[ohgodplsno]

> Benchmarking isn't forbbiden. What's forbidden is public disclosure of the benchmark results, unless there is enough information for somebody to replicate the tests.

All good then. You're totally allowed to taste the food I cooked to see if it's poisoned. You're just forbidden to tell anyone or show symptoms.

What the actual fuck is wrong with people actively defending these blatantly anti-user behaviours? Google won't pay you for it, you don't need to shine their boots.

[jsnell]

I find it strange that between posting stuff that was just blatantly untrue vs. fact-checking the claims, you find the fact-checking to be the offensive part.

If the author would just have told the truth, there would have been nothing to reply to. Unfortunately telling the truth would have generated a lot less outrage, and thus a lot less engagament for tweets and the HN submission, since the reality was so much more reasonable than what they claimed. For example for the case of benchmarking, what's in this TOS is actually a lot more permissive than e.g. the benchmarking clause of the AWS TOS which nobody has problems with. It's just a blatant lie to equate that to "no benchmarking".

This shit took space on the HN frontpage that could have gone to something cool instead. It also has zero chance of affecting any change because it was all untrue. Like, the entire reason to complain about things like this is to use the court of public opinion to get it fixed. But in this case there is no fix to be applied, because the terms are substantially different from what the author described.

[SecurityLagoon]

It's to stop people claiming poor benchmarks without providing any way for people to replicate and confirm. It's the way that basically all benchmarks are done to be taken seriously.