OpenEuphoria

Poll Debrief

• Posted by Peter Robinson <indorlaw at yahoo?com.a?> Sep 15, 2007
• 587 views

Hello all

Following on from the recent polls, a few issues were raised but left
unresolved. They don't have to be resolved now but I thought I'd draw attention
to them while the issues are fresh.

In a thread entitled “Postscript – Sequence of Types” started (by me) on
27/8/07, Derek raised the idea that the syntax 'sequence of sequence of etc'
contains redundancy - i.e. since the word 'sequence' is the only possible word on
the left of an “of”, a simpler syntax should be considered. The idea seemed to
get some support, but the discussion diverted to a second idea of constraining
the length of a sequence within the same declaration.

In and around the second poll on Types within a Type, the idea re-surfaced that
there should be a record or structure declaration available that could do the
same thing as the proposed type declaration, and may be preferable. There seem to
be a few people who like this idea, but it does raise the question – Is it
proposed that a stucture/record declaration would be a drop-in replacement for a
type declaration? For example, could you include boolean expressions within such
a declaration? Is it a proposal that should be addressed at all?

It has been suggested that the proposal for typing within a type-or-sequence
declaration is OO-driven, but I don't see it that way. This kind of thing was
part of Pascal, wasn't it, and certainly part of C. It just allows you to declare
an object in the same way that you picture it in your mind.

Proposals that look anything like an OO language tend to cause controversy on
the list, but I haven't seen the detail addressed very often. I wonder whether it
would be better to address aspects individually rather than assuming that it's
all indivisible. The main features of OO are:-

encapsulation of data  (which was addressed in the recent polls)
encapsulation of routines relating to the data
inheritance of data and routines by objects and sub-objects of the same type
polymorphism; i.e. A single name can have multiple meanings in an inheritance
context.

I think it would progress matters if the community examined these issues and
decided one way or the other whether they have any merit within Euphoria. I
believe a Yes or No answer would represent progress.

This is just an invitation. I'm not proposing another poll unless some broad
concensus is reached.

Cheers
Peter Robinson