Open anermina opened 7 months ago
@anermina I think the problem here is that the clause reference should be "Clause 8, second paragraph" by default, not "Clause 5".
As you know, we do (kind of) have page references functionality, which we have introduced to enable page references for indexes, and we have used that for BIPM.
https://github.com/metanorma/metanorma-bipm/issues/67
This functionality exists in the backend, but has quite deliberately NOT been exposed to end-users until now, because we have not wanted to encourage page-only cross-references: they are bad practice. We do that only for generating ISO and BIPM index entries, and we expose it ONLY in BIPM.
But a page reference in PDF, replacing any built in referencing, is enabled by adding the attribute pagenumber=true
to the xref
tag for the cross-reference. In BIPM, this is done as:
https://www.metanorma.org/author/bipm/topics/markup/#cross-references
<<{{anchor}},pagenumber%>>
I am not encouraging us to spread this functionality to the rest of Metanorma, especially for an edge case like this. Cross-references should always be to clauses, for the very simple reason that page references change, and are print-bound (which is why HTML and DOC completely ignore the tag). We should not be encouraging users to behave as if this is 1950.
ISO Rice Amendment (2023) cross-references a page within the document.
Should we support cross-references in the reviewer notes? PDF currently does not add a hyperlink that can be clicked.
Considering this cross-reference is actually added to point to a modified clause, and that clause gets generated within the same page in Metanorma, I found it more appropriate to use a cross-reference to the amendment's section the comment points to, i.e. to use the following markup ("Page 2 of this Amendment" is omitted):
However, this markup leads to generating a clause number (if not specified differently within the anchor), while the clause numbers are "hidden" in amendments.
Therefore, I think the following markup would be the best to use (and is currently applied), but I am not sure whether it satisfies ISO's requirements.
Please let me know what you think would be the appropriate way to encode cross-references like this one.