[MODEST] comments

[dina]

Dear modest colleagues,
  
We read the paper resulting from the AMNH meeting and learned 
a lot about the dynamical evolution part. As regards the 
interface between stellar dynamics and stellar evolution, 
only the third option (i.e., "Maximal Interface") seems 
reasonable (or feasible) to us, mainly due to the large 
differences in time scales both between SD and SE and also 
among SE courses of different stars. The practical solution 
would be to have evolution tables (or arrays) for all stars,
as an initial library for "primordial" stars (as Ron and 
Onno call them) and calculated ones for collision mergers 
(produced on the demand of SH). Thus SD and SE should be 
completely independent and should communicate only via tables. 
A point which might still require some consideration is 
whether a collision merger should be evolved all the way or 
only for a time within which it is likely to undergo another
destructive collision. However, as far as we understood,
each star is expected to undergo such a collision about once 
in its lifetime, so it seems that in any case long term 
evolution of a merger product would be required.

As to Ron's suggestion that some phases of evolution need 
to be parametrized (mainly the AGB phase), we fully agree, 
but in this case, too, this should be done within the SE task. 
Thus the final, full-scale tables for primordial stars may 
result from different types of calculations, but SD and SH, 
which make use of them, need not be aware of the different 
methods employed. It may even be worthwhile  considering 
"toy-tables", rather than "toy-SE-codes", to play and 
experiment with.

With best wishes,

Dina and Attay