Hi Alex:
A couple of comments inline ...
All the best, Ashok
On 5/23/2013 5:03 PM, Alex Heneveld wrote:
Thanks all for joining the call. Helpful to me. My notes from
memory, followed by slide links, followed by chat log.
- general agreement in Alex's very high level animation slides
[1,2] *except* for AT's directly referring to PCT's
- lots of discussion around AC v PC
- Alex to open an issue allowing AT's explicitly to refer to PCT's
and A's to PC's, so you don't have to always go through an AC(T)
(example a site which offers Wordpress blogs as PCT's -- ACT's
would just be posts and data, it's the PCT's which are
interesting)
To keep the model simple, perhaps all we need is to allow void
components
with requirements
- general agreement in Gil's slides though *except* for
requirements perhaps too simple
- discussion about clients being able to figure out valid matches
based on metadata: nice for this to be possible when practical,
but also may end up with ways the client can ask the server ...
not in scope for now but a likely early extension (perhaps
same/related to modifying Assembly Templates so that one is using
their favourite components)
- Gil okay with typed requirements; Alex happy, though a simple
type which just looks at capabilities and attributes and string
matching is a good thing
Could you give an example where typed components are useful? I
think this would help the discussion.
- people seem generally happy with DP as a bag of components each
with requirements
- nobody seems opposed to yaml anymore and some people strongly in
favour
My concern is not with YAML but with allowing both JSON and YAML.
Early on we decided that we would not support both XML and JSON due
to the added complexity
and picked JSON. Let's use the same argument here and pick either
YAML or JSON
- Alex to update proposal in line with latest thining
Best
Alex
[1] Alex slides Keynote -
https://docs.google.com/file/d/0B3XurVLRa7pIVGtNWFdCdzhZc0U/edit?usp=sharing
[2] Alex slides PDF -
https://docs.google.com/file/d/0B3XurVLRa7pIQVpTR3A3dG9ybXc/edit?usp=sharing
[3] in Gil's email sent 19.02 BST
Alex Heneveld (Cloudsoft)1: howdy all
Alex Heneveld (Cloudsoft)1: suggested agenda:
Alex Heneveld (Cloudsoft)1: - components and requirements
- pdp zip
- references and functions
- including files
Alex Heneveld (Cloudsoft)1: 0 - discuss where we have consensus and where not
Alex Heneveld (Cloudsoft)1: attendees - alex derek gil ashok tom krishna adrian
Alex Heneveld (Cloudsoft)1: (mark?)
Tom Rutt (Fujitsu): Mark: are you on the call?
Mark Carlson (DMTF): Sorry. Adrian Cockcroft stopped by just as I was joining the call. We were talking over old times at Sun.
Mark Carlson (DMTF): NetFlix has an open source PaaS offering, and I was recruiting him for CAMP...
Scribe: no worries . . getting Adrian onboard would be a major win
Scribe: drug him if necessary
Adrian Otto (Rackspace): more like a toolkit, and less like a problem from what I've seen.
Adrian Otto (Rackspace): s/problem/platform/
Mark Carlson (DMTF): Also hit up another old Sun buddy - Tim Bray, on joining CAMP. He did an OAuth2 tutorial here for a keynote. Seems to be Mr. Identity at Google these days.
Alex Heneveld (Cloudsoft)1: @Mark both would be major coups -- good luck
Scribe: if you could even get Tim to blog something half-favorable about CAMP - that would be something
Tom Rutt (Fujitsu): The current "model" is that an ACT is owned by an ACT, while the PCT is owned by the Platform. Currently the ACT can have usage pointers to PCTs, and the ACT has containment pointers to the ACTs. There is no current pointers from Assembly template to PCT
Tom Rutt (Fujitsu): can a PDP be used to instantiate Platform Component Templates in a Platform?
Mark Carlson (DMTF): @Tom - do you mean load the code for Platform Component?
Tom Rutt (Fujitsu): 5o clarify: can a PDP be used to create a PCT resource based on metadata in the PDP? I thought the Platform has preconfigured PCTs, which are referred to by ACTs?
Alex Heneveld (Cloudsoft)1: so PDP will typically cause creation of new ACT's which may reference may reference existing PCT's or my create newly created PCT's
Mark Carlson (DMTF): Seems like there would be an ACT and a PCR that specifies the Wordpress PCT
Scribe morphed into Gil Pilz (Oracle)
Krishna Raman (Red Hat)1: The multiple war scenario is probably why Tobias suggested that we need ACT's to specify PCT requirements
Krishna Raman (Red Hat)1 morphed into Krishna Raman (Red Hat)
Krishna Raman (Red Hat): Sry all, have to drop off.
Alex Heneveld (Cloudsoft)1: thanks all
|