r/ex/w experiments

[John P. Morgan]

On Mon, 4 Aug 2003 00:18:06 +0100
Peter Dobcsanyi <p.dobcsanyi at designtheory.org> wrote:
 
> Many of the
>computational functions provided by the GAP and R packages are (or will
>be) re-implemented in Python in this framework 

Peter, can you tell us why this is being done? If GAP and R are calculating 
everything needed and we have readers and writers for both packages, why expend 
the effort to rewrite in python? I don't want to intefere with your business, 
just to understand. :-)

>1)
>
>We should not rush the publication. It seems we can meet the deadline
>with the ext-rep doc, but now, I feel we need a few weeks "cooling off"
>period before publishing. Say a publication around the first week of
>September would be a reasonable target. (In time we wouldn't lose too
>much, everybody is on holiday in August, not to mention that I'd like to
>go away too :-)

Delay is okay with me, but can we agree to be 99% certain of no further delay? 
It is VERY important to keep moving.

>
>    - JP and Leonard that could be a good opportunity for you to use
>      your Reader/Writers for a real thing and re-establish the R-GAP
>      connection what you used before but now on top of ext-rep.
>      I have the most minimal Readers in my mind, just to get the
>      necessary job done. What do you think?
>

This may be possible, but I am not too concerned with the R-GAP connection of 
this Spring since it does not mean a lot for long term (it was for a particular 
set of computations). Is there a general need for R-GAP communication?

>2)
>
> My experiments showed that my worries about complicating things
>with the Reader and the impact of loosing DTD-s were unwarranted.  
>
>(Now, I run for cover and hope JP would pull only his previous Writer :-)
>

ARRGH! :-)


JP