Re: Rise of Node

You're describing two things. One is performance, the other is whether
Javascript can handle the complexities of sending back validation errors.
The former is a valid concern that should be tested and honed.

You could call an RPG program from Node.js that performs the data
validation by performing 8 SETLL operations and checking for the existence
of foreign-key relationships. But you appear to be advocating the use of
JavaScript to perform that type of data validation. Right?

The latter,

well, I am scratching my head. Why wouldn't Javascript be able to query
eight tables and use logic to determine what message to return to the user?

I think most people understand what 8 SETLL operations would look like in
an RPG program, perhaps each paired with a CHAIN or an API call to retrieve
an appropriate error message. Most people understand that the validation
might require a few milliseconds of CPU time, at most.

But it's not clear how you'd do that type of validation in a Node.js
application. What would the JavaScript look like? Would you try to validate
all eight foreign-key relationships in a single cycle of the event loop, or
design the application to use 8 cycles of the event loop?

Does that sound like a reasonable question to ask a Node.js developer? It
seems like Node.js developers are rightfully concerned about bogging down
the event cycle with too much work. Right?