Adapt Learning Community: Mixing saved LMS output when updating the adapt package

Hi Adapters,

I have a small question that I wonder about.

When I have an Adapt package uploaded in an LMS (let's say Moodle) and people are busy following this course. During the period people are following this course, I receive some feedback which makes me to add a article or block or component to this course. When I re-upload this adapt package and people continue with this training package their earlier given answers to questions in the course are all mixed up or even opened again while the question had only 1 attempt.

When I look in Moodle I do see that al the previous given answers are still there.

Any suggestions to avoid this? Should I do something different when I wish to update a currently published course?

if you add any blocks to the course you need to ensure that they have new, unique, tracking ids.

You need to be particular careful modifying questions in a live course, this is because the code that restores the question data uses the question's position within a block to record where it is, so if you do something to change that then it will get very mixed up

we are thinking about how to improve this.

Show parent | Reply

Hi Matt,

We are building the adapt courses with the Builder (not possible to apply custom tracking ids right?) and when we look at the .js files in the build package we do see that tracking ids are unique.

Indeed we do not alter the questions in the blocks.

Still we see a mixing of returning content.

Hope for a solution soon! But in the meantime what would you suggest we could do to avoid this from happening? Of course we try to avoid having to update training packages, but still...

Show parent | Reply

So it's not changing the tracking ids between versions in any way? i.e. for any particular question you're having problems with, does its containing block still have the same tracking id as it did before and is it still in the same position within the block i.e. left/right/full?

Show parent | Reply

Hi Matt,

When you mentioned the tracking ID's for the blocks, I started to look at the json files. And what happened is that the new block was inserted in the beginning with tracking ID 1, all other ID increased 1. (By the way I work with Jay)

Show parent | Reply

Well that would certainly explain the behaviour.

Sounds like it's something that needs to be logged as an issue with the AT as inserting a new block shouldn't change the tracking IDs of existing blocks.

Show parent | Reply

Hi Matt, where do you think it should be logged? Github or Atlassian?

Github, I can do. Atlassian not (no account). Cheers, John

Show parent | Reply

Definitely github for the AT.

Maybe just drop a query into the chat for the AT first, just to check this is a bug that needs logging.

You can actually log tickets in (our) Atlassian Jira without having an account - but we only ever used this for the framework and plugins, and have recently decided to move back to using Github.

Show parent | Reply

Hi all,

The authoring tool (re)generates the block _trackingId values on every publish. I've raised a ticket in GitHub for this issue:
https://github.com/adaptlearning/adapt_authoring/issues/971

Regards,

Brian

Show parent | Reply