Re: Naming RFC: Properties, property sets, presets

[Wols Lists]

On 13/07/20 10:29, Urs Liska wrote:
> Am Montag, den 13.07.2020, 19:02 +1000 schrieb Andrew Bernard:
>> Property subset is good.
>>
> 
> In a way, yes, namely by keeping within the semantic domain.
> However, while a property set actually defines the available properties
> (along with predicates and default values) the "preset" is concerned
> with *values* for a subset of the property set.

Does the "property set" define the default values? In which case, the
property subset also defines the default values for the items it contains.

We're basically thinking along the systemd lines, aren't we, where the
distro defines the default values in a default setup directory, and then
the admin defines the over-rides in a local setup directory?

If both property set, and property subset, define default values then
that makes perfect logical sense. From what you said above, this seems
to be the case. The "property subset" is a proper subset of the
"property set" and over-rides any settings in the "property set".

If you want to say that the "property subset" cannot define predicates
or new values, that's fine. It *IS* a *SUBSET*, after all :-)
> 
> After some thinking I disagree with the notion that preset is wrong in
> itself. The thing I'm referring to isn't a wrongly-worded mathematical
> set but a set of values to preset (in the ordinary English meaning of
> setting in advance) a subset of available properties.
> 
> However, I agree that it would be unfortunate to have "set" in such
> different meanings in "property set" and "preset".
> 
Except I'm a native speaker. To use the word "preset" in this context
jars *horribly*. The implicit assumptions built into the mind of a
native speaker will make this a lilypond "special case" use of language
and while non-native speakers will simply learn it, native speakers will
have great difficulty unlearning the normal meaning of the word.

Cheers,
Wol