Missing documentation for Tensor class?

[ernesto...@gmail.com]

Hi Everyone,

I am not 100% certain of this, but it looks like there is some documentation missing from the Tensor class for v 9.5.0 - the first entry in the contraction section is for the l1-norm, although some of the text appears to be for contractions in general. This documentation does, however, exist in the documentation for 9.4.0. I had a look in the change log and couldn't see anything about the methods being deprecated, so is it possible there is something wrong with the docs?

While I am here: The change log for 8.4.0 marks double_contract for Tensors as being deprecated, but the method is in the documentation for 9.4.0 and it is not marked as deprecated. Was the function reinstated?

Regards,

Ernesto

[Bruno Turcksin]

Ernesto,

Yes, something is wrong with documentation. I don't know why the functions don't show up.

I am not sure what happened with double_contract being deprecated. Maybe it was a different version maybe the function was reinstated.

Best,

Bruno

[ernesto...@gmail.com]

Hi Bruno,

Thank you for confirming my suspicions. I had a look at the header file and all seems to be in order there. Please let me know if there is a way I can help fix things for a future doc update. I am, obviously, okay with using documentation from the last release myself.

Regards,

Ernesto

[Wolfgang Bangerth]

On 12/21/23 23:48, ernesto...@gmail.com wrote:
>
> Thank you for confirming my suspicions. I had a look at the header file

> <https://github.com/dealii/dealii/blob/0da331b3df2ccd5f39b93fe824dd3c9c3a419f7f/include/deal.II/base/tensor.h#L2273> and all seems to be in order there. Please let me know if there is a way I can help fix things for a future doc update. I am, obviously, okay with using documentation from the last release myself.

Ernesto:
I took another look at the online documentation, hoping that Matthias's recent
update to doxygen 1.9.8 helped. But it did not.

We have over the years looked at this and related issues many times and never
really managed to address them in a way that would make doxygen either give us
the complete documentation, or at least provide us with warnings we could
address. I know it is not a satisfactory answer, but I do not know what else
to do than just say "we don't know how to fix this".

Best
W.

--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/