>> Does the above idea help resolve this concern, which I admit is
>> perfectly legitimate?  It allows a specific status object to belong to
>> many groups.
>
> If 'convention can be a list, that solves it.
>
> But doesn't testing for the existence of specific properties relating to
> the desired convention also solve it? If you have a non-#f 'errno
> property, you can do errno-related stuff to it;

Aha, here's a further point of - not really disagreement, but different
priorities.

I would bank heavily on picking property names (going into the Scheme
Registry) that are hard to misuse. For example, I wouldn't coin a 'line
property because it doesn't say whether it's a line number, a line of
text as a string, or something else. 'line-number clearly says it's a
number.

Likewise, 'sqlstate or 'errno are names that clearly refer to standard
terminologies. 'errno is vaguely risky ("error number" is a generic
concept) but 'errno is so universal that there's little risk of
confusion, especially as we'll document it in the Registry.

The tentative list in the registry has 'facility and 'severity. These
are deliberately vague. The concepts exist in syslog, but also in VMS
and other systems. The principle is that vague concepts are fine, but
their names have to be equally vague. A 'vms-facility or
'syslog-facility property could have the same information with a more
clearly defined meaning. Or a 'which-facility property with 'vms or 'syslog.

If property names are chosen haphazardly so that they clash easily, my
proposal is nullified and doesn't make sense.