[Pharo-project] [COTDC] 12 - PluggableListMorphByItem

Igor Stasenko siguctua at gmail.com
Tue Feb 15 21:23:09 CET 2011

On 15 February 2011 21:12, Stéphane Ducasse <stephane.ducasse at inria.fr> wrote:
> I would love that you show me that. I will ask alain to show me that too.
> I want to learn that and we fix them.

most of them are just adding indirection to some state and/or behavior.
At cost of adding a lot of extra ivars.

In general there is a quite visible pattern: every time people think
that it's cool to make some morph's properties
customizable, they adding ivar + accessors to PluggableXXX morphs.

And since there is virtually no limits of how many "useful" new or
custom properties could be added to widgets,
the tendency of growth of these things to some horrid mosters become
higher and higher.

So, the question arises: should we rethink the model instead? And
think towards adding extra/fresh properties to
existing class(es) instead of making subclass with weird names which
look like a piece of junk.

> Stef
>> On 15 February 2011 19:03, Gary Chambers <gazzaguru2 at btinternet.com> wrote:
>>> "Essentially useless in its current form. Probably intended to eventually
>>> deal with actual objects in the list rather than indexing/string-forms."
>> every time i looking at those pluggableXYZ. and it always one thought
>> don't let me rest:
>> - we don't need them.
>> They are implementing/perverting functionality of base widgets.
>> Because for most of them, they addding no value comparing
>> to base widgets, but just some extended interface(s) on top.
>>> Regards, Gary
>> --
>> Best regards,
>> Igor Stasenko AKA sig.

Best regards,
Igor Stasenko AKA sig.

More information about the Pharo-project mailing list