On 2022-03-11 18:16, Hans Hagen via ntg-context wrote:
> Hi,
>
> It has been a while since there was an updaste but there it is. The 
> most significant 'changes' are:
>
> - more automatic adaption to double sided page properties in metafun 
> (lmtx) ... blame sreeram and me for messing up
>
> - more extensibve support for indic fonts which sreeram has documented 
> on the wiki (fonts, translutaration etc)
>
> and
>
> - improved math rendering to be documented by by mikael in an upcoming 
> manual ... it tooks a couple of months so it is supposed to have some 
> impact.
>
> before we move on the the structural compoments and see what can be 
> done beter there, we decided to release the first stage that deals with:
>
> = fixing font properties; we don't expect fonts to be improved so 
> that's the route we follow
>
> = extensive inter atom spacing (we have many more math classes, some 
> in the engine, some in context) and everything can be controlled (the 
> old pairwise spacing primitives have been replaced by a more general 
> mechanism, including support for fast setup of categories and 
> inhertance from other classes)
>
> = idem for inter class penalties, including granular control over line 
> breaks using a system of (in/de)creasing penalties
>
> = auto-fenced inline material across lines and linebreaks in molecules 
> (read: complex atoms)
>
> = taking stretch and shrink in encapsulated atoms into account in 
> linbreaks (using some new box related mechanism that is also available 
> in text and will be explained later)
>
> = spacing at the edges that should not be there but always was there 
> as side effect of the engine is now removed; as with other new 
> features this can be controlled (we have rulesets for atoms to which 
> we can add more in the future)
>
> = more detailed control over (skewed) fractions, (flattened) accents, 
> (sized) fences, radicals ... there was already a lot there but it now 
> combines with mentioned spacing (when relevant the context macros have 
> been adapted to control this which is a bit cleaner)
>
> = we have a new built in muskip (tiny) but as we have an inheritance 
> system as well as binding to variables one can set up extensive 
> spacing models register and constant binding; the currently set values 
> are based on math typesetting references; in the future we might have 
> additional ones
>
> = although we did handle primes already rather transparent in mkiv / 
> lmtx, inconsistency in fonts made us decide to make primes first class 
> properties of atoms and molecules which means that we now have: pre 
> super/subscripts, post super/subscripts and primes; this permits nicer 
> anchoring under parameter control
>
> = there are more font parameters to control matters (they are not in 
> open type but we sort of assume they are); the options that we had to 
> control how the engine deals with fonts are now driving the process 
> (without them being set not much is applied) and we basically assume 
> open type fonts (where one can enable engine options for those kind of 
> fonts but also for traditional); some experimental options for testing 
> are gone
>
> = in the process tracing for what is done in math by the engine has 
> been extended (mostly for ourselves)
>
> = instead of auto patching math fonts most is now under control of the 
> goodie files; user can (as usual) make their own, but also ask Mikael 
> to add missing bits and pieces (wrt dimensions)
>
> = most known math fonts are dealt with but there are a few exceptions: 
> we no longer consider asana to be useful (it doesn't pass our qa and 
> needs too much patching); erewhon is also troublesome (dimensions, 
> missing shaped) as is concrete (idem) so we decided not to waste too 
> much time on them now
>
> = we probably need to freeze math fonts in the distribution and only 
> update when we know what has been changed
>
> = we got rid of quite some artifacts that normally go unnoticed (like 
> tiny added spaces that result from always applied spacing even if not 
> needed and/or to compensate for glyph features like too much or little 
> side bearing): by fixing inconsistencies in fonts we could get rid of 
> all and therefore inter atom spacing is what does the job; once you 
> know it, you see it and there is no going back
>
> = because we zoom in we notice these things but much went unnoticed 
> (or was just accepted) for years; it means that we think some of the 
> things we did qualify as math micro typography (like it or not) and it 
> does mean that documents can render differently (unlikely for display, 
> propably also for inline unless you have plenty of math in a paragraph)
>
> == our objective has been to make it easier for users and to get rid 
> of explicit spacing; for that we also introduced commands for 
> differentials (that have their own class), imaginary numbers (also 
> have their class) and we added exponential as class; fwiw, digits and 
> factorial are classes too so again spacing has been optimized for them
>
> = whatever i forgot in the meantime mikael will add to this list
>
> There is a rough description in the 'ontarget' document. We hope you 
> enjoy it, at least we did (and so, as we're not finished yet).
>
> Hans
>
> -----------------------------------------------------------------
>                                           Hans Hagen | PRAGMA ADE
>               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
>        tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
> -----------------------------------------------------------------
> ___________________________________________________________________________________ 
>
> If your question is of interest to others as well, please add an entry 
> to the Wiki!
>
> maillist : ntg-context@ntg.nl / 
> http://www.ntg.nl/mailman/listinfo/ntg-context
> webpage  : http://www.pragma-ade.nl / http://context.aanhet.net
> archive  : https://bitbucket.org/phg/context-mirror/commits/
> wiki     : http://contextgarden.net
> ___________________________________________________________________________________ 
>


With regard to math fonts, I notice that the goodies file for Libertinus 
expects version 6.8 of the font, but the current version (since February 
2021) is 7.040, and much work appears to have gone into the math font.

Two questions, then:

 1. Is Libertinus Math considered a "useful" math font?
 2. How does the goodies mechanism work when the expected version
    differs from the found version?

(I understand that keeping up with evolving fonts is somewhere between 
difficult and impossible.)

I also note that the typescript for libertinus is quite old and does not 
reflect the semibold variants that have been available for some time. I 
prefer them to the bold variants, and my own typescript for libertinus 
does incorporate them. I am happy to share it if anyone wants it.

-- 
Rik