And speaking of ALGOL

[Paul Koning]

> On Aug 12, 2015, at 1:39 PM, Nigel Williams <nw at retrocomputingtasmania.com> wrote:
> 
> 
>> On 12 Aug 2015, at 11:24 pm, Paul Koning <paulkoning at comcast.net> wrote:
>> ...
>> Did Algol come after the hardware?  I always thought of the hardware as having been customized for their Algol, but admittedly I don’t actually know which is chicken and which is egg.
> 
> It is suggested in the oral history at UMN.edu that the B5000 was designed as an ALGOL machine and Burroughs had the idea that only compilers would generate machine code, so they made the B5000 compiler friendly, and the system would have an OS to manage resources, so it was designed around drum/disk being an intrinsic part of the system.

That sounds right.  For one thing, it is clear if you study the hardware manuals that the system is NOT secure against machine language programmers.  Security comes from the fact that the compilers other than ESPOL will not generate code that compromises security, and the ESPOL compiler is protected so that ordinary peons do not have access to it.

I found this out when I tried to write a program that reads foreign format tapes, in particular past tape marks.  Algol can’t do that — either that, or the consultants couldn’t figure out how.  I started looking at other languages, but when I started asking questions about DCALGOL I got a whole lot of pushback from the system staff. They viewed questions like that with extreme suspicion.

	paul