[Info-vax] Open source usage, Was Python and various libraries updated

[Arne Vajhøj]

On 8/13/2020 5:34 AM, gérard Calliet wrote:
> Le 13/08/2020 à 00:47, David Goodwin a écrit :
>> On Thursday, August 13, 2020 at 8:03:34 AM UTC+12, Arne Vajhøj wrote:
>>> To me the biggest problem with VMS and open source is not VSI
>>> but the community.
> It's a big problem, not the biggest. The two problems are mixed.
>>>
>>> There are way too few that are willing to spend time
>>> on VMS open source.
>>
>> If open source projects on OpenVMS are important then VSI probably 
>> needs to actively encourage participation. Perhaps if some of their 
>> own things were open source and easy to contribute to people in the 
>> OpenVMS community would get involved.
> It is my point. I just hope VSI could do the minimum so community could 
> take the opportunity.
>>
>> As far as the wider open-source community goes, unless big changes are 
>> made that's probably a lost cause. OpenVMS is too obscure and too 
>> proprietary for most to bother with even if they have heard of it and 
>> are aware its still maintained.
>>
> It is our responsibility to make OpenVMS known and liked. We can develop 
> a form of curiosity about OpenVMS. Young ingineers are often inquisitive 
> people. And for-long-time developments, reusability are becoming hype.
> 
> I think of this situation as something like a dead lock:
> 
> VSI doesn't help encouraging collaboration on Open Source, and thinks 
> there is not sufficient strength outside that can be usefull,
> 
> The community - little for now - doesn't engage with strength on 
> collaboration, and is discouraged by the lack of interest on 
> collaboration by VSI.
> 
> Something has to be done to unlock this. Community License is a very 
> good point. Making more standardfully available Open Source developments 
> will be the next good point.
> 
> More generaly VSI and the community have to quit the old paradigm of the 
> marvelous company and the passive users. We have all to rebuild trusted 
> collaboration, and adopt some ways from the Open Source paradigm. And we 
> have also to be more conscious of the specific qualities of OpenVMS 
> which can be "selled" to other people.

There are some things VSI can do. Some things they can do alone
and some things they can do together with the VMS community.

But VSI does not have the size to contribute large number of
engineers to open source work.

Google can put 1000 extra engineers to work on open source and
it will be white noise in their financial result.

VSI can't.

What VSI could do and in my opinion should do:
* make a public pro-open-source statement by senior management
   to send the right signal
* have a person spend some time lobbying open source projects to
   include VMS fixes in main distribution
* establish some processes for getting VMS fixes for open source
   done by VSI out to the community and getting VMS fixes for
   open source done by the community back to VSI (as workaround
   until everything hopefully eventually will end up in main
   distribution)

 From a practical perspective a lot of this could become much
easier if VSI created some Github repo's.

But now we start talking about something that requires
engineering hours that VSI may not be able to allocate
right now.

And to repeat myself: in my experience the VMS community
is not very willing to spend time working on open source.
I suspect that one can count the number of people that has
made VMS open source contribution within the last 5 years
on 2 hands.

So do not expect miracles even if VSI establish some
framework to assist the community.

Arne