> Michael G Schwern wrote:
>
> > Also, just being able to tack flags onto a variable means each
> > variable (that has a flag) would have to carry around a hash!
No; it has to carry around a vtbl (or vtbl-like thing).
User-level attributes are (or should be :-) accessor methods
to some underlyin
Michael G Schwern wrote:
> Also, just being able to tack flags onto a variable means each
> variable (that has a flag) would have to carry around a hash!
Or it could mean that every new flag is associated with a hash of
names of things for which it is set -- in which case
$something:at
At 01:46 PM 9/18/00 -0400, Michael G Schwern wrote:
>On Mon, Sep 18, 2000 at 04:04:56PM +0100, Simon Cozens wrote:
> > On Mon, Sep 18, 2000 at 10:51:52AM -0400, John Porter wrote:
> > > I would think that if it could be done at all,
> > > it would only be in extension (formerly XS) code.
> >
> > W
On Mon, Sep 18, 2000 at 01:46:57PM -0400, Michael G Schwern wrote:
> I'll have to go with John here. Mucking about with variable attribute
> flags (with the exceptions of the fixed set which have been proposed)
> falls firmly into the "hard but possible" realm of things. It should
> be done in X
On Mon, Sep 18, 2000 at 04:04:56PM +0100, Simon Cozens wrote:
> On Mon, Sep 18, 2000 at 10:51:52AM -0400, John Porter wrote:
> > I would think that if it could be done at all,
> > it would only be in extension (formerly XS) code.
>
> Why? I don't want to go to C just to add a flag to a variable.
Simon Cozens wrote:
>
> (The deadline for collecting ideas passed two weeks ago. Why is this all
> still going on?)
Because there are still many worthy ideas which have not surfaced yet.
Which is the higher priority?
--
John Porter
We're building the house of the future together.
At 04:04 PM 9/18/00 +0100, Simon Cozens wrote:
>On Mon, Sep 18, 2000 at 10:51:52AM -0400, John Porter wrote:
> > I would think that if it could be done at all,
> > it would only be in extension (formerly XS) code.
>
>Why? I don't want to go to C just to add a flag to a variable. That smacks of
>ma
On Mon, Sep 18, 2000 at 10:51:52AM -0400, John Porter wrote:
> Are all the possible attributes going to be predefined, or can the
> user define new ones?
The user should be able to do anything they damn well like. This is,
allegedly, Perl, which means it's about making it easy to do what the
pro
Peter Scott wrote:
>
> How about an attribute for hashes:
>
> my %foo : fixed;
Has anyone talked about the ability to access the
attributes attached to a variable?
Are all the possible attributes going to be predefined, or can the
user define new ones? I would think that if it coul
