[Info-vax] RMS record metadata, was: Re: Re; Spiralog, RMS Journaling (was

[Jan-Erik Soderholm]

Den 2016-06-23 kl. 10:50, skrev lawrencedo99 at gmail.com:
> On Thursday, June 23, 2016 at 8:10:37 PM UTC+12, Jan-Erik Soderholm
> wrote:
>> But the core business applications doen't have those wildly dynamic
>> SQL needs, the operations are quite well defined and tested. And we do
>> not want the performance overhead with dynamic SQL, of course. Doing
>> things like the Python code below does is totaly out of the question
>> in high volume, high performance code and applications.
>
> You do realize the query only needs to be constructed once each run, and
> can then return a thousand or a million or a billion records, don’t
> you?
>

OK. You just makes me more convinced that you do not really know
what you are talkning about.

In a business application case it is usually a question of
changing a single (or maybe a coupl e)record in a database, and
doing it at a more or less high rate.

I also said that in an ad-hoc query and reporting scenario,
Python works quite well.

> Here
> <https://en.wikibooks.org/wiki/Python_Programming/Databases#Looping_on_Field_Breaks>
> is another example of a pretty traditional data-processing scenario:
> accumulating statistics on sales by employee.
>

Not anything that is done 10s or 100s of time per second.
Maybe once per day, or on-demand when someone wants to see it.