Re: Object Oriented Euphoria
On 12 Dec 2004, at 19:07, Matt Lewis wrote:
>
>
> posted by: Matt Lewis <matthewwalkerlewis at yahoo.com>
>
> Kat wrote:
> >
> > On 12 Dec 2004, at 12:51, Matt Lewis wrote:
> >
> > >
> > > posted by: Matt Lewis <matthewwalkerlewis at yahoo.com>
> > >
> > > Kat wrote:
> > >
> > > > Also, dynamic changes in methods and any data field while the program is
> > > > running, with changes (or additions or deletions) propagated thru all
> > > > descendants instantly.
> > >
> > > I don't think there's any way to do this (unless I start mucking around in
> > > the back end, which destroys my ability to be compatible via preprocessor
> > > or
> > > possible future 2-step binding). I don't see an example of this above.
> > > Could you give an example of how you might do this?
> >
> > 1) one fine day in Oz, we don't know of the playtpus
> > 2) the next fine day in Oz, we find it
> > 3) the following day, supervisors are raising heck because the platypus
> > cannot exist in the known classifications.
> > 4) a confusing day in Europe, those nuts in Oz came up with a platyus thing,
> > and we haveto make a new class for it without rebooting reality.
>
> I understand why this is useful. I'm just not sure how it could be
> implemented--other than, I suppose, some sort of eval function. I've
> thought a little about this, but I'm not sure how to do it...
>
> Do you have an idea of what an implementation might look like?
In the Eu source, no, sorry. Since RobC won't put in an eval(), i can't think of
a way to make this happen in Eu. With eval(), it would be simple to drop
everything into associated lists, use routine_id and var_id and eval() to add
classes, methods, etc while the app is running. It's so difficult to come up
with every scenario of input and expected output in real world interactions,
but with mirc it's almost trivial because i can $eval() the latest data and
methods of handling it.
Kat
|
Not Categorized, Please Help
|
|
