[om] Concatenation of lists ??

[Andrew Solomon]

> Thanks for your suggestion, what do other people think?

Here's my perspective as a software developer of how new symbols and CDs
should come about:

I was writing a web service to render Hasse diagrams the other day
and ran into a bunch of things I didn't know how I ought to do in OpenMath.
I wrote to the list and received some good suggestions about how to do it
and then went back to my code, using these ideas to produce 
Hasse diagrams from my phrasebook.

That is,
[Step 0: preliminary discussion on the list]

The next step is a subtle one which bears thinking about. It's one thing
to say (cd=poset name=hasse_diagram) is a symbol which takes such and
such arguements, and another to actually produce a concrete example of
a hasse diagram from your phrasebook and see what it looks like when 
the elements are of various concrete types. When we are confronted with
a bunch of examples we might then see provisos or caveats which should be
added, or conventions which should be followed when using this symbol.

Therefore I think the next step is to:

[Step 1. show the list a few examples of OMOBJs produced from an 
actual phrasebook]

No doubt, the list will perceive infelicities in the way 
particular examples are represented, and the software developer will 
see how their suggested changes to the way an object should be represented
can be accommodated in the software and come up with counter suggestions:

[Step 2. The list discusses the merits of how a symbol
works (and its implementation in a phrasebook). They 
agree on semantics and usage for the symbol.]

Finally,

[Step 3. The CD editor should have the duty of
deciding when agreement has been reached and coalescing the 
discussion on the list into a final CD.]

In practice the editor would do this by sending an email to the list 
with PROPOSAL in the subject line, describing the proposed symbol
definition, and then when objections have died down (or been over-ruled;)
the symbol gets etched in stone.


I really think this would be the way to develop new symbols - driven 
by phrasebook development. If the CD editor (who is this again?) agrees
we could give this process a trial run with the hasse_diagram symbol
and see how it goes.


Andrew


--
om at openmath.org  -  general discussion on OpenMath
Post public announcements to om-announce at openmath.org
Automatic list maintenance software at majordomo at openmath.org
Mail om-owner at openmath.org for assistance with any problems