Hacker News Reader

Comment by joselreyes | Hacker News Reader

I actually think Apple’s repair system is one of the reasons they have success. Most customers can’t repair their stuff so having a trustworthy place to go to is a really nice safety blanket. Compare that to most other phone or computer manufacturers where you have to ship your stuff or do it yourself and risk voiding warranties. I imagine eventually Tesla will offer an Apple-esque “Certified Repair” service option to increase their ability to help customers with non-Tesla employees.

View on HN

I think the biggest issue with Apple's "repair" model is the amount of repairing that actually goes on. As a result of the increasing board integration of Apple computers and Apple's repair policy, a single resistor kicking the bucket means a full board replacement that will set you back a good fraction of what you paid for the computer in the first place, at an Apple certified repair center. Even worse is the outright hostility to 3rd party repair shops who instead opt for (more difficult), but ultimately less expensive (for the custom and the environment) component-level repair. You won't be getting any documentation or spare parts from Apple.

I don't think I've had a single resistor repaired in any of the thousands of servers I've owned in the past 20 years.

I picked a component off the top of my head. Water damage is almost always the culprit in these repairs, which doesn't affect most servers (except for these [0] maybe?) Louis Rossman's channel on Youtube [1] can give you better information than I can.

[0]: https://news.ycombinator.com/item?id=17444076 [1]: https://www.youtube.com/user/rossmanngroup

I dont think you shipped a whole Rack back because a single fan was broken either.

Fans are a FRU. The complaint was about sub-fru items.

There are no sub-fru items in Apple ecosystem, in fact there are no frus either.

I’ve had capacitors repaired.

Haha, of course you did - It's always a capacitor![1]

These mature cheap reliable mass-produced electrolytic capacitors of which we rarely see fault today, sure had quite the rebellious stage[2] [3].

Interesting, from the announcement on badcaps.net[3]:

"Back in April 2002, I had an idea to start a business (actually spinoff of another business) servicing motherboards at component level amid the mass capacitor failures plaguing the PC industry."..."The store boomed from its launch all the way up until about a year ago. Over the last 6 months or so, I have sold very few capacitors. Once my stock depletes, I will be shutting down the Badcaps.net store."

That wasn't an accident, but some company trying to cut corners on what was a well-understood manufacturing process.

Your servers probably aren't running at 95C under load and have all components ventilated.

You probably don’t bring your server to hazardous environments where it can get drinks spilled on it.

For laptops this kind of environment is common; in fact most of Louis Rossmann’s repairs are about liquid damage.

While I agree the lack of repair documentation and available spare parts is a concern, Isn't the whole point of integration (at least on a chip/board level) to increase reliability?

I'm a mechanical engineer masquerading as a software developer so I could be talking out my ass, but I'm fairly sure that board failure rate is proportional to the amount of discrete components on said board.

It is hardly ever the boards themselves that fail but the connections between them and that is the reason for further integration.

Reply on news.ycombinator.com