3 ms·
I'm consciously avoiding Factorio because I know what it would do to me. It already took one of our team members. As for automation: I'm surprised that scaling
by Tade0 5y ago
I'm consciously avoiding Factorio because I know what it would do to me. It already took one of our team members.
As for automation: I'm surprised that scaling is presented as the main - or actually only - reason for automating. I for one don't automate for speed, I automate for correctness.
The other day we discovered that some of the tens-of-megabytes JSONs we're reading on a regular basis have an error: single-element arrays were replaced by that element, so e.g. `[1]` would become just `1`. That was okay for the parser on the receiving end, but not for the recently introduced more efficient serializing module.
I could just go over those files and search all such occurrences by comparing them to the reference type definition - would take less than an hour provided I didn't make any mistakes. But I'm only human, so I opted for something else, namely: drawing from the pile of rich manure that is the npm repository and finding modules for:
1. Generating JSON schemas from TypeScript type definitions.
2. Validating JSON files against JSON schemas.
Armed with these tools I modified the files until the validator confirmed full correctness. Took longer than expected, but now not only can I guarantee that these files are error free, but any future ones as well.
- mbrameld 5y agoI think scaling is presented as the main - or actually only - reason for automating because it's actually the only reason for automating in the game. Nothing that you can automate in the game can be done incorrectly.