[scaphandre]

I had also have a slightly unresponsive home button, on an iPhone 4.

Try running the Apple Calc app, then hold the power button until the 'slide to power off' prompt appears and then let go and hold the home button gently until the calculator app force-closes.

A previous reddit thread advised that procedure, apparently to trigger some 'recalibration'. It's possible it's placebo effect, but it appears to have slightly helped for my home button. http://www.redmondpie.com/how-to-recalibrate-iphone-home-but...

[WalterBright]

I can attest that using denatured alcohol, which I use for cleaning lenses and scanners, will destroy your ipod.

[mikeash]

It works even better if you do it while hanging upside down and while your spouse or significant other sacrifices a live chicken.

Seriously, it's a physical switch. There's no "calibration". "Recalibrate" a physical switch by doing some magic ritual in the Calculator app? If you believe that worked, I have some really high-quality HDMI and SATA cables you might be interested in purchasing, for very reasonable prices.

[scaphandre]

Ha - I agree that that procedure seems preposterous. I have seen it listed in several forums, and only considered it after several others had reported success. I figured it didn't cost much to empirically give it a go. I present weak evidence that it appears it may have helped for me.

For what its worth, I am doubtful. There did appear to be some subjective improvement in responsiveness, but I didn't benchmark it. It is not the Calculator app in itself that functions as this, I had originally read any stock Apple app would do it.

Software reset does fix this for some, so I don't think that some button state polling problem relating to software is inconceivable. Possible, but unlikely.

More info here - they acknowledge it is likely bs. http://lifehacker.com/5909709/recalibrate-your-idevice-home-...

[st3fan]

Here is a theory. Switches, specially push buttons, need debouncing in software. If a switch has problems then the characteristics of the debounce might change. Which could possibly be fixed by recalibrating.

(Not by holding a switch down of course, you would need to push and release the switch many times to see a debouncing change)

[mikeash]

Poor debouncing could conceivably cause a single push to register as two, or vice versa, but it's not going to cause a push to fail to register at all, which is part of every complaint I've heard so far.

I also see no reason that debouncing would require calibration. Basic debouncing just ignores any additional events that occur within a set period. That would just be set at the factory. There's a huge gulf between the timeframe of multiple events occurring due to switch mechanicals and the timeframe of legitimate ones.

Finally, the idea that somehow the Calculator app triggers a secret magic calibration step is just absurd. Even if you can somehow swallow the idea of software calibration of a simple mechanical switch, that detail makes it clear that this is nonsense.