Exhaustivity error

[nkalivoda]

When I hit "get results" after checking the constraint (not Gen condition) Exhaustivity on the interface, I get this message:

This is presumably because it doesn't have a category argument. We need to either (a) add category options or (b) make sure it just works as a category-insensitive constraint.

It is currently defined in the info as (b), and I think we should stick to that, because I can think of at least two versions that could take a category argument (Exhaustivity-parent or Exhaustivity-child, basically analogous to the different types of NonRec).

[jbellik]

Have you pulled and rebuilt? We discovered this bug last week and Max put in a fix for it. I can't replicate this issue on the current version of the master branch.

This did make me realize that we don't have those other versions of Exhaustivity on the interface; I think we have them implemented on the back end so I should look into that.

On Mon, Sep 21, 2020 at 1:06 AM Nick Kalivoda notifications@github.com wrote:

When I hit "get results" after checking the constraint (not Gen condition) Exhaustivity on the interface, I get this message:

[image: Screen Shot 2020-09-21 at 10 02 45 AM] https://user-images.githubusercontent.com/12573882/93744214-c3ddbb00-fbf1-11ea-8130-2942e9a2342d.png

This is presumably because it doesn't have a category argument. We need to either (a) add category options or (b) make sure it just works as a category-insensitive constraint.

It is currently defined in the info as (b), and I think we should stick to that, because I can think of at least two versions that could take a category argument (Exhaustivity-parent or Exhaustivity-child, basically analogous to the different types of NonRec).

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/syntax-prosody-ot/main/issues/464, or unsubscribe https://github.com/notifications/unsubscribe-auth/AC7ROKT6L2JH7K7GZFKXPPLSG4CQRANCNFSM4RUHH2IQ .

--

-- Jennifer Bellik, PhD Post-doctoral researcher & lecturer UC Santa Cruz https://people.ucsc.edu/~jbellik/

[nkalivoda]

Ah, good that it's already been fixed. I found while using the interface on your website.

And yes, it would be great to put the other versions up. Can't remember exactly what we did before, but I'm thinking:

1. Exhaustivity-parent (category-insensitive) Assign a violation for every node of level k that has a child of level n<k-1.

2. Exhaustivity-parent(pCat) For every node pCat, pCat at level k of the prosodic hierarchy, assign a violation if pCat has a child of level n<k-1.

3. Exhaustivity-child (category-insensitive) Assign a violation for every node of level k that has a parent of level n>k+1.

4. Exhaustivity-child(pCat) For every node pCat, pCat at level k of the prosodic hierarchy, assign a violation if pCat has a parent of level n>k+1.

[jbellik]

I just updated the interface in master to have the category-insensitive Exhaustivity parent. We already have category-free Exhaustivity child. I agree it would be godo to add the category-sensitive versions.