use the lmfdb/knowls inline display technology instead of popups

[kohlhase]

In many cases, we use popups in MathHub, e.g. for "goto definition" (when it works). http://lmfdb.org has a much nicer display technology: in-place expansions (they are even recursive). Go for instance to http://www.lmfdb.org/knowledge/show/ec.additive_reduction and click on "number field", and then in the extension on "label". Isn't this nice?

We should steal (learn from and integrate) from this to make MathHub nicer. I am not sure where this actually belongs, MathHub? JOBAD? MMT? but someone can tell us.

cc: @tkw1536

[kohlhase]

I really like the little "permalink" feature too, it is very non-intrusive and obvious what it means.

[akohlhase]

This would indeed be a push towards good usability!

[kohlhase]

@johncremona told me that the details about how the system works are at http://aimath.org/knowlepedia/

[m-iancu]

This looks like it's a very interesting library for us since we (want to) use transclusions a lot. But we have to figure out how this integrates with JOBAD which is our de-facto library interactive documents. Do we integrate it as a JOBAD module, or do we use it to override the left-click event of JOBAD (will that even work ?)

In any case we should make it work, it's much better than the module we use now.

[JohnCremona]

http://www.lmfdb.org/knowledge/history

will show you all the most recently edited knowls in order of time. It seems to have a fixed number of entries, but if that was a difficulty then I'm sure it could be changed. In fact I know it could be since I just looked at the code ;)

John

On 27 January 2016 at 11:53, m-iancu notifications@github.com wrote:

This looks like it's a very interesting library for us since we (want to) use transclusions a lot. But we have to figure out how this integrates with JOBAD which is our de-facto library interactive documents. Do we integrate it as a JOBAD module, or do we use it to override the left-click event of JOBAD (will that even work ?)

In any case we should make it work, it's much better than the module we use now.

— Reply to this email directly or view it on GitHub https://github.com/KWARC/MathHub/issues/188#issuecomment-175581253.

[kohlhase]

In any case, if we can use the knowls library, then we can generate knowls from the glossary and these could then be referenced from lmfdb seamlessly.

[JohnCremona]

Michael, the knowls I have edited or created today, following our work on the MMT versions, are as follows. I hope you can extract the necessary text from these. Some of these are only relevant over Q (e.g. Rouse), some only over number fields (e.g. Obstruction class).

Label for an elliptic curve over ℚ

Elliptic curve over a field

Tate module of an elliptic curve

Minimal discriminant

Isomorphism of elliptic curves

Isomorphism between Weierstrass models

Endomorphism of an elliptic curve

Obstruction class of an elliptic curve

Local minimal discriminant of an elliptic curve

Global minimal model

Integral model

Endomorphism ring

Algebraic closure of a field

Absolute Galois group

Isogeny between elliptic curves

Conductor of an elliptic curve over ℚ

Rouse classification Rouse label

On 27 January 2016 at 13:14, Michael Kohlhase notifications@github.com wrote:

In any case, if we can use the knowls library, then we can generate knowls from the glossary and these could then be referenced from lmfdb seamlessly.

— Reply to this email directly or view it on GitHub https://github.com/KWARC/MathHub/issues/188#issuecomment-175617954.