On Tue, 15 Jan 2002 [EMAIL PROTECTED] wrote:

> there was? i did not notice that ...

stephen decided that not enough people cared about requiring
5.6.1 to call a vote on the topic, and nobody pitched a
fuss. that's good enough for me.

> but you can specify types in another way, attributes is
> just the cleanest approach since types can be tagged
> onto the variables.
>
> without types you won't be able to automatically compare
> data structures. they tell you wether you have to
> compare the values the hashref is pointing to (array
> type) or the reference itself (ref tyoe) as it was in
> your previous example.

i think the question is, how much comparison do we really
have to do? i'm not convinced that the extent you describe
is necessary. but then i didn't study computer science, (for
very long anyway), so i might be talking out my ass. maybe i
should eat some acid and contemplate the mathematical and
philosophical underpinnings and implications of object
identity.

:)

Reply via email to