[Info-vax] Any stronger versions of the LMF planned ?, was: Re: LMF Licence Generator Code

[Arne Vajhøj]

On 8/9/2021 12:39 PM, Bill Gunshannon wrote:
> On 8/9/21 8:34 AM, Dave Froble wrote:
>> On 8/9/2021 7:59 AM, Bill Gunshannon wrote:
>>> On 8/8/21 9:23 PM, Dave Froble wrote:
>>>> On 8/8/2021 7:07 PM, Bill Gunshannon wrote:
>>>>> Not exactly true.  While modern BASIC has c ome a long way it
>>>>> will never escape its roots which were not in the IT Production
>>>>> world.  Like Pascal it was intended to teach concepts and,
>>>>> believe it or not, one of those concepts was not programming
>>>>> per se.
>>>>
>>>> Interesting statement.  Coming from someone who admits to not being
>>>> familiar with the language.
>>>
>>> Misunderstanding, again.  What I am is not an expert in VMS BASIC.
>>> I have used BASIC since the Kemeny/Kurtz days.  I have done real
>>> production work using BASIC on everything from Micros to Mainframes.
>>> I have done business, financial and engineering programming in BASIC.
>>>
>>>> Perhaps older implementations of Basic were as described.
>>>
>>> I stated that modern BASIC had improved but that doesn't change the
>>> original purpose.  The idea of "fixing" these languages (like they
>>> also tried with Pascal even after the original author of Pascal
>>> gave them an alternative) is little more than a band-aid when you
>>> consider there were/are languages designed to do the work people
>>> try to do with these languages.
>>>
>>>> Some time back, as I recall things, some of the compiler people at DEC
>>>> asked the question, "why cannot every language be able to do what
>>>> others do?".  The result was the implementation of many new features
>>>> in Basic.
>>>
>>> Exactly.  "Lets put a band-aid on the language rather than do the proper
>>> software engineering task of choosing the right tool for the job."

>>> Simple: Choose the right tool for the job.
>>> BASIC, like Pascal was intended to teach certain concepts.  It was
>>> not intended as a production language.  Production languages existed,
>>> even when BASIC and Pascal were created.  In those days, new languages
>>> weren't just ego trips.  They were designed for particular tasks.
>>> They should be used for the tasks they were designed for. That is
>>> a major part of real software engineering.
>>
>> You seem to be implying that VMS Basic is not a "right tool for the job". 
> 
> Depends on the job.  I only have an inkling into just what the
> program(s) you do in BASIC are or do but I suspect from a real
> software engineering standpoint BASIC was the wrong tool.  What
> made it the right tool may have been just your familiarity with
> it.
> 
>>         Does your opinion (that's what it is) out weight the opinions 
>> of others? 
> 
> My opinion based on decades of research by people much better at this
> stuff than I am, but, yes, my opinion.  Does the CDC's opinion on how
> to handle COVID outweigh ordinary people's opinions?  They claim to be
> backed by "science".  I, too, claim to be backed by computer science.
> In the end, everyone is free to do as they please.  The discussion is
> purely academic.  I have done many things in many different fields of
> endeavor that I was told (usually after the fact) were impossible.
> 
>>           There have been and still are many serious applications 
>> implemented using VMS Basic.  Are all those people who use VMS Basic 
>> "wrong"?
> 
> "Wrong" is another of those words that can be more subjective than
> objective.  Could those applications actually be better if done
> in a more suitable language?  Probably.  Are they doing the task
> that was needed to be done.  Yes.

I guess you can take 3 different approaches to evaluating Basic.

The anecdotal approach.

Basic was created in the mid 60's specifically for teaching.

Dartmouth Basic was not that useful for real world usage.

And computer scientists soon started criticizing Basic. Including
Dijkstra "It is practically impossible to teach good programming to
students that have had a prior exposure to BASIC: as potential
programmers they are mentally mutilated beyond hope of regeneration."

And that must mean that everything Basic will forever be bad.

The engineering approach.

Does the language has the data types, control structures, IO
support etc. to enable writing business applications.

Dartmouth Basic did not.

But VMS Basic seems to have what is needed. Data types
including decimal. Conditional and loops. Sequential and
index-sequential IO.

There are a few warts: implicit type by last character in
variable name, lack of thread safety etc..

But overall VMS Basic is no worse than other VMS
procedural language.

Good solid 1980's technology.

And if one has to chose one of those traditional
1980's VMS procedural languages (Fortran, Cobol,
Basic, Pascal, C) then Pascal and Basic seems
by far the most obvious to use - they will result in
less more readable code than the other.

VB.NET Microsoft's latest incarnation of Basic
is a full blown OOP language with some FP support.

Good solid 2000's technology.

Microsoft's previous incarnation VB6 and VBS had
some limitations but were definitely usable as well.

The empirical approach.

VMS Basic has a piece of the VMS application development
market - I don't know exactly how how big - but while
Ada, PL/I etc. are gone then Basic is still supported, so
Dave is not the only user.

VB6 and VBS (in ASP) must have been one of the worlds
most used programming languages 1995-2005. It must have
been usable.

VB.NET is definitely lacking behind C# but are still widely
used. It must be useful.

++++

Trying to evolve existing languages into a different type
of language is not always a success.

I believe nobody in the Cobol community liked OO Cobol.

Fortran 77 to Fortran 90 was really an entirely new language
and a lot of users were lost in the process.

The Ada 83 to Ada 95 made the language very complicated.

The FP stuff added to Java 8 is not pretty.

But other are more lucky.

Object-Pascal and Delphi may not be super elegant addon
of OO, but the users liked it.

VMS Basic and the later MS Basic implementation also
managed to add totally new features (full OO for VB.NET,
somewhat OO for VB6/VBS) and be popular with
users.

Basic seems to be good language to extend.

So Jon Reagan - when do we see OO in VMS Basic?

:-)

Arne