[Info-vax] OpenVMS in the future, Open sourced or Closed? Intent is to keep it...

[Stephen Hoffman]

On 2015-05-02 14:29:55 +0000, David Froble said:

> People like to throw around names for things.  Sometimes others aren't 
> very clear what the names mean.
> 
> To throw a question back at you, I'll ask, just what do you mean by 
> "open source"?
> 
> 1) Free software?
> 
> 2) Source code available to those who want to make modifications?
> 
> 3) Something else?
> 
> As usual, the devil is in the details.  So perhaps you could expound on 
> the details of what you are asking for?

Ayup.  Who is going to have the time and skills and servers and build 
environment and the funding to work on the code, too.   Linux started 
out and built a base in another era.

What would attract developers to OpenVMS?

Open-sourcing some software package does not magically instantiate 
herds of programmers ready and willing to work on the code, for that 
matter.

There's a huge pile of people and process and infrastructure required 
before that's even particularly feasible, too.

I'm one of the few folks around that has ported the existing OpenVMS 
build environment to another cluster, and that was a couple of weeks of 
effort to get the code transferred and databases and the rest of the 
baggage relocated and configured and running for just creating partial 
builds — which is what I needed, not full system builds — and I knew 
how the then-current OpenVMS build worked.

The source code and the build are just the tip of the effort involved 
in open-sourcing, too.  A successful project needs bug- and 
change-tracking, source code reviews and tracking, distributed source 
code management, getting legal reviews and open-source reviews and 
individual source code contributor legal sign-offs, documenting the 
build and the testing environments, and a whole host of details.   
Somebody to oversee and coordinate the efforts, too — without peeving 
too many of the potential contributors.  All surmountable, of course.  
None of these will be quick, and this would not be a cheap project.

But finding a bevy of programmers willing and able and self-funded to 
work on OpenVMS, and in numbers large enough to matter?    That's a 
tougher problem.

At present, the SourceForge efforts involving open-source code on 
OpenVMS are quite small.   There's some work underway to make 
contributing to open source on OpenVMS easier.  Right now, just the 
set-up effort can be a complete pain in the ass, and I know how to deal 
with (most of) that.   Many folks just won't be interested.

> Keep in mind that VSI may not be able to do some things with the code 
> they got from HP.  There also might be (I'd hope not) restrictions 
> agreed upon new code written by VSI.

Ayup.   As much as having the option and the source code access would 
be nice, HP and VSI undoubtedly value and are interested in preserving 
their current exclusivity.   HP probably at least through 2020 or 2025 
or so, per their roadmap.

 Having code escrow would be an interesting approach to the issues that 
some folks here are probably most concerned about.  But that's fodder 
for another time.


-- 
Pure Personal Opinion | HoffmanLabs LLC