[Info-vax] Fun: Object Pascal on VMS

[bill]

On 9/3/2024 6:27 PM, Craig A. Berry wrote:
> On 9/3/24 2:59 PM, Arne Vajhøj wrote:
>> On 9/2/2024 7:51 PM, Lawrence D'Oliveiro wrote:
>>> On Mon, 2 Sep 2024 11:35:16 -0400, Arne Vajhøj wrote:
>>>> On PC:
>>>>
>>>> C:\Work\VMS\objectpascal>ppcjvm demo.pas
>>>> ...
>>>>
>>>> On VMS:
>>>>
>>>> $ java -cp demo.jar:fpcjvm.jar "demo"
>>>> F1=0,F2=
>>>> F1=123,F2=ABC
>>>
>>> VMS can no longer self-host its own development stack any more.
> 
> It depends on what is meant by "its own development stack." It certainly
> does self-host the tools used to build the OS itself.
>> There are lots of compilers and interpreters available on VMS.
>>
>> There are also a bunch that are not available on VMS. That
>> bunch include FPC.
>>
>> If someone wanted FPC to run on VMS, then they could
>> port it. It is open source.
>>
>> FPC has really nothing to do with VMS, but compiling to
>> Java byte code is a very broad targeting cross compiler.
>> Which I took advantage of.
> 
> And I think everyone who wasn't trolling recognized that.  The JVM
> brings a big party with it wherever it goes.  .NET would bring a
> slightly smaller but still nice-to-have party.  With LLVM as a
> foundation for the current crop of VMS compilers, many things are
> possible (but nothing is easy).
> 
> Has anyone actually tried porting FPC to VMS?  It looks at first glance
> like a lot of it is written in Pascal, so I assume it would need to be
> cross-compiled initially.
> 

Why?  Wasn't the VMS Pascal compiler ported?  I would think it
would be a lot easier porting something written in Pascal
compared to something written in C.

bill