I cannot agree with Lawrence's conclusion that the CCS compiler is
OK. It seems to be OK, if used carefully, for the smaller RAM sizes.

WARNING: The CCS compiler DOES NOT work correctly (even the latest
version) for four-bank parts using the straightforward "long" address
compiler mode.

The insurmountable bugs that I found are in the large RAM support
mode.

I have three open bug reports with CCS.  Telephone support is not
available in fact.  And they have not answered my emails AT ALL, even
to say something like "sorry, you're 20th in line".  NOTHING.  (In
case you are wondering, I paid for support- what a sucker.)

> "you get what you pay for"
Is probably the driving force.  CCS is a small company and
seems not to have adequate resources to support the product.  There
is ONE support person, AFAIK.  Which would be OK, if they actually
did support, but he seems not to be available much.

> I write some good, compact code with it and I am continually surprised by
> the tricks it knows.
I does write good code for the mainstream cases on small RAM parts.
I chose CCS over Hi-tech because I2C support was built in
(preprogrammed in the library).  Unfortunately, it doesn't work right
in Master mode (contrary to claims in the manual.)

And finally, the manual is awful.  It is incomplete, has significant
typos, tons of unsignificant ones.

So my conclusion is that for hobbiest level cost, reliablility and
support, and size of project, it may be adequate.

But for pros who are using the compiler to do a project that is a)
Big, OR b)on a limited schedule OR c)needs to be verifiably correct,
I cannot recommend it.

------------
Barry King, KA1NLH
Engineering Manager
NRG Systems "Measuring the Wind's Energy"
Hinesburg, Vermont, USA
barry@nrgsystems.com
"The witty saying has been deleted due to limited EPROM space"