[Info-vax] Programming languages on VMS

[Jan-Erik Soderholm]

Den 2018-01-24 kl. 15:09, skrev Simon Clubley:
> On 2018-01-23, Jan-Erik Soderholm <jan-erik.soderholm at telia.com> wrote:
>> Den 2018-01-24 kl. 00:36, skrev already5chosen at yahoo.com:
>>>
>>> That's not what I meant. I had in mind all code-generation tools also
>>> running off-VMS, ether on desktop, or, for large projects, on non-VMS
>>> build server. Only minimal debugger agent runs natively.
>>>
>>
>> Ah, OK... :-) The jury is still out on that one, I guess... :-)
>>
> 
> That depends on whether it is better to treat VMS as an embedded
> target or as a host machine that you can log into and develop
> software on the operating system itself.
> 
> The setup already5chosen describes is absolutely normal for headless
> embedded development but OTOH, with VMS you would still need
> to be logged into the machine anyway to do all the other things
> associated with software development and testing for the type
> of software that typically runs on VMS.
> 
> Simon.
> 

Yes, I have done some development for 8-biters, and of course,
it is a whole other kind of environment where you produce ready
made "hex files" that are "flashed" into the processors. But is
there any reason we cannot try to keep the discussion around VMS?

In VMS, most of the infrastructure is already there. Compilers,
the linker and so on. It is the actual editing of code that is
somewhat behind. It is probably easier to interface some client
IDE environment to a VMS backend for builds, then to "learn" some
desktop IDE (and tools) to understand the VMS envionment.

Now, for *some* development environments, such as those that are
JVM based, it might be easier to do the full build on some other
environment and just deploy the JAR (or whatever) to the VMS system.

But I do not see the VMS specific Fortran, Cobol and Basic compilers
beeing ported to Windows or Linux and still producing VMS code.
I might be wrong, of course. Has happend before...