Re: Get a definition for a constant in a source member into a CL program

On 2017-06-10 11:38 AM, Mark S Waterbury wrote:

...
The key phrase here, (and so, my "quote of the day") is, "Premature optimization is the root of all evil." Hard-coding such symbolic values in programs is just another form of "premature optimization." (It's a "Pay now, or pay later" kind of thing...)

That "premature optimization" quote is one of my faves! It's one of the quotes listed on my work home-page.

But I don't agree that hard-coding symbolic values counts as premature optimization. I don't actually even understand the statement enough to argue against it. Using named constants just seems like obvious goodness to me.

Ohhhh, following the policy of "think before clicking send" ... Are you saying that it would be better to use a database to store all the values that we currently use named constants for?

I guess that could make sense for some "constants", but it could lead to serious hoop-jumping in some cases, such as using constants for definitions. And if overused, it could lead to one of the 3% of performance issues that do need to be dealt with.

And I think it would only count as premature optimization if there was already a system or policy in place to get "constants" from a central place like a database or srvpgm, and a programmer decided to just go with a named constant.