On Wed, May 6, 2009 at 11:08 AM, Derick Eddington
<[email protected]> wrote:

>Why aren't you
> complaining about how in PLT an explicit phasing (for --- expand) does
> not instantiate the library at run-time?  But you want an equivalent
> import form in implicit phasing to do so...

In PLT different phases have different instances, so a library can be
instantiated
at expand time but no at run-time. In implicit phasing there is a
single instance
for all phases, so I expect an imported library to be instantiated
both a compile time and
at runtime, even if it may be a waste (I am not arguing it is not a waste).

>> >> Suppose I am importing a library
>> >> purely for its side effects
>> >
>> > That is a major misuse of importing.  Importing is purely a lexical
>> > scope thing.  It's far more confusing to have to figure-out that a
>> > lexical import is being done to implicitly cause side-effects at
>> > execution time.  If you want to execute something in an imported
>> > library, say so (in code), like you would for a zillion other cases.
>>
>> You say so, but at least the PLT people think in a different way.
>
> I'm pretty sure they'd agree that importing *only for causing
> side-effects* (which is what we were talking about) is a major misuse.

Let me declare officially that I am not fond of side effects and I
dislike designs
based on side effects; nevertheless, I use side effects for logging
and debugging, so I do have modules depending on side effects when
developing/debugging an application.

> As increasingly seems to be your usual, you're twisting the conversation
> into a maze in order to appear to defend what you previously said, when
> in fact you don't address it.  Or maybe you're just not paying
> attention...

Don't take it personally, Derick. We can just agree to disagree on the
point of what is the "right" import semantics.

>> I say the meaning of "import" is open to interpretation in the
>> Scheme world and I do not know of any other implementation
>> using the meaning in the strict sense of Ikarus.
>
> After searching all four R6RS documents for every occurrence of
> "import", I now know they only talk about importing in the context of
> lexical scope.  Hmm...

Let's close the discussion on this point, since I do not think
it would be a productive one.

Reply via email to