
11:06:03 [silvia]
to make it easier for the server to determine the components
11:06:22 [silvia]
doesn't stop a server from trying to interpret a url that has more than just the components url encoded
11:07:09 [silvia]
I think we can safely prescribe url encoding the way Philip has currently done it and it follows section 2.4
11:08:05 [jackjansen]
I am not sure I agree. I am not convinced that we can remove the quotes for queries.
11:08:17 [silvia]
that is a different matter altoghether, I think
11:08:26 [silvia]
what problem do the quotes solve?
11:08:36 [jackjansen]
escaping ampersands
11:08:42 [jackjansen]
11:09:42 [silvia]
I seem to remember having had a problem where the quotes introduced more confusion than helped, but cannot now find it
11:09:49 [silvia]
anyway, I guess we're out of time :)
11:10:06 [davy]
Action: Erik to summarize the discussion on the quotes in a mail or on the wiki
11:10:06 [trackbot]
Created ACTION-150 - Summarize the discussion on the quotes in a mail or on the wiki [on Erik Mannens - due 2010-03-03].
11:10:19 [Zakim]
11:10:20 [davy]
11:10:21 [jackjansen]
11:10:21 [Zakim]
11:10:23 [Zakim]
11:10:27 [Zakim]
11:10:28 [Zakim]
IA_MFWG()5:00AM has ended
11:10:29 [Zakim]
Attendees were +0329331aaaa, mhausenblas, Yves, Erik&Davy, +31.20.592.aabb, jackjansen, +329331aacc
11:10:33 [RRSAgent]
I have made the request to generate davy
11:10:56 [davy]
chair: erik
11:10:58 [RRSAgent]
I have made the request to generate davy
11:11:18 [silvia]
thanks everyone!
11:11:34 [silvia]
btw: we're only talking about quotes for track and id dimensions, right?
11:11:50 [davy]
11:12:42 [silvia]
davy: is your implementation using quotes?
11:13:04 [davy]
11:14:25 [erik]
rrsagent, draft minutes
11:14:25 [RRSAgent]
I have made the request to generate erik
11:14:32 [silvia]
ok - I don't think I mind either way actually
11:14:41 [erik]
rrsagent, please part
11:14:41 [RRSAgent]
I see 7 open action items saved in :
11:14:41 [RRSAgent]
ACTION: raphael to adjust the milestones in the charter [1]
11:14:41 [RRSAgent]
recorded in
11:14:41 [RRSAgent]
ACTION: troncy to adjust the milestones in the charter [2]
11:14:41 [RRSAgent]
recorded in
11:14:41 [RRSAgent]
ACTION: Jack to identify and add in corrib any missing test cases for temporal fragments [3]
11:14:41 [RRSAgent]
recorded in
11:14:41 [RRSAgent]
ACTION: mhausenb to add all MF WG members to corrib [4]
11:14:41 [RRSAgent]
recorded in
11:14:41 [RRSAgent]
ACTION: mhausenb to add a copy TC functionality [5]
11:14:41 [RRSAgent]
recorded in
11:14:41 [RRSAgent]
ACTION: mhausenb to come up with a fix for overview vs. edit single TC [6]
11:14:41 [RRSAgent]
recorded in
11:14:41 [RRSAgent]
ACTION: Erik to summarize the discussion on the quotes in a mail or on the wiki [7]
11:14:41 [RRSAgent]
recorded in
11:15:05 [davy]
indeed, me neither