Closed zepumph closed 3 years ago
I did a large refactor. I don't feel like this needs to end up in bad text or anything, because I feel like the convention will be followed. @jessegreenberg do you feel differently? Anything else here?
I believe that this rename may have created backwards-compatibility issues with release branches. I just did a checkout of the 1.2 branch of atomic-interactions, tried to load phetmarks, and got the following error:
The full path for the abbreviated part in the screenshot is:
http://localhost:8080/perennial/data/interactive-descriptions
It can't find it because the file has been renamed. I think this rename may need to be reconsidered or some way of supporting backwards compatibility needs to be worked out.
BTW, I removed the ready-for-review label.
I labeled this as "high-priority" because it's messing with my ability to test release branches, and I assume others are in the same boat or will be shortly.
Two things to note:
url: '../perennial/data/interactive-descriptions'
to url: '../perennial/data/interactive-description'
(just drop the 's') in phetmarks.js
Overall the rename looks good to me.
For the compatibility issue with old phetmarks I am not sure. "Interactive Description" is being used publicly now on the website to brand this set of features so I think we should keep the rename.
The file rename in phetmarks could be relatively painless as maintenance releases go, maybe we should commit to that. @zepumph what do you think, do you see another way?
@zepumph and I discussed - @zepumph had a thought that the fixing commit for this could be applied to all release branches without actually publishing them. This would mean that we would go all the way up to updateDependencies
in the MR process then leave commits unpublished.. Unfortunatley, these unpubilshed commit would then get flagged for the next batch of maintenance releases when someone runs checkBranchStatus
. Is there a way to make a commit to a release branch and flag it as acceptable?
Otherwise, we agreed that a batch of maintenance releases seems like the best way forward. I will check in with @jonathanolson about this.
Unfortunatley, these unpubilshed commit would then get flagged for the next batch of maintenance releases when someone runs checkBranchStatus. Is there a way to make a commit to a release branch and flag it as acceptable?
Please feel free to do this! I'm not concerned about that at all.
OK thanks @jonathanolson - I will go ahead with this then.
Dependencies have been updated for release branches in https://github.com/phetsims/phetmarks/issues/48. The only exception a phet-io specific release branch of molarity which will be taken care of separately in that issue.
I verified that phetmarks is launchable from atomic-interactions 1.2 SHAs again. @jbphet reassigning to you to verify if you wish, otherwise I think this can be closed.
I tested this by checkout out the 1.2 branch of atomic-interactions and testing phetmarks, and it worked fine. Thanks for the fix @jessegreenberg and @zepumph. Closing.
I hear @terracoda over in https://github.com/phetsims/a11y-research/issues/145#issuecomment-614259137. We should make sure to fix this across the project
Currently there are 23 usages of
interactiveDescriptions
, and 19 usages of "interactive descriptions" (case insensitive).@jessegreenberg whoever get's this this first.