[Info-vax] Free Pascal for VMS ?

[Stephen Hoffman]

On 2018-05-12 19:46:16 +0000, seasoned_geek said:

> On Saturday, May 12, 2018 at 11:38:36 AM UTC-5, Stephen Hoffman wrote:
>> On 2018-05-11 00:27:59 +0000, seasoned_geek said:
>>> 
>>> Most business applications existing today on OpenVMS are littered with 
>>> API calls.
>>> 
>>> I wrote the only C++/OOP application in existence at one Fortune 50 
>>> OpenVMS client. Everything else is BASIC and FORTRAN. Mostly BASIC.
>> 
>> You're discussing the past.  That's dead and gone.  It got OpenVMS 
>> where it is, and it won't change the trends.   The question for VSI — 
>> beyond keeping the installed base reasonably happy — is about 
>> increasing usage of the platform, and procedural app designs are not 
>> where most developers folks are headed.  Then there's that the itemlist 
>> is an utter disaster of a design given
> I'm talking about the installed base still running production for 
> billions if not trillions of dollars in business each year. It's not 
> dead. It's not gone. It will be running payroll, doing accounting, 
> processing orders, managing warehouses and a hundred other tasks long 
> after both of us are dead and gone.
> 
> Calling it the past is calling it abandoned, yet, it is the only thing 
> funding VMS.

Keep doing what DEC, Compaq, HP and HPE very reasonably did with 
OpenVMS.  Keeps most of those existing apps and existing folks from 
porting, too.  That's probably a nice business for a decade or two, 
too.    This if you're not looking to attract new folks and new apps.   
If you follow the existing and longstanding approach toward the 
installed base and toward complete upward compatibility, there's no 
reason to expect a change in the longer-term trends for OpenVMS.   This 
because — as I've occasionally commented around here — making the shift 
from the installed base to new apps is the most important task for the 
VSI folks past a successful port, and it's bigger than the port in the 
long term, and the key to the longer-term survival of OpenVMS.   Why?  
Because I'm continuing to encounter folks porting apps off of OpenVMS.  
Folks that know about VSI, too. That's not a sustainable trend for any 
commercial operating system platform.  Not without new applications to 
offset those folks departing, or those apps that are being retired and 
replaced.  Expect that the folks at VSI will be advertising whatever 
new apps and new ISVs and wholly new deployments they can discuss, too.

While capricious API breakage is a very bad idea, offering complete 
compatibility was not and is not a good trade-off, long term.  Which 
means trade-offs; decisions to keep the old mistakes and eschew new 
work, or to replace and deprecate and break.   Because the VSI budget 
and the available schedule time and available staffing are not infinite.





-- 
Pure Personal Opinion | HoffmanLabs LLC