Discussion about this post

User's avatar
gold's avatar

We will just move one more level up in the hierarchy. Once upon a time, code was code. Then it was assembler. Then it was source. In each case it was the formalism in which functionality was managed, where changes to functionality would be expressed.

As more code is generated, the asset will be in the specs and reqs from which the code was generated. Hell, we'll look at the source code the same way we look at machine or assembler code now -- very rarely and only if there's some fairly specific problem. We'll adjust the inputs and rebuild.

Will the specs and reqs become a liability? Well, sure. Or they certainly could be. But hell, we can make a mess out of anything, amirite?

Expand full comment
1 more comment...

No posts