alt Hacker NewsMaybe one day, but it doesn't look like we are very close yet. From the OP article, they handed it the working linux driver and asked it to just make this FreeBSD compatible, but it could not. Looks like it took OP a significant amount of work over 2 months to get something that seems to work.
What is interesting is it seems like the work resembles regular management, asking for a written specification, proof reading, etc.
Replies
plagiarist • yesterday at 10:53 PM
To make these things work you do need to write a spec and figure out what unit tests will prove it actually did what you want. Even then it will take a bunch of shortcuts so it's best if you're a domain expert anyway.
lazide • yesterday at 10:48 PM
Aka, the hard part.
> What is interesting is it seems like the work resembles regular management, asking for a written specification, proof reading, etc.
That's how I've been using the bot for years. Organize tasks, mediate between them, look for obvious-to-me problems and traps as things progress, and provide corrections where that seems useful.
It differs from regular management, I think, in that the sunk costs are never very significant.
Find a design issue that requires throwing out big chunks of work? No problem: Just change that part of the spec and run through the process for that and the stuff beneath it again. These parts cost approximately nothing to produce the first time through, and they'll still cost approximately nothing to produce the second time.
I'm not building a physical structure here, nor am I paying salaries or waiting days or weeks to refactor: If the foundation is wrong, then just nuke it and start over fresh. Clean slates are cheap.
(I don't know if that's the right way to do it, or the wrong way. But it works -- for me, at least, with the things I want to get done with a computer.)