[Fiware-oasc-etsi] CIM ISG ToR proposal

[Lindsay Frost]

Dear Juanjo, thank you for your email below, agreeing to the version I sent at Di 30.08.2016 20:39,
but reminding to discuss/add your paragraph re an open source implementation. I attach the docx.

Dear all,
Concerning Juanjo's paragraph below.  I am not confident that the mechanism
" until they provide a reference to a publicly available open source implementation of
their contribution" will assure the desired open-source result.

Examples of difficulties:

a)      I can well imagine that Company A might successfully contribute API
specifications  related to Topic A, but not be happy to reference/endorse a complete
implementation which they think does a bad job of Topic A ... but all the other CIM members
think is good enough (or the only running code maybe ;-).  That would mean that the Topic A
features would need to be removed from the final specification.

b)     I can well imagine that in a long API specification, and Data Models, it will be very
hard to identify which exact contributions came from Company A. There is a famous example
in 3GPP where the order of two bits was reversed in a radio configuring protocol and the
patent behind it ("higher efficiency for xyz ... by ordering of bits in coded radio messages")
became Standards Essential. What happens in the API if Company A proposes the feature A,
but Company B slightly re-orders the parameters ... is Company A remembered as the author,
or is Company B ? Probably you want both A and B to endorse the implementation. But that
ends with all contributors to the specification needing to endorse the (same) implementation.

Can anyone please suggest some other mechanism ?

Thank you
Lindsay
________________________________________

From: Juanjo Hierro [mailto:juanjose.hierro at telefonica.com]
Sent: Dienstag, 30. August 2016 21:07
To: Lindsay Frost; serge.raes at orange.com<mailto:serge.raes at orange.com>; Franck Le Gall
Cc: Ernoe Kovacs; Martin Bauer; Juanjo Hierro
Subject: Re: [Fiware-oasc-etsi] CIM ISG ToR proposal


Hi,

  I agree to take this as new draft.

  Point is that whether we also agree to include the paragraph I was proposing (again for your convenience):

Members and Participants of the ETSI ISG CIM agree that their contributions will not evolve from preliminary version of the Group Specifications up to stable versions of the Group Specifications until they provide a reference to a publicly available open source implementation of their contribution (developed by them or third parties) that can be used by applications developers.

  My intention is that this doesn't mean that referred open source reference implementation become "part" of the spec.   The intention is to a) clarify that we are going to work on specs that are validatd through implementation (i.e., part of the "driven by implementation" approach, rather than allowing people to arrive and come with ideas when their companies have no intent to implement anything or even to adhere to the implementation some open source initiative like FIWARE is developing) and MOSTLY b) somehow make it implicit that contributions to the specs will be at the end of the day granted by contributors as royalty-free, otherwise an open source implementation would not be viable.   Note that I was proposing to say "they provide a reference to a publicly available open source implementation" and I have deliverately avoided to say "they provide a reference to a publicly available open source reference implementation".   I made this on purpose to avoid that ETSI claims that if we are talking about "reference" implementations then we are talking about implementations somehow "bound" to the specs and therefore whose development should be governed also by the group (in practice meaning we have to go for an OSG).

  But that is a discussion I'm happy to take after consolidation of the proposed draft as new draft.  BTW, fine to carry out the discussion in the more general list.

  Cheers,

-- Juanjo

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.fiware.org/private/fiware-oasc-etsi/attachments/20160831/9aca8f1c/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 160830 draft of ETSI ISG proposal for cross-cutting Context Information ....docx
Type: application/vnd.openxmlformats-officedocument.wordprocessingml.document
Size: 659235 bytes
Desc: 160830 draft of ETSI ISG proposal for cross-cutting Context Information ....docx
URL: <https://lists.fiware.org/private/fiware-oasc-etsi/attachments/20160831/9aca8f1c/attachment.docx>