Re: A question Re: Level 1 <---> level 2 mappings; arithmetic versus applications
> From: "Svetoslav Markov" <smarkov@xxxxxxxxxx>
> To: Dan Zuras Intervals <intervals08@xxxxxxxxxxxxxx>,
> P-1788 <stds-1788@xxxxxxxxxxxxxxxxx>
> Date: Sat, 10 Jul 2010 13:16:04 +0300
> Subject: Re: A question Re: Level 1 <---> level 2 mappings; arithmetic versus applications
> CC: Nate Hayes <nh@xxxxxxxxxxxxxxxxx>
>
>
> Dear Dan,
>
> your detailed posting convinced me in
> your intention to find a proper place for
> the mid-rad format in the standard.
>
> Up to now I was anxious about a possible preclusion
> of the mid-rad form in the standard which IMO
> will be certainly a big mistake.
>
> I congratulate your idea to look for a compromise.
> It seems to me that the idea about two "abstract data
> types" (explicite and implicite) might be the correct
> approach to such a compromise. I shall follow with
> interest the developments in this direction and shall
> try to be helpful as much as I can.
>
> Best regards,
>
> Svetoslav
>
Thank you Svetoslav.
I cannot take credit for another man's work.
The idea to separate abstract datatypes into
explicit & implicit types was Nate Hayes'.
The rest of us merely pounced on it as a
possible means to the compromise we seek.
John is working out the language as we speak
but I have high hopes for this approach.
It gives us the conceptual framework to have
some datatypes pinned down to the very last
bit (the explicits) whilst allowing others
(the implicits) the freedom they need to
implement their approach to intervals.
Of course, that is my opinion of it all.
You may find that each of us touches the
elephant in a different spot. We are still
discussing the nature of the beast.
But it feels good to me.
Yours,
Dan
P.S. - You should know that we have discussed
your words as well as others who have posted
here. It has helped. I thank you for keeping
an open mind to what is naturally a work in
progress. That too helps.