Chester (lamchester.delete@this.gmail.com) on November 24, 2020 11:45 pm wrote:
> Foo_ (foo.delete@this.nomail.com) on November 24, 2020 2:34 am wrote:
> > Chester (lamchester.delete@this.gmail.com) on November 23, 2020 2:15 pm wrote:
> > > - 7z compression/decompression - 7z is far more popular than xz
> >
> > As already pointed out elsewhere, 7z and xz both use LZMA under the hood.
> > Of course, 7z may use more than LZMA, but talking about them
> > as if there were disjoint workloads doesn't make sense.
>
> Fair enough, they both use LZMA under the hood. That makes xz a touch
> more relevant. But then, why not test both like Phoronix does?
>
> > Besides, saying 7z is popular is extremely one-sided. 7z isn't popular at all
> > on Linux for example. And while Linux is not very common overall on the desktop,
> > it's quite common on the desktops of some categories of professionals.
>
> I was mostly focused on people buying machines for home use.
>
> But yeah, professional use is a completely different story. From what I've seen,
> companies will sample hardware and run their specific workload before buying
> a batch of servers. So review sites are probably less relevant to them.


The few large companies can afford to buy samples of possible choices or they some times may obtain access to sample hardware from the vendors. Therefore they may do their own tests and choose correctly.

The many small companies and individuals cannot afford to buy samples of hardware that might prove to be a bad choice.

That is why review sites are valuable, even if they very seldom perform the tests that e.g. I would do on hardware that I want to buy. So one still must use the available published tests to roughly estimate the results of the tests that one is really interested in.