[Info-vax] C99 updates to CRTL

[Dave Froble]

On 7/31/2019 8:29 PM, Simon Clubley wrote:
> On 2019-07-31, Arne Vajhøj <arne at vajhoej.dk> wrote:
>> On 7/31/2019 7:24 PM, Stephen Hoffman wrote:
>>> Interesting.  Y'all like RMS though, right?  Because RMS allows you to
>>> use different apps to access the same (RMS-based) database files.  That
>>> makes RMS and its interpretation of the file formats... common.
>>>
>>> RMS is stuck about thirty years back in terms of higher-level access,
>>
>> Hmmm.
>>
>> I would also call the RMS SYS$ calls which is what you have to
>> use in C/C++ a bit cumbersome.
>>
>
> If you were British, I would call that an excellent example
> of British understatement. :-)
>
>> But the Pascal interface is in my opinion as good or better
>> than interfaces to modern NoSQL databases.
>>
>
> How well do the current crop of NoSQL databases handle resizing
> fields in a record and adding fields to a record without having
> to rebuild programs or having to make source code changes to allow
> programs to continue using the existing data ?

Well, that's a good question.  It depends on several things.

In my DAS database product, changing data record layout is possible, it 
required a copy of the data records, and there are utilities to perform 
these tasks.  But no, not on the fly, as might be possible in Rdb and such.

Data definition is part of the database.  However, nothing stops a 
programmer from hard coding a data length.  Nothing stops that in 
databases that allow on the fly data field resizing.

Even when allowed, what happens when a program uses information from the 
database for data lengths, when the length is changed while the program 
is running?  That's a rather slippery slope.

Maybe some "features" can be rather dangerous?

-- 
David Froble                       Tel: 724-529-0450
Dave Froble Enterprises, Inc.      E-Mail: davef at tsoft-inc.com
DFE Ultralights, Inc.
170 Grimplin Road
Vanderbilt, PA  15486