Re: scratch/igc: Implications of MPS being asynchronous

[Pip Cet]

"Stefan Kangas" <stefankangas@gmail.com> writes:

> Eli Zaretskii <eliz@gnu.org> writes:
>
>>> Do you have a better word to propose here than "asynchronous"?
>>
>> I don't think this aspect of MPS is very important to us.  So I would
>> suggest simply dropping that part.
>
> Do you suggest dropping the words "synchronous" and "asynchronous", or
> not documenting this at all?

In general, I think "asynchronous", "synchronous", or "concurrent" are
unlikely to be useful terms, since there are competing definitions.  The
quote you suggested is fine, because it defines what "asynchronous"
means explicitly; even for that quote, though, someone is likely to
substitute their own unusual definition of "asynchronous" and argue
about the circumstances in which MPS does or does not meet that
definition.

Participating in such arguments is a waste of time.  Some people say 0
is a positive number, and I say it isn't, so just define which version
of the term you mean or avoid it in the first place and say
"non-negative" instead.  (Very rarely, mathematicians finally reach a
consensus on such questions, as has happened with "is 1 a prime", "is 0
even or odd", or "is infinity a real number", but for every such case
there are dozens of others in which conflicting definitions remain in
use).

The "don't make no-GC assumptions" rule is vitally important, it's a
major difference to traditional GC, and it needs to be documented.
Thanks for doing so.

Pip