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

[gérard Calliet]

Le 10/08/2020 à 19:52, Dave Froble a écrit :
> On 8/10/2020 1:12 PM, gérard Calliet wrote:
>> Le 10/08/2020 à 17:25, Arne Vajhøj a écrit :
>>> On 8/10/2020 11:09 AM, gérard Calliet wrote:
>>>> Le 10/08/2020 à 03:50, Simon Clubley a écrit :
>>>>> On 2020-08-08, Arne Vajhøj <arne at vajhoej.dk> wrote:
>>>>>> And open source could be used for some of he truly critical
>>>>>> stuff. Linux is becoming very common for SCADA today. And
>>>>>> they could use GNAT for software development.
>>>>>>
>>>>>
>>>>> That reminded me that VSI hasn't said anything recently about the
>>>>> possibility of Ada support coming to x86-64 VMS.
>>>
>>>> Last meeting (UK users group by zoom). Bret Cameron: we have no plans
>>>> on Ada.
>>>>
>>>> Since 2015 I had to rebuild Gnat Ada for Itanium OpenVMS, I didn't
>>>> get any collaboration from VSI (except from individuals).
>>>
>>> VSI has limited resources. They have to prioritize.
>> Agreed. But.
> 
> One has to feel that VSI understands their priorities better than anyone 
> else.  For example, look at the inclusion of DECnet in x86 VMS.  VSI 
> went to their customers, and they listened.
> 
>> It's now for more than five years I'm only saying "help us to help you".
> 
> Ya know, I really do agree with this concept.  So, what's it take for a 
> little bit of help for you?  Then a little bit of help for me? Then a 
> little bit of help for *
> 
> I do understand their refusal to start down this rather steep slope.  At 
> least until the port to x86 is completed.
> 
> I offered to provide some help with the lock manager.  But I understand 
> their position of "no new anything until it is ported".
> 
> Your, and my, help to VSI takes some of their valuable time.  That's the 
> reality.
I have to say: 1° snip (don't no why they say that when something is 
hard to be said ("ils sont fous ces anglais"))
2° Your help is not like my help :) : you propose something where they 
will have a lot to do to take your help, I speak about helps where their 
effort is a lot lighter.

> 
>> Ada cannot be a priority. ok. Some invest in Ada for you. Take your
>> phone, perhaps for a very low effort something can be done and worth it
>> for all.
>>
>> It's my general opinion on interest of VSI being more fair in the Open
>> Source realm. Just do a little bit more to gain trust and help from a
>> large enthousiast (and new) community.
> 
> See above about "the port isn't everything, it's the only thing".
> 
>> The times have changed - it's a pain for us thinking about the huge
>> success of internet compared to the Digital decline - and business
>> around Open Source paradigm proved it to be very good.
> 
> Depends.  For generic stuff, yeah, great.  For specific stuff, no.
> 
>> *If* you play the game as it is played by the majority. Is someone using
>> now MySql? No. MariaDB. Why? Oracle didn't play the game.
>>>
>>> People can disagree on how they prioritize, but everybody
>>> should be able to agree that they have to prioritize.
>>>
>>> And I understand why Ada is not a priority. I believe that
>>> the Ada approach to safe programming is good. But the world
>>> chose a different path many years before VSI was founded.
>> Have a look at how Adacore business is good and growing. And  the huge
>> role of community in their success. They play the game.
> 
> Yet when I look around, no Ada.  Users are not monolithic.  What is good 
> for you is not necessarily good for others.  Well, except the port, 
> which is essential for everyone.
> 
1° snip
2° see above "you speak about what you know, as I do"; no Ada you know, 
some Ada I know which could be sufficient to have a little company doing 
things for that; and taking a miminum help (just for example saying "we 
know x who does something: one sentence); and a day another customer I 
don't know but that VSI knows says: "I need Ada, elsewhere goodbye".

1° snip
2° can I try something not all the computer scientists (they talk only 
with computers) are able to do: understanding a chain of ideas, a little 
bit long and abstract?

The problem is about "the port is everything". I don't agree and I think 
this theorem is false, and causes of a lot of troubles we are in.

a) in the beginnings the port was a mean, not an end. *if* (i) we want 
to go ahead on VMS we have to (ii) port to x86. {the mean (ii) and the 
end (i)}

a') the mean is easy to describe and cannot be thought of as easy and on 
a short journey to get
a'') the end is very difficult to describe (what are the reasons we 
think it is good to continue with VMS) a lot more difficult to expose, 
and in a contradictory situation: *because* VMS is a long term solution 
we have to wait now in a difficult situation with it {because it is 
good, we can support it is not so good *?*}

b) now our end is that the mean will be achieved, because... I forget

c) our end is the port: we say to the customer "you have to choose VMS 
*because* it will be ported to x86" {at that point the customer thinks 
we propose him a tool which is good *only* because it can do later... 
what all the others tools can already do}

d) because the hypothesis places us in a contradiction, it is a false 
hypothesis, so: we cannot say the port is everything.

 From 2014 to today and from today to 2022, the port has not, cannot be, 
will not be everything.

And the marvel is that althought we think the contrary, we know that the 
port is not everything. The real good things are the second lives of 
Alpha environments, the rebuild of a real support, the entousiasm of the 
old guys... and the *way* of quality the port is made. It is a complete 
revival of a computer ecosystem, with a myriad of topics involved. And 
just at its beginning. The current possible trap is confusing the mean 
and the end, observing only the -in-two-years- mean as a end, and 
stumbling on a gap our obsession prevents seing.

Thanks for the opportunity :)

Gérard Calliet