Michal Vasko

Yes. In short, `OPTICAL_CHANNEL` identityref is missing its prefix.

Everything relevant is mentioned in the [specs](https://www.rfc-editor.org/rfc/rfc7950.html#section-9.10.3). I am sorry but I am not here to explain the basics of YANG.

It seems you will have to merge current `devel` manually first for the ASAN and UBSAN test to pass. But the TSAN test failure will be more difficult to fix...

We have no documentation to explain the locks, you will just have to ask me. But I am not sure what exactly you are referring to. The thread that is...

> It looks that all locking goes through _sr_rwlock() or sr_rwunlock(), and that sr_rwlock_reader_add and sr_rwlock_reader_del are never called directly. Right? Then whenever the readers list is accessed, the r_mutex...

> It seems that the identityref derived-identity2-delete is evaluated in the namespace of the deviating (source) module instead of the deviated (target) module. What makes you think it should work...

It seems almost none of these short character classes are supported in the way XML Schema defines them as their meaning differs from Perl regexes (which is what we mostly...

> So the recommendation should be to replace these by [0-9] and [...a-zA-Z0-9]. That is exactly the Perl semantics of these. So what exactly is the problem and what would...

> YANG defines patterns differently from what you implement. Well, I would hope it all works except for these character classes. > So instead of saying that YANG and W3C...

No problem, warning added.