andrewducker (
andrewducker) wrote2023-04-24 12:00 pm
Interesting Links for 24-04-2023
- 1. What did Diane Abbott think would happen when she claimed that Travellers, Irish and Jewish people don't experience racism?
- (tags:labour racism OhForFucksSake )
- 2. The Scottish Question: an overview
- (tags:Scotland politics law )
- 3. This is how "We won't need coders" conversations always sound to me.
- (tags:coding comic funny true )
- 4. A Simple Exercise to Strengthen the Lower Esophageal Sphincter and Eliminate Gastroesophageal Reflux
- (tags:exercise health )
no subject
They say "specification", in situations like this, but I suspect what they really mean is "incomplete and vague set of requirements". What they want to write is about the level of detail of "Let me manage my finances visually; make it easy to see [this or that kind of report]; under no circumstances [make some particular UI goof that really pissed me off in the last program I tried]".
By the time you've translated that into a full specification of what the program will actually do (as opposed to what it won't do, and/or how the user will feel about the experience), you're at least most of the way to code. But I think that is part of what the optimists want to have done for them – they not only want Magic AI to do the boring stuff like unit testing, debugging and wrangling arcane syntaxes and type systems, they want to wave their hands and airily foist off on Magic AI the whole problem of making their only-half-described desires precise in the first place!
no subject
no subject
Somewhere in my gigantic (~750) box of Nancy Buttons is one that IIRC goes,
"Writing a program that conforms to the spec is easy. Writing a spec that says what you mean is impossible."