On Fri, 20 Jul 2001, Jim Linnemann wrote:

> Dear Mike,
> 	Thanks very much for the Handel-C stuff; I've browsed through it and find
> it quite interesting.  I'll talk with some of my other colleagues and maybe
> check university pricing.
> 
> o in reading my notes, I was confused on one point: did you say that the
> interactive simulator was not up to a full scale project, but rather was
> only a good learning tool?  If so, where should the simulation be done?

	What I meant to say is that with literally no effort one can use
	the simulation mode to understand why simple code fragments don't
	work the way they were foreseen to work.  To debug a realistic
	system, however, one needs to generate the input vectors for
	the complex system and understand the output vectors generated. 
	The latter is no different from what you would have to do for 
	VHDL. For simple cases, Handel-C is much easier than VHDL; for
	complicated systems they are not very different.
> 
> o Would you recommend use of Handel-C in, say, a junior-level physics
> electronics course as a teaching tool, or is it too removed from the hardware
> to be useful for that?

	Absolutely. I think the level of abstraction provided by Handel-C
	is the right way to get newbies thinking about parallel
	processing.
> 
> o your numbers sounded like Verion 3.0 makes you pay a substantial price
> from removing some limitations.  Are they claiming they will improve it so
> it's as fast and efficient in resource use as the earlier version?  Because,
> as usual, we're always in danger of pushing the hardware limits.
> 
	It is improving, and some of our compilation times were
	exacerbated by limited memory on the host (64MB). Things
	get much better with 256 MB.


	regards
	Micheal