>write some beautiful piece of code to solve a hard business problem
>"anon that's too complicated, why dont you just do this instead"
>"because if i do that some of the output data will be objectively wrong"
>"its ok we dont care, we dont really use those fields anyways"
i hate it here

>>105583073 (OP)
case of maintainability > performance?

>>105583
>>105583073 (OP)
Remove those fields then?

>>105583321
it's not even maintainability vs performance, it's "maintainability" vs correctness
what they want me to do is LITERALLY wrong

>>105583580
yeah thats kinda what i meant
idk id think of it in this terms: "the client is always right"
theyre the ones paying for the product, and they will be the ones using it
if they want broken shit, give it to them
you could add a readme describing the behaviour or something so that the next maintainer knows whats up if you wanna be extra-thoughtful

>>105584177
>if they want broken shit, give it to them
nigger

>>105584177
You have never had a job. If it's broken by their request you still get sued, not them.

Why are you even writing code in the first place?
There's AI for that.

>>105584185
>>105584210
>You have never had a job.
not in IT
but if youre liable then ask for the request to be put in writng and youre covered
when you ask for that make sure to record the conversation so that you have leverage

idk thats what i would have done
>dodnt work in it
and by the sounds of it i dodged the mother of all bullets

>>105583580
Stop caring about theoretical "correctness"

>>105584210
You're a moron

>>105584296
i think hes just an OCD autismo

>>105583073 (OP)
Sometimes mildly wrong but well understood is better than completely correct but a conceptual black box.

I've had these types of things happen, and usually I'm missing something. Here it may be
>we don't want these fields to become actionable right now
>the wrong version is someone's vanity "pet" thing and fucking with it would piss them off
>maybe fixing it would best be done later as "subsequent" work and included as a sure thing into a clusterfuck of other tentative things
>maybe your code is absolutely schizo autism incarnate and this is just a more polite way of telling you

>>105584247
i shit in your glorified chatbot for technologically illiterate cattle
any more questions?

>>105584177
>if they want broken shit, give it to them
yeah at the end of the day i'll have to do that. but it's just stupid.

>>105584779
i agree
but its how it is
there might be also other considerations at play
like i said, maintainability
if you wanna do things right maybe you could set unused fields to null
but talk that through with your lead/boss obviously
the unintended changes to fields corrupt the data anyways, making it worthless
so you might as well set it to zero in your operations