Hello everyone,
Today's design call will be an informal continuation of the discussion on the direction for Data Services that we had from the earlier developer status call. A high level overview of the proposed direction is as follows:
> • We think "pure clean" data may be the wrong approach - we just need to make it easier for people to extra only what they need.
> - we believe getting closer to couchDB is the right approach (ie Agilix had mentioned they are not concerned with well formed... just wants everything they can recognize as DC conformsTo and an ASN, Jim did that with just CouchDB for AMPS plugin)
> • Current 'service' definition is currently enough.
> - if anything we need to invest some time in making it easy to not deploy indexes with services to limit resource usage for inactive node services.
> • Focus on making Map Reduce views and lists more accessible
> - build some libraries to make certain things easier (isXML, isTrustedIdentity, hasAlignmentData.... etc)
> - possibly support more than Javascript view server (Jim's Java Viewserver, and the existing python viewserver) with these libraries.
> - provide education materials for teaching the community how to design Map Reduce specific to the Learning Registry domain (pull by date, by resource locator, indexing strategies, etc)
> • Make native CouchDB replication possible from public node to encourage the benefits of standing up your own node.
> - need to look at removal of Admin Party to protect writes.
> - node operators who need security (like DoD) can use defined distribute instead to restrict access where needed.
The conference line number is:
> +1-800-605-5167
> 436796#
Jim Klo
Senior Software Engineer
Center for Software Engineering
SRI International
e. jim.
...@sri.com
p. 805.542.9330 x121
m. 805.286.1350
f. 805.546.2444
