CMI 5 Working Group Meeting Minutes (Washington, DC) – April 15 16th, 2015 - AICC/CMI-5_Spec_Current GitHub Wiki

CMI 5 Working Group Meeting Minutes (Washington, DC) – April 15-16th, 2015

Virtual Attendees

  • Chris Amyot - ICOM Productions
  • Jason Haag - ADL
  • Duncan Welder - RISC
  • Catherine Cormier - Halogen
  • Charles Touron - ATSC
  • Bob Hess - ATSC
  • John Dowell -
  • David Pesce - Exputo
  • Russell Duhon - Saltbox
  • James Sanderson - Autodesk
  • Julio Colon - Edu Carte
  • Simone Vidotto -
  • Algernon Corbett - Nova Southeast University
  • Andrew Downes -

On Site Attendees

  • Ray Lowery - Pratt & Whitney
  • Andy Johnson – ADL
  • Craig Wiggins - ADL
  • Art Werkenthin - RISC
  • Ben Clark - Rustici Software
  • Henry Ryng - INXSOL
  • Bill McDonald – Boeing
  • Jim Boyd – Electrical Training Alliance

(If anyone was missed, please let us know)

Meeting General

The cmi5 specification working group got together for a face-to-face meeting on April 15th & 16th at the ADL Colab Facility at Alexandria, VA (Washington, DC area).

The reason for having a physical meeting was to accelerate spec progress for the initial release of cmi5 (to be called “Sandstone”).

The meeting had two daily (2 hour) web conferences to update the larger community on specification progress and to solicit more feedback.

Weekly web conference meeting will also continue at the normal schedule resuming April 24th at 10:30am Eastern US Time. All are encouraged to continue participation thru this venue and the Github issues list.

Specification Changes

Satisfied Verb for Blocks & Course – As a result of discussions during the onsite meeting, it was agreed that a verb was needed to record complete/passed (i.e. “satisfied”) to the LRS for a Course and for individual blocks within a course. This addresses the “rollup” issue for recording at course or block level.The “Satisfied” verb will be used in a statement written by the LMS to the LRS when all AU’s (in a block or course) are satisfied based on their “moveOn” criterion. Block Id’s were also added to the course structure to support the recording of “Satisfied” statements.

Multiple “attempts” that record Passed - In the use cases discussion (and on the Github issues list) it was noted that the rule preventing multiple “Passed” statements from being recorded by a given in AU (in the same registration) caused an issue. In non-regulatory training it is common to allow the learner to “try again” to get a higher score (even though they previously “Passed” with a lower score). To address this use case, the course structure will include a “passIsFinal” attribute for each AU. The LMS will write this attribute to the StateAPI prior to AU launch so the AU can determine if it allowed to issue multiple Passed statements.

Remove verbs - Suspend & Resume– It was determined during the meeting that “Suspend” & “Resume” verbs were unnecessary given the function of the State API and the other statements recorded during an AU session.

Remove OAuth– Given some of the issues and uncertainty surrounding OAuth, it was determined that cmi5 will only use “basic authentication” (using the Fetch URL) for the initial release. Future releases will include other authentication methods (including OAuth).

Remove Activity Types – I was determined that ActivityTypes were too domain-specific to be included in cmi5 statements. There were removed from the runtime specification.

Requirement for reporting ALL data recorded – Conformance language was added to ensure that the LMS would provide a means for providing access/reporting ALL data recorded by the AU and LMS. Access to the data would be limited by LMS access control policies. Access to all data recorded by AU’s (even non-cmi5 statements) is key cmi5 goal to support extensibility.

Branding

The proper spelling and capitalization for CMI-5 is now “cmi5” (all lower case – no dash) The proposed logo concept for cmi5 will have an xAPI “theme”(using the puzzle piece symbol). Draft of this design will be shared later in the weekly web conferences.

Version Identification

cmi5 versions will named after geologic rocks. The first release will be “Sandstone”. Major changes to the specification that affect interoperability will result in a new “rock” name. Minor changes to the specification will include numeric serialization following the “rock” name (e.g. “Sandstone 3rd edition”).

Product Road Map

“cmi5 Sandstone 1st edition” is planned to be release in early May 2015. Sandstone is intended to be a release for developers to validate with working implementations/prototypes. It is planned that the next major release of cmi5 will be Sept 2015 (“rock name” to be determined). This release is expected to be usable for fully functional products (in cmi5 compliant LMS systems, content creation tools, and content).