Random factory reset

[Andy Fox]

After many years of owning my HDR T2 I had my first ever experience of a factory reset screen when I turned it on this morning.

All back to normal in about ten minutes using the custom firmware, so a big thank you to everyone who makes it possible.

 

[Black Hole]

For anybody stumbling across this thread:

The following links to a checklist of things to do to ensure a HDR-FOX is restored to its preferred state after a reset. It would be most convenient if adequate preparations are made (such as noting down the preferred settings, and/or installing defensive CF packages) before a reset occurs – that means NOW!

https://hummy.tv/forum/threads/is-there-a-good-time-to-retune.7624/page-2#post-103888

 

[/df]

Isn't it time to build @prpr's suggestion, which stops the box from going into the wizard when it gets confused, into the boot-settings package or at least reference it from the quoted post?

 

[Black Hole]

The problem (to my mind) is that this is not proven.

 

[af123]

The only thing that I think we know is that reformatting the flash configuration areas using the system flush update seems to reduce the frequency of these. My suspicion is that corruption triggers the problem, just skipping the process and continuing does not seem appropriate.
I don't think there is a check and repair utility for JFFS2.

 

[Black Hole]

reformatting the flash configuration areas using the system flush update seems to reduce the frequency of these.

While not wishing to be picky, this is pushing it. You can't make a statistically significant observation of a reduction in frequency from an already very low frequency, without an extremely long period of data gathering to compare rates of incidence between users who perform a regular system flush with a control population of those who don't.

 

[af123]

While not wishing to be picky, this is pushing it. You can't make a statistically significant observation of a reduction in frequency from an already very low frequency, without an extremely long period of data gathering to compare rates of incidence between users who perform a regular system flush with a control population of those who don't.

There are at least three reported instances of boxes which were experiencing very frequent resets (weekly or monthly) before flushing, and not afterwards. The Humax code drops into the setup routine based on values it reads from the configuration database; it is not much of a stretch to assume that corruption is involved.
Personally, I used to get one or two a year on each of my three boxes and have not had any for several years since I added a flush to my pre-Christmas preparation process. No, it's not statistically significant; you'll find all of think, seems to and suspicion in my post!

 

[Black Hole]

I'm not saying a flush can't help, only that it is difficult to prove it.

 

[/df]

The problem (to my mind) is that this is not proven.

I haven't had the Wizard appear when this setting was in place. Have other people had experience with it?

If someone has installed the CF and boot-settings, there's no way that the Wizard can improve their experience. Either the system comes back up with tuned channels etc despite what the Humax startup code thought, or it doesn't, in which case the CF user can choose how to fix the problem (restore backup, system flush, etc).

There should probably be a warning in the Webif, as for crashes, if the Wizard flag was set and has been cleared by boot-settings -- if only so that the CF user can give the box a smug look.

 

[Black Hole]

What if Humax chose to fix detected data inconsistencies by forcing a factory reset, which we then countermanded without fixing the underlying problem? The consequence could be vague unreliability rather than an obvious fault.