Hacker News Reader

Comment by KarlKemp | Hacker News Reader

The mechanism I described isn't one to disagree or agree with. It's math: For <x> even parts with chance of failure on any given day <y> (including some variability <z>), the expected first failure is at some point long before <y>, because

    P(no parts fail within the first two years) = 
        1 - P(<y>)^((number of parts) * (number of days))

Improvements in <z> have a huge impact here, quite obviously.

View on HN

What you provided isn't (just) maths, it's a modelling assertion: that failures are independent from part to part, and independent from day to day.

I'm skeptical about your first assertion -- think of the effect of a power surge. I don't agree with your second assertion at all, from my experience of cheap devices that fail not long after their warranty expires.

> The mechanism I described isn't one to disagree or agree with.

And yet you failed to account for the embodied energy in building the product and the fact that maintaining any appliance or equipment requires repairs. So there is quite a bit to disagree with.

It is much more wasteful to build a product with no methods of repair, noting that a large number of components will statistically fail due to variability of tolerances or usage patterns.

A 50 yr product requires maintenance, replaceable parts and access to documentation. All of these are lacking in our modern consumer environment. You can blame this on complexity, or designs towards sleekness or what not but the truth is that product failure after a short amount of time (planned obsolescence) creates a reoccurring customer base, and the ability to pass off cheaper parts that don't need to be able to handle removal (glue vs. screws, etc.).

Sure but its fine for individual parts to fail if the manufacturer will repair/replace that part. Spares of easily user-replaceable parts can even be included with the product, like spare fuses often were.

It's not necessary to achieve zero failures, but it creates incentives for manufactures to use higher quality parts.

I am supportive of setting minimum lifespans for certain classes of products and to expwct manufacturers to be liable for that performance.

it is not the main point but requiring the manufacturers to repair an item after 50 years cannot be done.

the factory that built that refrigerator heat pump from the 70' is unlikely to still be making it.

as of today it would be already a big improvement if the was a law saying: "if a part is no longer produced or sold by the IP owner then it is legal to recreate and sell clones of that part"

> the factory that built that refrigerator heat pump from the 70' is unlikely to still be making it.

Yes but why? Why do we have 400 different models of heat pump for every fridge out there? I can understand that once every decade or so a new tech comes out that's worthy of developing parts from. But even then the new parts could be retrofitted into existing designs.

> "if a part is no longer produced or sold by the IP owner then it is legal to recreate and sell clones of that part"

That would be a huge improvement. What would be an even better improvement would be to abolish so-called IP entirely and to mandate actual durability for all parts sold. Think about lightbulbs: the parts/socket are standard (no IP involved), every factory can produce them, and yet we still have crazy bad durability.

Reply on news.ycombinator.com