# [SLIM] Minutes IETF 95

THURSDAY, April 7, 2016,
16:20 - 17:20 Afternoon Session II
Chairs: Bernard Aboba, Natasha Rooney
WG Datatracker Web Page: http://datatracker.ietf.org/wg/slim/charter/
Note Taker: Chris Newman

## Preliminaries

## draft-ietf-slim-multilang-content-00:

There was a call for volunteers to help test multipart/multilingual on more clients (will repeat on list). Realistically need to support both message/rfc822 and message/global.

Target update for about mid-May for updates to this draft.

## draft-ietf-slim-negotiating-human-language-01:

Randall: Changes to document were mostly clarifications. Declined to make changes that made the proposal more complex. Would really like to not add things. Some disagreement over how to interpret wording for lang attribute.

## draft-ietf-slim-use-cases-01:

Updated today with some cleanup including proper references. Needs some quick review. Randall has sent comments on the document. Brian agreed to review document and comment.

## ETSI Statement (Gunnar):
Reference to the importance of the SLIM work at ETSI.

* Q: If ETSI is publishing use cases, does the IETF need to publish a use case document?
* A: No reference to IETF use cases document by ETSI.

Gunnar sent mail about ETSI use cases that are not well covered.

Discussion:
* We have seen a number of grouping mechanisms in various ways, they're complicated and tend not to be implemented and don't seem to help as much as we thought they would.
* The simpler we keep the document, the more likely it is to be implemented. That makes it easier to extend it based on real needs.

## The Current "lang" attribute

Discussion: 
* The current lang attribute is defined to cover content. There was previous agreement in a face-to-face that we could not use the existing attribute.
* Order of importance has meaning for static media (e.g. newscast with multi-lingual content)
* Would really like us to try to separate requirements from mechanisms. If we decide we have a requirement to do more complex cases then we need a mechanism. If we decide we don't have such a requirement, then we can have a simpler mechanism. 

## Grouping preferences

Discussion:
* We don't think we have requirement to do grouping. Would prefer to start simple and go back later if we find we need the ability later.
* Can add grouping mechanism later.
* We also may not need order or grouping.
* Don't believe we solve enough for the necessary use cases. The simple use cases can be done with the 'lang' attribute. The new attribute should handle the complex use cases so it's different.
* Do we have a way of coming to consensus on which use cases are required vs. optional. Given that, can we reach consensus on if technology fits use cases.

Chair: please send uses cases to list to update use cases along this line.

Discussion:
* need to see if we have rough consensus on simpler mechanism earlier.
* extraneous media is not a problem.
* In addition to getting right media, also need right resources, such as an interpreter.
* that is not a signalling problem.

## Next Steps
Take this discussion to the list for those who cannot be present at IETF95 to comment.