Re: Thread/mutex/&c names

[Vladimir T. <vtz...@gm...>]

On 7/13/09, Sam Steingold <sd...@gn...> wrote:
> I think I was not quite clear on this issue.
>
> On a few occasions, I used the following paradigm (not sure this
> deserves such a grand name): an object would have a slot NAME,
> and the following invariants would hold:
>
> (eq (symbol-value (foo-name foo)) foo)
> (eq (foo-name (symbol-value sym)) sym)
>
> while I am not prepared to extol the virtues of this approach,
> ISTR that it was quite useful for i/o, data presentation, structural
> updates (when change-class was inconvenient) &c.
>
> I think it would be useful to permit the name slots to be symbols, not
> just strings.

Agree.
Is it ok to implement it via %record-ref?

Vladimir