[Info-vax] Reimplementing VMS, was: Re: HP adds OpenVMS Mature Product Support beyond the end of Standard Support

[Johnny Billquist]

On 2014-02-02 14:16, JF Mezei wrote:
> On 14-02-02 14:25, Johnny Billquist wrote:
>
>> Thus this false rumor spreading needs to be stopped.
>
> Your speculation that HP DID NOT evaluate of port of VMS to 8086 has
> less validity than my statement that it did evaluate it even when I
> remove what I was told from an informed source.

You are suggesting that HP did more than do an evaluation, in that you 
suggest some code was written.
That the did some kind of evaluation at one point or another would be 
expected. But that is more a case of asking the question "can it be 
done, and how much work would it be". Some people would then sit down 
and get some rough answers to this.

At no point is any code actually being written as a part of this. It's a 
feasibility study.

>  From a business point of view, when HP is negotiating the end of IA64
> with Intel, it makes perfect business sense to evaluate the
> costs/complexity of porting  the software that relies on the platform
> being killed.

Yes.

> If evaluation of the port of VMS was not done, it means that the
> decision to kill VMS had been made closer to 2004 than to 2009 when
> dismantlement of VMS engineering began. (final firings were early 2010)

This part sounds like just nonsense. What has things done in 2004 have 
to do with anything? You can do a feasibility study at any point. DEC 
could have done such a study way back in 1995 for all we know, and HP 
could have done one in 2013. Does that mean that in 2012, you mean that 
the decision to kill VMS was done in 1995?

> What I was suggesting is that the group who meets/talks to HP about the
> future of VMS request some of the results (at technical level) of the
> feasability study to port VMS to x86.

Yes. And that don't mean a single line of code was written.

> The work to find a way to implement the 4 protection levels, interrupts,
> drivers etc may have already been done, at least in concept, and this
> would save a tremendous amount of time. No point wasting the work that
> was already done by the then very qualified VKS engineering team/person.

We already know this can be done. Studies of how to implement VMS with 
less than four levels of memory protection was done already in the 90s.
Just look at the VAX virtualization project done by DEC. It's mentioned 
in the VARM, and there is also a paper on what changes was required to 
VMS to make it run on this virtual VAX, which actually do not have four 
physical protection levels.

Or you can go to any number of papers written on the topic elsewhere. Or 
you can just, easily, do a design in a few minutes on paper on how you 
actually do this. I've even written about this in this forum in the 
past. There is nothing magical or tricky about it. The only question is 
how to maximize performance. Getting it to just work is trivial.

> It is also possible that work miught have uncovered a big show stopper.
> (although doubtful),

Yeah. There are no showstoppers. But it requires work. A lot of it.

> Would the group be hurt by asking for this ? If the worse that can
> happen is either "sorry, you can't have that" or "sorry, it doesn't
> exist", what have you got to lose ?

The problem is when people start thinking that DEC/Compaq/HP wrote some 
code to actually get VMS running in x86. Which does not exist.
Studies are a different thing than code.

	Johnny