Re: Much 16 bit Ada useage?

[John McCabe <john@assen.demon.co.uk>]

stt@houdini.camb.inmet.com (Tucker Taft) wrote:

>>>However, it seems that the European production of the MA31750 will
>>>stop, so new projects are likely to use the European and space
>>>qualified implementations of the SPARC (ERC-32) or the ADI 21020
>>>DSP. (A port of GNAT for the latter would be nice. Does it exist?)

>: I don't know if it exists, but Tartan were working on Ada for the
>: SHARC (AD 21060) before being bought by TI, so I don't know what's
>: happened to that.
>
>Intermetrics took over the SHARC Ada contract when TI bought Tartan.  A
>beta version of the product is now available from ADI 
>(lori.berenson@analog.com).
>
>However, the SHARC is definitely not a 16-bit processor.
<..snip..>

Of course - I wasn't trying to imply that it was. My comment was simply 
in response to the previous comment regarding the 21020. What I was 
trying to get over was that, at the time I looked at it, no one produced 
an Ada compiler for the 21020, and the closest I could see to that being 
a possibility was the Tartan SHARC compiler.

>: I can still see a market for 16-bit processors in space applications,
>: and I'm sure they'll be supported by Ada compiler - although maybe not
>: Ada95.

>If there is genuine interest, an Ada 95 compiler could be developed
>relatively quickly, using either the GNAT/GCC approach, or the
>Intermetrics AdaMagic front end generating ANSI C as an intermediate.
>The basic problem is finding that first customer who wants to
>help foot the bill...  We welcome such inquiries.

Last I heard, TLD were supposed to be getting together with ACT to come 
up with an Ada95 compiler for the MIL-STD-1750 processors build on the 
GNAT technology. I don't know if that happened or not. Anyone else know?

I was never particularly impressed with TLD's Ada 83 for MIL-STD-1750 as 
a product, but when the code it generated worked, it was s**t hot.