> New version handles 10x the throughput
Probably could have achieved 5x the throughput while using the team skills by just going from wsgi to asgi.
Two months to refactor something that already worked fine? The product owner must have been on vacation. That was a very expensive project that produced zero business value.
It produces little product value sure. However on the flip side the eng team basically spent two months of salary teaching this employee some incredibly important lessons around being pragmatic and crafting software that can be maintained by the team.
I've wasted many 2 months on way worst tbh, at least he got something used by users with little, but still some, improvement
I don’t know if I would see ”Now I'm the only go person so guess who gets paged for everything” as an improvement.
"better" is a vague term and working hours are limited so clearly some things are more worth than others but
It's very easy to make the wrong conclusion from a post like this. Better software is achieved through small decisions that compound over time. And bad software often happens because shortcuts compound too.
Well, at least if usage explodes it won't look like such a bad idea.
That's a really big if.
It is great to see a negative or neutral experience get some publicity showing the real consequences. We often see the publishing bias of a success story but rarely do you see 'did the thing and meh was the result' which is likely 99% of reality. Thanks for this!
[dead]
[dead]