MathML schema must change, allow definitionURL on semantics

[sbmlsecretary]

SBML L2v3r1:

The MathML subset schema does not allow 'definitionURL' on the <semantics> element. If we follow through with making the changes for issue #1745160, we must change the MathML schema.

Note that this is potentially a serious problem: it changes the syntax of SBML. On the one hand, we should not be doing syntax changes in Releases (only in versions), but on the other hand, the current spec is inconsistent and it could be argued that the change is a correction.

Reported by: mhucka

Original Ticket: "sbml/sbml-specifications//105":https://sourceforge.net/p/sbml/sbml-specifications//105

[sbmlsecretary]

Logged In: YES user_id=862059 Originator: NO

I agree that the change should not be taken.

Original comment by: shoops

[sbmlsecretary]

Logged In: YES user_id=862059 Originator: NO

I view this a correction. To lessen the impact on older documents we should provide a default for definitionUrl that explains that the attribute was missing prior to L2V3R2

Original comment by: shoops

[sbmlsecretary]

Logged In: YES user_id=641982 Originator: NO

OK. Part of the realisation that we needed definitionURL on semantics was because of SBO. The MathML schemas for SBO and SBML NEED to be the same. As it stands you cant grab the math from SBO and be sure it will be valid in SBML!

I agree that definitionURL on semantics can be considered a correction - it probably should have always been there! But I have two thoughts:

1) Since libSBML is now very level/version specific its easy enough for me to let definitionURL on semantics be allowed in l2v3 and not in l2v2. However a change in the schema means that this is inconsistent - since the MathML schema is not level/version specific.

2) If we are going to change the MathML schema we should bring it completely in line with SBO; otherwise we will be discussing a similar issue again!!! However, that is a syntax change.

Original comment by: sarahkeating

[sbmlsecretary]

Logged In: YES user_id=343670 Originator: YES

I think the point #1 that Sarah is making is that if libSBML starts checking whether definitionURL is valid according to level/version, then it would be inconsistent with the MathML subset schema, which *wouldn't* be level/version specific. I am not sure what to do about that. The only thing that comes to mind that won't introduce more problems is just to ignore it. After all, the chances of anyone even having used 'definitionURL' on <semantics> is vanishingly small.

As for #2, I agree it's a good idea, but the practicalities of the situation are such that we won't be able to sort out additional MathML changes in time.

Original comment by: mhucka

[sbmlsecretary]

Logged In: YES user_id=862059 Originator: NO

I agree with the proposed change and that it should be done.

Original comment by: shoops

[sbmlsecretary]

Logged In: YES user_id=641982 Originator: NO

Mike is correct the chances of this being an issue for anyone is very small. [The fact that until libSBML-3 libsbml didnt support semantics at all and no-one ever noticed would imply that there arent many people using semantics with or without a definitionURL!]

However you could also argue that we dont change the schema and this release has the inconsistent spec; since if we change the schema we make all previous l2 specs inconsistent :-)

But since we encourage people to use the latest spec I wont argue that. Change the schema as a "correction".

My reservations about the mismatch between SBO and SBML Math still stands but I agree that for R2 thats not a feasible change.

Original comment by: sarahkeating

[sbmlsecretary]

Logged In: YES user_id=641982 Originator: NO

I agree with the proposed change and that it should be done.

Original comment by: sarahkeating

[sbmlsecretary]

Logged In: YES user_id=1045203 Originator: NO

I agree with the proposed change and that it should be done.

Original comment by: lenov

[sbmlsecretary]

• milestone: 678078 --> Accepted:_Changes_with_conformance_implications
• status: open --> closed

Original comment by: mhucka