[Info-vax] 64 bit DCL ?

[David Froble]

dodecahedron99 at gmail.com wrote:
> Sorry if this goes against the grain of those that love DCL but
> wouldn't it be better to let DCL die a natural death?

Since I have use for it, no.

> If it's time and effort being expended on 'OpenVMS Next', then I
> don't believe it's DCL that should get the attention

I would agree that VSI's main focus should be the OS and compilers.

> As much as DCL has brought OpenVMS along, it's time has well and
> truly come and gone as a modern glue script for OpenVMS

Just because someone says so, doesn't make it so, in general ...

Opinions are great things.  Everyone is entitled to at least one, and 
everyone thinks their own are correct, and other's are not.

> My vote would be to spend the time and effort into Python.

See above, re: OS and compilers.

But volunteers are great.

> A lot of the work has already been done getting python running on
> OpenVMS and I believe getting Python 3 onto OpenVMS is also underway.
> Having something like Python 3 as it's native CLI I think would have
> spin-offs longer lasting than investing more time into DCL

Don't force your choice down my throat.

Now, as a CLI, who could argue with options?  But the current users are 
using DCL, for the most part, because it's what they have.  Should they 
be pushed elsewhere, without it being their desire?  I think not.

> Powershell looks nice as it has built-in's for direct web output
> creation but Python can do similar through libraries and Python over
> time keeps incorporating more and more useful aspects into it's core

I might have to look at Powershell, if it works on VMS.  Read too many 
references to ignore it.

> Here's a rough comparison between the uptake of Python versus
> Powershell. Yes, I hate these types of comparisons as well but it
> gives an brief snapshot
> 
> http://www.techwars.io/fight/python/powershell/
> 
> My fear of introducing 64bit DCL into the current DCL image is:
> 
> 1. It will break something 2. It is time and effort into a scripting
> language that is dead and too limited to take OpenVMS into the
> future, IMO 3. It will not help promote the move on from DCL, just
> prolong it's entrenchment in OpenVMS

Perhaps it is DCL that is what some VMS users prefer?

I would agree, don't break anything.

The statement about "move on from DCL" is not a universal opinion.

> I'd rather see a clean cut so as to leave existing DCL as it is and
> invest time and effort into a universally accepted language that
> would also help the port of further code onto OpenVMS going forward

Volunteers are welcome.

> Then there is the topic of fostering rapid development on OpenVMS.
> Libraries libraries libraries are needed. At present there is simply
> too much old outdated libraries around and very little from an OO
> nature, even the C++ implementation left a lot to be desired. If
> Python is supported fully and developed on OpenVMS natively, then
> porting other libraries should be easier

There is at least one problem with using other people's libraries.  You 
also get their bugs and security issues.  Sometimes without the 
capability to determine if such exist.  I'm sure there are others.

Nor am I a huge fan of OO, just for the sake of being OO.

> I know it will be a lot of work to hammer out a full set of Python
> libraries for OpenVMS that could replace and go beyond DCL but IMO we
> don't want to just replace DCL, we want to better it and have a
> native language that can take us forward. Yes, Python lacks on the
> web side but I think it has more going for it than Powershell, as
> good as Powershell is. Python is OO based as well (PowerShell is too
> but not as fully implemented).
> 
> Having an OO based language as it's CLI should overtime help foster
> more robust scripts and more importantly, more script re-use and
> could foster the use of Python modules specific to OpenVMS. Imagine
> OpenVMS being filled with libraries that were not just examples to be
> cut and paste into your scripts but actual callable routines that
> were part of the OS itself that could be simply called :-). This is
> what gave lexicals their edge IMO, simple to use and call and could
> be called straight from the CLI. They were good enough for most
> things and that's the key
> 
> Scripting languages have become prominent because they are now good
> enough, same as why Linux and it's derivatives are seen as good
> enough when compared to OpenVMS. We need to not just play catchup
> with bringing OpenVMS into the modern world but to leapfrog ahead and
> IMO, that doesn't involve keeping DCL as OpenVMS's CLI

I'd agree with the concept.  I'm not so sure that I'd agree with what 
some propose as "better".

> We need to look beyond existing OpenVMS users (myself included) and
> their comfort levels but look to what will attract new people to
> OpenVMS. It's not going to be 64 bit DCL. People these days are not
> willing to invest time and effort into learning just another language
> unless they see it has longevity and possibility a job at the end of
> it. I've seen this trend in people asking on forums about whether or
> not they should learn language X or Y. Sorry, but DCL isn't going to
> get a look in.
> 
> Python might go someways to attracting people willing to port
> libraries and code to OpenVMS but DCL no matter how enhanced will not
> (unless it's ported to other OS's and the chances of that happening
> are less than zero!)

I've installed Python on one of my VMS systems.  I was able to print "Hi 
there".  That's as far as I got.  Maybe it exists, but I could not find 
any documentation or tutorial.  Perhaps when Python is as well 
documented as DCL ....

> OpenVMS needs to move away from being so different to Linux while
> keeping and enhancing what makes it awesome. Moving to a
> cross-platform scripting language like Python should go some way
> (maybe a long way) to helping with that task
> 
> My 2c worth

Perhaps it is that difference that makes VMS, in my opinion, so much better?

If VMS is my only platform, why do I care what *ix does?