zlacker

I have no objection to your overall sentiment, but, TBH, much of the reason they do this is it is trivially easy to use very authentic looking benchmarks to effectively lie. The net is that the academic ideal of a free market is effectively unachievable in real life - perfect information does not exist.

I don't know how to get around this problem of course.

>>TheIro+(OP)
I agree that we can never get to the ideal of a free market, but at the very least, we can ban business practices that are expressly designed to undermine the free market. We already have anti-trust laws that prohibit companies from some practices, such as price-fixing or false-advertising. We really need to strengthen such laws to ban modern practices such as benchmark-bans.

>>TheIro+(OP)
much of the reason they do this is it is trivially easy to use very authentic looking benchmarks to effectively lie.

...and that is what the companies themselves do for their own products when they show benchmarks, to make them look better than they really are.

>>TheIro+(OP)
Yes: lies, damn lies and statistics - but I feel like there should be some sort of truth that comes out of the aggregate of these things.

>>TheIro+(OP)
So it's only ok for Oracle to lie in their own benchmarks?

>>TheIro+(OP)
We can fix the problem of a possibly unfair benchmark with open source. Anyone can inspect the code of the benchmark used, build and run it themselves, criticize it or build an alternative.

>>TheIro+(OP)
> much of the reason they do this is it is trivially easy to use very authentic looking benchmarks to effectively lie

Have you ever paid attention to how GPUs are marketed, or the dirty tricks the drivers use under the hood? In spite of this, it's still pretty easy to find out which GPU at a given price point is faster.