[Info-vax] VMS Software needs to port VAX DIBOL to OpenVMS X86 platform

[Simon Clubley]

On 2020-12-26, Arne Vajhøj <arne at vajhoej.dk> wrote:
> On 12/26/2020 5:52 AM, Simon Clubley wrote:
>> On 2020-12-23, John Reagan <xyzzy1959 at gmail.com> wrote:
>>> On Wednesday, December 23, 2020 at 1:54:02 PM UTC-5, Stephen Hoffman wrote:
>>>> expectations and the current market for operating systems. And VSI
>>>> doesn't have MICA source code, and anybody that does have that source
>>>> code is unlikely to share it.
>>>
>>> Please no.  Somebody will ask me for a PILLAR compiler (which was derived from the EPASCAL compiler - compare the PILLAR and EPASCAL manuals)
>> 
>> I've read the Pillar manual and I liked it.
>> 
>> It had a lot of good things in it and if it had been implemented during
>> the 1980s it would have been a _very_ viable alternative to the C
>> language that existed at the time.
>> 
>> If it got established, it could have changed computing history and today
>> we could now have had a viable, and established, Pascal-like alternative
>> to C when writing operating systems and other low-level code, with
>> everything that would imply for system security.
>
> Maybe.
>
> But there were languages with Pascal syntax to chose from.
>
> Modula-2, Ada 83 etc..
>

None of those were used as the system implementation language in a
primary operating system from a major vendor at the time.

Ada is actually an example of what could have been for Pillar because
Ada was a language that did get established in the defence industry due
to the mandate and if the Ada compiler situation was far better than it
currently is, it might still be much more popular generally these days.

If Pillar had been used in a DEC operating system at the time, it could
very well have been the jump-start such a language needed to get
established for use within operating systems and other low-level code.

Just look at what happened to Pascal for application programming
(but not OS programming however) when Turbo Pascal and then Delphi
came along.

>> Given that the Pillar manual was never officially released but sort-of
>> "leaked", I wonder if anyone would get into trouble (copyrights, etc)
>> if they implemented Pillar using, for example, LLVM ?
>
> Re-implementing from functional specification has traditional been
> considered OK.
>

Even if the specification was not officially released, but just leaked ?

Simon.

-- 
Simon Clubley, clubley at remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.