IETF 64 Vancouver, Canada
RMT WG Minutes

Many thanks to Magnus Westerlund for taking these minutes.

Off the Agenda
--------------

    Allison asked what was happening with SDP flute draft.
    Magnus: It is being updated and we could expect an update the next few days.

Status update
-------------

See presentation slides.

Moving to Standard Track:

Brian Adamson: it would be appropriate to move TMFRC to Standard
Track as well, as NORM can use a normative reference to it.

TESLA:

Vincent Roca: implementation is progressing, but no draft has been
submitted yet. A few more weeks than it should be available.
Lorenzo: TESLA not on the charter need to start discussion on that
topic. So please take this to the list.

Moving NORM to Standard Track -- Brian Adamson
-----------------------------------------------

   draft-ietf-rmt-pi-norm-revised-00
   draft-ietf-rmt-bb-norm-revised-00

See presentation slides.

Lorenzo: Changes section should be included to help reviewers with
noticing what has been changed and reduce the workload of review.
Allison: This are the changes and why it should go forward. Section
should be called "Changes since RFC XXXX". Some documents need to go
together so there are no meaning to send in documents until the
dependency are ready.

Consensus on the above topic was the following:

   - all submissions transforming an Experimental into a Standard Tack
     RFC should contain the following:

     * a section "Changes since RFCXXXX"
     * A common note explaining the rationale for moving to Standard
       Track (this note will be the same in all document a will be
       drafted soon by the WG).

   - Dependency order should be enforced in the submission to the IESG
     (submit dependent documents later).


Mark: On the EXT_FTI it is defined by the FEC scheme. Updated FLUTE
has changed.  Brian: As NORM is not a ALC protocol it defines it in a
similar way but do not reference it as it

Mark: ALC/LCT registers its own IANA registry, should NORM use the same
registry?
Lorenzo: Yes, two different registries should be defined.
Allison: Write clear instructions to ensure that they are correctly
placed in the IANA registry so these ends up under the RMT umbrella.

Consensus on header extensions:

   - establish 2 IANA registries (one for ALC/LCT and one for NORM).
   - Write clear instructions to accomplish this in both ALC/LCT and
     NORM, choosing appropriate (hierarchical ?) names for the
     registries.
   - all documents that allocate header extensions (for ALC, NORM or
     both) must contain specific IANA instructions to do so in the
     "IANA" Section.
   

FLUTE Update -- Stephan Wenger
------------------------------

  draft-ietf-rmt-flute-revised-00.txt

See presentation slides.

Stephan: Changes slides needs to go into the draft.
Magnus: Has anyone reviewed the URN used to identify the FDT XML schema.
Allison(side note) Needs to go to URN mailing list.
Mark: Should reference the FEC BB not the basic FF schemes.
Allison: Media type and URN review can happen in parallel in with the
WGLC.  Send notice to both those mailing lists that can be found under
non-WG mailing list at IETF page.

FEC BB -- Mark Watson
---------------------

  draft-ietf-rmt-fec-bb-revised-02
  draft-ietf-rmt-bb-fec-basic-schemes-revised-01
  draft-ietf-rmt-bb-fec-raptor-object-03

See presentation slides.

Lorenzo: Is there a changes section?
Mark: IS it ready for going to IESG.
Lorenzo: A new WGLC does not seem necessary. Revise and declare
intention to submit.
Allison: If this doesn't have dependency it is fine to submit it. Only
drafts that has dependency shouldn't be submitted until all documents
are ready and then sent to IESG as a bunch.

basic-schemes

Lorenzo: FEC-raptor-object will be sent to IESG as soon as the BB is ready.

LCT Revised -- Mark Watson
--------------------------

  draft-ietf-rmt-bb-lct-revised-01
  draft-ietf-rmt-pi-alc-revised-01

See presentation slides.

Mark: No comments was received on proposal to RMT and has been
implemented. So if you do not agree to the changes then speak up.
Lorenzo: IANA notes in FLUTE, ALC needs to have registration of their
header extensions.  Brian: Then NORM should also define its header
extension registry.

Allison: IETF Review (replace IETF consensus) is a good level. As this
forces review within IETF.
Mark: Sounds reasonable
Magnus: I think IETF review might be reasonable for header extensions.
But what about fully qualified FEC schemes and there requirements?
Allison: There are a lot of options on how IANA can handle these
things.

Long discussion on alternatives. Allison providing some alternatives
and Magnus arguing for greater flexibility for IETF external
organizations to register things. Some alternatives to IETF
specification required was: IETF expert to approve specification
required as complement IESG approval of specification required

Lorenzo: Alternatives
1. IETF consensus
2. IETF consensus with bypass with publication required and expert review.
3. Something else

Lorenzo choice between the first two alternative will go for a 1 week
WGLC.

Mark: Proposal for IANA rules for LCT header extension: IETF publication
required.  No comments.

Mark: Please report if LCT changes of sender current time field would
cause problem.  LCT draft is not available due to errors with the
publication.

TMFCC -- Mark Pullen
--------------------

  draft-ietf-rmt-bb-tfmcc-05.txt

See presentation slides.

Mark Pullen (MarkP)

Lorenzo: Will a new draft be resubmitted?
MarkP: Yes, as soon as deadline for receiving comments has been passed.

LDPC -- Vincent Roca
--------------------

  draft-ietf-rmt-bb-fec-ldpc-00.txt

See presentation slides.

Mark: Do you mix source and repair symbols?
Vincent: No, they will be of only one type.
Lorenzo: IS this ready for WGLC?
Vincent: No, next version
Mark: To achieve good performance it most likely that you need to send
source and repair symbol packets in random order. Draft should
document the consideration between sending source symbols first, or
sending them in random order.
Mark: Please remove statement in relation to IPR that are not the
standard IETF boiler plate.
Lorenzo: Mark is probably correct, will double-check.

RS fully specified code -- Vincent Roca
---------------------------------------

  draft-lacan-rmt-fec-bb-rs-00.txt

See presentation slides.

Lorenzo: Thanks very much for doing this overdue piece of work. Need
to formally accept this as WG item. Vincent, please send a request to
the WG mailing list.


File Aggregation -- Vincent Roca
--------------------------------

  draft-neumann-rmt-flute-file-aggregation-02.txt

See presentation slides.

Mark: TOI question. FLUTE revised has no structure for TOI. Should all
text be removed about the TOI in this proposal?
Vincent: Yes, seems reasonable.

Joerg: Running a FEC scheme over for example a website where certain
parts need to change is counter productive.

Lorenzo: A Question is if this should be a WG ITem. IT is a bit off charter.
Mark: It is useful and should be supported.
Joerg: Would you like to get other gains like compression, by doing zip on the
files.
MArk: The question do you need to expose for the FDT layer that these are
separate files?
Lorenzo: Please send mail to the list discussing these issues.

Timestamp LCT header extension -- Stephan Wenger
------------------------------------------------

  draft-jansky-alc-timestamp-extension-00.txt

Lorenzo: Why a timestamp and not a long lived identifier?
Vincent: What is the goal for this document. Why restricted to a specific
protocol and the chosen.
Mark: LCT does not have a mechanism to indicate changes, however FLUTE
does. So you would only have to listen until the next FDT repetition
comes around.  Lorenzo: This discussion should be included in the
draft. This could apply to norm.
Brain: Norm has an instance id that can indicate changes. But it is
limited and will wrap.
Mark: The LCT header registry is an example.  And until that is
published we live in the current world. And the WG can publish things
without registry request.

Lorenzo: please start a thread on the list to decide if this will be a
WG document.