This SHOULD NOT be news.
Maybe you guys should revisit your school books.
No, this comment is not "too harsh" or elitistic.
This SHOULD NOT be news.
Maybe you guys should revisit your school books.
No, this comment is not "too harsh" or elitistic.
Would you be ok if the people who worked on your house didn't know about construction basics -- and worse, created stuff that's widely known to be ineffective?
> Customers PAY that company for their services.
If the impact of poor programming is material, customers will send their money elsewhere. Customers evaluate products in a marketplace; unless the sale explicitly mentioned the developers' qualifications, it's not like they are defrauding customers.
Once again, it's OK that you (person, programmer) don't know that. But as a company, there should be a strong guarantee that you (company) have the knowledge to build a good product. It's fine for a company to have some people who don't know this stuff, but not to not have anyone capable (and in a position that allows them to) notice these kind of mistakes.
> Maybe programmers should educate themselves in the basics before working on products people depend on?
That remark clearly puts the onus on individual workers and there is no qualifying statement that anyone who doesn't know this particular thing should be allowed to work, under supervision.
Ever heard of the principle of charity? Programmers should educate themselves, and that includes senior supervision. And they should make sure they used what's best for the job when the ship while still learning -- and if they're unsure, they could check with someone...
Nope, didn't wrote off any of those. In fact I encouraged them, saying programmers should educate themselves (and all of those are ways to do that).
What I did write off was letting selling ineffective staff by interns, apprentices, on-the-job trainees ship to customers...
Do customers pay for a quality product or to have whatever done by cheap or unpaid internets shipped to them?
any serious hoster filter out ddos with a router. routers are basically build to test IP against subnets, so they can basically run each IP address against a million of hardware subnet testers all running in parallel and decide what to do depending on the result.
And if you don't have a router, any OS worth its salt already have a radix tree implementation for its routing table. The only thing you need to turn a Linux routing table into a ddos filter is to enable reverse path filtering and then add blackhole routes with iproute2.
Guess what, sometimes it's needed. Even knowing that a given data structure exists is sometimes enough.
It's 5000 times faster than an sequential list lookup.
It is a good discussion of radix trees, though.
But it's not OK for a company to offer a service built in such a naive way (naive from a comp-sci point of view). To me this kind of mistakes are a symptom of having a poor CTO, poor hiring/dev practices, etc.