New numbering fallback messages

[Oscar Levin]

I'm seeing some new "Fallback" messages when building a project that has its divisions level set to 0 (and I imagine also 1).  For example:

* PTX:FALLBACK:   numbering level set for open problems ("2") is greater than the maximum possible levels ("0") configured.  The default value will be used instead

However, I don't have anything set for open problems (this is from a publication file that predates open problems being able to get a different number than blocks).  

I understand this this is just an information message, but I worry it will be confusing to a publisher since it looks like a message telling them they made a mistake setting numbering, but then unable to find where they made that mistake.  And why would we not use the default value when nothing is set?

Could we let the publisher variables xsl set the the default value of the element-specific number levels to be the minimum of 2 and the value of divisions level?  It seems like this is the "default" that the message is referring to anyway.  Then only output the warning if the user actually has set those values too high.

[Rob Beezer]

Thanks, Oscar. You have set me off on a quest.

To make numbers (now in assembly, late), we need to know the publisher variables
for this.

To make the publisher variables we need to know the defaults.

Defaults are based on "document type" (does a book have parts?).

The document structure can't really be known until we have resolved version
support (in -assembly, early).

Your requested fix is not too much work, but investigation has led to a lot of
consolidation and safety work around this delicate situation. Likely ready soon.

Rob

> --
> You received this message because you are subscribed to the Google Groups
> "PreTeXt development" group.
> To unsubscribe from this group and stop receiving emails from it, send an email
> to pretext-dev...@googlegroups.com <mailto:pretext-
> dev+uns...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/pretext-
> dev/3d0c56bb-3a31-40d9-84fc-0858e54e1241n%40googlegroups.com <https://
> groups.google.com/d/msgid/pretext-
> dev/3d0c56bb-3a31-40d9-84fc-0858e54e1241n%40googlegroups.com?
> utm_medium=email&utm_source=footer>.

[Rob Beezer]

OK, this was a latent problem, since about 2021. It is only being exposed now
that we have three new famlies of "counters", such as the possibility of running
open problems on a "distinct" numbering scheme. So old publication files were
not set up to squelch this unanticipated behavior.

Worse, the message said the default value was being used, and it was - *even if
it was wrong* in the given situation. Again, somewhat disguised if explicit
values were set in the publisher file, consonant with the (low) division level.
So worse than a cryptic message, also buggy behavior.

When there is a low overall division numbering, the defaults for various
numbering schemes are computed, but if higher, then clipped to the lower value
for divisions.

Authors will get a FALLBACK message when their publisher file does bad things,
but when the publisher file is silent for a family of numbered objects, there is
no message about how defaults are adjusted to a (low) divisions level.

https://github.com/PreTeXtBook/pretext/pull/3074

Thanks very much, Oscar, for the report. Lots of good maintenance work as a result.

Rob