Skip to content

Comment on How a few screws cost $2000 and a 240GB multinodes cluster cost $50parent

Comments

I disagree here. While it is a testing component, part of any valid test is that it be replicable. If you are running down to 'Home Depot' and picking up some random screws with who knows what specifications, how exactly are you going to replicate the design here? Perhaps they will never need to do it and it is a genuine one-off design, but putting on the line a potential multi-million (billion?) dollar project? In the grand scope then, spending $2000 (which I doubt WAS the real cost here) on some screws just gets a shrug out of me.

The safety cover is not part of the device under test and I doubt it interacts with it except during a catastrophic failure. It does not need to be replicated exactly the same in a future test.

Yeah sure - but where do you stop? Where do you draw the line and say "this bit doesn't need to be space-rated"?

Reductio ad absurdum says the table it sits on is "part of the test", and the power supply it runs off is "part of the test", and the engineers car that he uses to get to work to run the test is "part of the test"...

I'd _strongly_ advise that a screw that hold together the enclosure of a load-simulating-test-jig is _not_ "part of the test" in the way that requires ISO9000 compliance (or space-rating).

I don't think the OP ever said it was a mistake. He said he wouldn't make that choice, but that doesn't mean he thought it was wrong for the company.

Instead he contrasted it to what you can do today as a startup.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.