Martin Dougiamas

I'm really sure it's useful or nice to name platforms specific here when you are really talking the strengths/weaknesses of LTI vs other methods for distributing content.

Moodle can handle all use cases here and so it has ultimate flexibility in this regard:

• It can have native content uploaded into it in a number of formats (including standards like IMS CP and SCORM).
  
• It can publish it's own content (both as individual activities or whole courses) using LTI standards.
  
• It can consume content published using LTI standards.
  
• It can consume content using any proprietary method (via plugins).

Obviously a course designer needs to choose among these when building a course.

Before we can talk about strengths/weaknesses of these, though, we need to define better what the OER content is:  Is it just single documents?  Is it a whole course (multiple resources/activities)?  Is it interactive?   Is it collaborative?  Does it allow grading?   I think there could be a nice matrix made of all this.

In general though, some of the disadvantages of LTI-based resources are that:

• They don't work well in offline cases, unless some way to copy/cache them is provided.
• They are generally not customisable/localisable/translatable by the teacher who is using them.
• They increase the possible points of network/server failure that can affect final course experience.
• They can be a little more difficult to set up and understand.
• They have some privacy issues with student data being sent to different servers.

Some of the advantages are:

• LTI resources can be maintained centrally but used equally on multiple LMSes 
• LTI resources lend themselves to business models based on subscription.
• LTI resources encourage standardisation (in the content itself as well as the format).

Totally agree there should be a whole bank of default scales.  Remember they need to be international and translated too.

Scales with percentages kinda crosses into letter grades for me, they need to be rationalised.

This scale is very old - in fact it was there from the first version of the forum.

While it was excellent in my original Moodle courses, and I do hope some people do think about using it I very much doubt that many do, simply because it's too far from their previous experience and it won't fit into their paradigms.

In any case I never had any particular wish to enforce it as a default scale - it just kinda hung around because no-one got around to coming up with a better one.

I would love it if those of you who care about these things came up with some better ideas and made an MUA project out of it.

Just, please, think about things that have some good solid proactive educational principles behind them and don't go for lowest common denominator stuff like Amazon stars or Facebook likes.

Yes that's correct.  Basically competencies are things that belongs to the learner, not the course, and they should be consistent that way.

Another option for the use case Mary mentioned, however, is to simply use the existing Outcomes system that is already there in core (and will remain as an optional thing).  Ultimately Outcomes end up in the gradebook for the course.

That's the plan, yes.   

But first we need a kickass assignment solution as that's where most serious manual grading takes place.