Re: Standard Queue status

[Jeffrey Carter <jeffrey.carter@boeing.com>]

Mats Karlssohn wrote:
> 
> Ted Dennison wrote:
> %<
> > Personally, I like his use of the "Direction" type to specify which end to start
> > from in his operations. I also like the ability to convert between lists and
> > unbounded arrays. I think those may merit putting in the final version. There
> > are also a number of operations in there that the strawman doesn't have which
> > some folks may find useful.
> 
> Yes, the use of "Direction" is elegant, and converting to and from
> unbounded arrays seems useful.

I don't know. Specifying the direction seems like control coupling.

I don't recall an "unbounded array" in his spec. I recall unconstrained
arrays, but that's a different animal. Converting to and from
unconstrained arrays is trivial to implement with the standard features
of a list, so I'm not sure it should be in the basic package; on the
other hand, I'm not sure it shouldn't be there if it's going to be
defined at all.

> After all one of my general rules is:
> "Don't bring in more than you need"
> That is, as much as I like nifty features, I want to be able to
> pinpoint exactly what facets of an object/alorithm-collection that I
> want to bring in for use when solving a problem. Even if
> instansiations becomes numerous or _somewhat_ more complicated.

There needs to be a balance. After all, the minimum features one needs
to build and use an unbounded list are provided by Ada, but I hope no
one thinks this is an argument for not having a list ADT. On the other
hand, the operations one can think of doing with lists are limited
mainly by one's imagination. So the general rule should be something
like: include the basic operations, operations that cannot be
efficiently implemented without directly manipulating the
implementation, and extremely common operations; exclude everything
else.

The problem is agreeing on "extremely common operations".

-- 
Jeffrey Carter