On Tue, Feb 21, 2017 at 3:34 AM, Zrubi <m...@zrubi.hu> wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA256 > > On 02/21/2017 02:59 AM, Oleg Artemiev wrote: > >> Attached file has more details. As usually I've replaced some >> potentially unique numbers w/ XXXXXX. > FYI: > > > The things you are hiding are NOT unique to your device. Instead they > are the exact device TYPE identifiers. Without those you can't even > tell what device you are using. Means without those ids the HCL is > pretty useless. > Details: > > BIOS: what you are masked is the BIOS version. > Different BIOS versions may affect some very important things like > vt-d, vt-x, TMP in my case chipset doesn't support it, not bios, thus no need to publish this. (this information is available via link on dicscussion that realtes to my old report, nothing has changed).
> VGA: you masked the PCI ID. It is the only real identification of the > device: > http://pci-ids.ucw.cz/ > NET: same as above. Without the proper masked ID you are not even able > to cheese the right driver. > SCSI: same again the masked id is the type number of your disk. Any of these as a single entry are not identifying. Combination of those could be enough to split target set from thousands to single hundred . That is not exact identify for law enforcement, but enough identification for targeted attack in "spotter" terms . At least combining that data razes ease of targeting a single computer from distribution center. That is enough reason for me not to post this information even when I trust Qubes team. > However this is actually not relevant in Qubes. Why then this is sent to HCL? Shouldn't we avoid unnecessary information like HDD details at all (or just ask for amount of free space (approximate, not exact, as exact value on system partition may help identify install) on assigned partitions)? > Of course you can still decide to not share those - it is up to you. Yes, I still do - see 'my reasons' below. > I just wanted to make it clear to avoid confusion. Thank you.I'll look for details on identification of devices w/ PCI ID. As about my reasons: There's difference between identification of a person from: *) vendor site (any vendor that has made some software ether running on my laptop) *) from some government agency like fbi/nsa/fsb/whatever *) from abilities that give just internet search for any curious hacker. It's not a big deal to publish that info, since I trust Qubes team and think I'm not that interesting person to spend time catching my personal data - that is probably useless for most of other people . Though I do not trust Microsoft, but I use their OS for gaming. I do not trust Google but I use a few android devices. I do not trust lots of other non-evil-by-intent organizations. Why? Lets imagine, that once, I become a person to catch by whatever reason. Then what? Then any information that I 've sent to the net is ready to use, stored for the hunter pleasure and I cannot erase anything that once has been sent to the net . Period. I see no reason to allow technical reports w/ my personal hardware details to any vendor (including Qubes) - all vendors usually operate on more info than them really need. HDD information sent via HCL seem to be useless to decide "is this laptop good enough to be able to Qubes?" Why not to remove it? We have a proverbial in Russia - "word is not a bird - once flew - no chance to catch". Information you reveal into public should be organized correctly independently to their current importance level , My idea is that if Qubes team wants to get additional information from users about spare parts - the HCL should get divided by at least two parts: 1) laptop model compatibility list (w/ less information about details (I guess within one model the hardware set is similar)). 2) hardware compatibility list w/ spare part list. I.e. if we want to know about some laptop model - one case, if we want to get a list of compatible boards, network adapters, etc. - another case. And very important - the second case is subject for anonymization - it should be hard to make a direct link between an exact spare part and a user reported it. Better even have in HCL a FIXME "Yes I'm okay for public link between my person and data from this report". If this is not what a user want - publish report anonymously. BTW, I'm okay for current link from HCL page to a thread - anyone can get my email and find out what a crappy hardware I use. ;) Currently reporter is directly pointed via link to google group discussion (if he/she agreed on that). I'm okay to send more details on my hardware once I'm sure: *) details are not easily traceable to my one laptop from *) reporter is not easily traceable from reported entry by vendor -- Bye.Olli. gpg --search-keys grey_olli , use key w/ fingerprint below: Key fingerprint = 9901 6808 768C 8B89 544C 9BE0 49F9 5A46 2B98 147E Blog keys (the blog is mostly in Russian): http://grey-olli.livejournal.com/tag/ -- You received this message because you are subscribed to the Google Groups "qubes-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to qubes-users+unsubscr...@googlegroups.com. To post to this group, send email to qubes-users@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/qubes-users/CABunX6OS%2BOYoUevYQo%2BKDi2YHF53Nvz2cgyYOmHUx44eTMpJrw%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
