Re: Lisp Serialization: GZ support?

[Paul Johnson]

The hang-ups I've got with hdf5 output are

1. Can't save dynamically allocated arrays of ints and doubles
and
2. Can't save Swarm Arrays of objects.

You've shown me how to do #1 in Lisp and I've patched Swarm to do #2 in 
Lisp. (I think I sent that patch recently, but if I didn't, please let 
me know and I'll send again).

To me, the  documentation for HDF5 from NCSA is still 
discouraging/overwhelming, and from looking through Swarm, I don't 
understand how Swarm is using HDF5 most of the time.  I don't have 
anything against HDF5, except that I don't have time/patience/training 
to understand it.  Even the terms they think are basic and obvious like 
"compound" are for me puzzling and mysterious.

I could anticipate your counter argument that if I'm willing to try to 
learn gz compression library, I ought to be willing to learn HDF5 
instead.  That is a good one.

But I've got a short-term deadline of proving serialization can be done, 
at least with the Lisp I've gotten that far.  I've got a medium term 
goal of running a model for a few months and then re-analyzing its 
output data.

pj

Marcus G. Daniels wrote:
> PJ> I don't know how to do the archiving output in HDF5. 
>
> For storing object references in HDF5 you can do the same kind of
> thing as with the Lisp serialization.  Make a global Map of integers
> associated with the (global, multi-referenced) objects and store the
> integers in the agents.
--
Paul E. Johnson                       email: address@hidden
Dept. of Political Science            http://lark.cc.ku.edu/~pauljohn
University of Kansas                  Office: (785) 864-9086
Lawrence, Kansas 66045                FAX: (785) 864-5700


                 ==================================
  Swarm-Support is for discussion of the technical details of the day
  to day usage of Swarm.  For list administration needs (esp.
  [un]subscribing), please send a message to <address@hidden>
  with "help" in the body of the message.