Nick wrote: > > It both motivated me that I should get an ARM laptop > > So firstly, the Chromebook Pixel isn't an ARM. Some other > Chromebooks are, but this one is x86_64. Oh! Sorry. I had thought all of the Chromebooks were ARMs. Knowing that it is and amd64 architecture isn't quite as exciting. :-) > The exciting things about this laptop for me are the screen > primarily, Wow. 2560 x 1700 at 239 PPI touchscreen is quite impressive! I can see the attraction. Even though it is widescreen it has some vertical pixels. > and also the freeness of the firmware [2] (though I lack the > expertise to take advantage of that, which is a pity as there are > some changes I'd ideally like). Hmm... It uses the free Coreboot. That is nice too. But only parts of it are free. Better than nothing to be sure. > > ... that dev-mode was stored in "battery-backed CMOS". But > > battery-backed CMOS should survive a main battery used to zero. > > Those are all new enough that a CMOS battery shouldn't be dead > > year. Therefore a main battery used to zero shouldn't seem to > > brick the unit. Is that not true on the Chromebook? > > The battery loss thing scares me too. It hasn't happened to me, > despite the battery completely running out on me quite a few times > (that was before I learned about the potential danger). All I have > to go on is the post I linked, plus a couple more [0] [1]. > 0. https://plus.google.com/111049168280159033135/posts/4nkSEmGoVF4 They cheated and didn't provide a battery to back up the CMOS!!?! > I read "battery-backed CMOS" as meaning the CMOS settings were > backed by the main battery. But perhaps I'm wrong. No. You have it right. Olof Johansson wrote: > It's caused by the Pixel not having a coin cell battery, so the > nonvolatile storage that's normally backed up by it is instead backed > up by the main system battery. For a system that has been sitting and > completely drained, the contents is lost, and with that the additional > developer mode options. But for many years battery backed up CMOS has used a CR2032 or similar on the motherboard to keep CMOS nonvolatile and to run the hardware clock. More recent systems use flash which is slow and can wear out but is non-volatile without power. According to this the Chromebook Pixel has neither. So if you truly lose main battery power then you really lose everything. > Either way though, it's clearly something that has burned others > (and you're right, I can't imagine a non-defective battery lasting > less than a year), so it's worth being very cautious about. Not present. Wow. I can't believe they designed it without any completely non-volatile storage of any kind. Although the mention the TPM and it being slow to access. I would take slow over a full system wipe any day. Would make me want to open the box up and find some place to tuck a CR2032 into it somewhere and wire the power up to whatever it uses for CMOS ram. > Probably it's technically possible to flash different firmware onto > the thing, and Google have been fantastic about releasing firmware > code [2], so it may be fixable. But it's beyond my resources to fix > it, sadly. I read this: Bill Richardson wrote: > We may be able to fix that in future Chromebooks, but changing the > verified boot security features generally requires a change to the > read-only BIOS, which isn't possible with an update. And so I think the machine can't be updated. In the old days we would pop out the ROMs and replace them. I doubt that is possible here but that is the type of thing that seems to be needed. The machine has an SD card slot. The main storage is only 32G. I think I would buy a 32G SD card and install it in the slot full time. Use the internal storage just as a shim to boot over to a 32G SD card. Then you would be safe. Bob
Attachment:
signature.asc
Description: Digital signature