• 0 Posts
  • 12 Comments
Joined 3 years ago
cake
Cake day: June 7th, 2023

help-circle
  • the thing is in the video tutorial he’s using password only(no tpm) and it still requires you to create a recovery key

    Never watched the video, so I’m just answering for the general case. Even with just a password, the recovery key serves a similar function. It’s a fail-safe if the primary protector is lost. It’s entirely possible that the password being setup will be given to a user and IT will still need a way in, if that password gets lost/changed.

    on Linux with luks, it never asked for me to save a recovery key.

    Ok, and? Yup, LUKS uses different defaults. Don’t get too caught up on “Linux does it this way”. Both ways have valid reasons and the Microsoft developers and LUKS developers picked different defaults. If you want, you can create a recovery.key file on LUKS. If you are using TPM as your primary protector, I’d actually recommend you follow Microsoft’s lead on this. If that TPM chip (or the motherboard it’s attached to) dies, you’re kinda fucked. If you are using a password, this is less of an issue. assuming you have that password stored safely somewhere not on that system.


  • It’s a design decision and one which is probably about right for many situations. TPM chips are not perfect, and there are a lot of situations which can cause them to fail or the hardware around them can fail (e.g. the motherboard). With the end result being that the data would be rendered unreadable. So a recovery key is a trade-off between confidentiality and accessibility.

    The design of the recovery key itself is also fairly good as a backup protector goes. The key is long and complex enough that brute forcing it is hard enough and would require long enough that it is functionally impossible. If your concern is that the recovery key could leak, the easy answer is “don’t write it down”. You could simply choose not to store it and while it would be a theoretical attack on the confidentiality of your data, practically it’s a non-issue. But, for those folks who feel that having an unlock code written down somewhere is an acceptable risk while mitigating the risk of the TPM breaking and rendering the disk unreadable, it’s a good solution. And I can personally attest to the usefulness in a corporate environment. People do dumb things with laptops all the time. And no matter how much we might tell people to use the network/cloud storage which is off the system and backed up regularly, there is always a parade of people whose system just died and contains the only copy of important project data.

    It’s important to remember that security is a long series of compromises between confidentiality, integrity and accessibility. All three are important. And you have to find the spot in the nexus of those factors which works for your situation.



  • In Maine having Platner flame out in such spectacular fashion did not help Democrats at all. That said, with a split of 48% to 45%, that leaves 7% either undecided or listing a different candidate. Couple that with a stated error of +/-3.7%, and the poll is putting Jackson with anywhere between 59% and 44% of the vote, if it were held today. The polling average has been pretty stable. And SSRS’s poll isn’t telling any different of a story. So, maybe this will be the year we are finally rid of Collins’s “oh I’m concerned, but not concerned enough to actually vote against Trump”.

    Michigan is similar. SSRS lists the race as 47% to 44%. This leaves a 9 point gap in Undecided/Other. And a stated error at 4.1%. That’s a lot of room for these numbers to swing. Though the polling average has moved a lot more in the past few months. It probably doesn’t help that the Democratic Primary took a bit to wrap up. But, a slight lead in a heavily Democratic national environment, with enough undecided voters left to shift the race, isn’t the worst place to be.







  • I think they buried the real lede here:

    The same survey also found that, for the first time in nearly a decade of its polling series, voters perceive the Democratic Party as having a better plan for the economy than the Republican Party.

    That we’re in a positive environment for the Democrats is a kinda “no-duh” situation. The out-party is usually favored during mid-term elections. There are also some well known economic issues which are working against the GOP (see: gas prices). So, we’d expect the Democrats to have an advantage. But that shift in trust of which party is better for the economy would mark a major sea change in opinion, if it sticks. As dumb as it may be, the American public has a habit of responding that they trust Republicans more on the economy than they trust the Democrats. If that changes, the Republicans would lose one of their major pillars of support.

    That said, I’m not sure I’d put all that much stock in polling yet. Americans have a habit of not really paying attention to politics until very late. The common wisdom being that most people aren’t paying attention until after Labor Day. Those of us subscribed to a community named “politics” are likely the exception and not the norm. So, polls and averages showing a D+6 environment are nice, but not all that useful yet. And things are likely to shift quite a bit before the actual midterms. Also, three months is a lot of time for Trump to do more stupid stuff or get stupidly lucky. We could be even more bogged down in the Iran quagmire or a deal could be struck which extricates the US, re-opens the Strait and brings down gas prices.

    All that to say that, this is interesting as a quick check on where things seem to be right now, but it’s probably not that predictive of where things will be in three months.