Of course ;-) Well, as it opens the world of IValues to objects unknown to us:
- (de)serialization would have to be 'open' as well, which prohibits certain kinds of optimizations and hampers simplicity of its implementation
- equality testing of IValues can not be fully controlled, since eventually it will boil down to equality of IObject values.
A design where we have fully canonical IValues could not benefit from constant time equality check if arbitrary(non-canonical) IObjects can
be nested inside other values.
- a general queasiness about not being able to fully control and standardize the shape and types of IValues, leading to analyses and visualizations not
being compatible or composable.
- me not having a direct goal for them anymore, which used to be the nesting of ATerm's in IValues. Since ITree will replace ATerm, I don't need this bridge
anymore.
Cheers,
Jurgen
On Fri, Oct 10, 2008 at 5:01 PM, Robert M. Fuhrer
<rfuhrer@xxxxxxxxxxxxxx> wrote:
Hi Jurgen,
Since you spent a while convincing me to add it, I think you're honor-
bound to spend a little time convincing me of the opposite now. ;-)
On Oct 10, 2008, at 7:25 AM, Jurgen Vinju wrote:
Hi Bob,
The object wrapper from pdb, I want to get rid of it.
I know, I know, I insisted on having it in the first place ;-)
Cheers,
Jurgen
imp-dev mailing list
imp-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/imp-dev
Cheers,
- Bob
_______________________________________________
imp-dev mailing list
imp-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/imp-dev
