2 ms·
My guess is the "independant discovery 1 hour earlier" wasn't truly independent. I'd guess that by messing with this stuff, it probably broke some internal sys
by londons_explore 7y ago
My guess is the "independant discovery 1 hour earlier" wasn't truly independent.
I'd guess that by messing with this stuff, it probably broke some internal systems that ended up firing alerts to engineers internally. Those engineers then 'discovered' the bug, and started fixing it.
Most systems at scale are designed to reject bad input, log it, but nobody takes any action.
A few systems have to process every record in order. For example, the billing system might go through every entry in a database table and add them to bills. If just one entry is malformed in some way and can't be added to the bill, it is retried. If there isn't success after a few retries, the whole process fails, and an engineer is paged to sort the problem.
I saw this kind of thing multiple times... You think you have fully sanitized every input, but someone always finds a way to add a 30 gigabyte surname to the addressbook, choose a profile image with negative dimensions, have a million devices share the same mac address, etc.