3 ms·

i've been saying for awhile now (to anyone bored enough to listen) that qubes style isolation is the only safe idea on offer these days. i just wish qubes were

i've been saying for awhile now (to anyone bored enough to listen) that qubes style isolation is the only safe idea on offer these days. i just wish qubes were usable. i'm fairly sure that 10 years from now people are going to be citing joanna rutkowska articles with regularity. her flatfooted denial that security is possible without isolation is the only sane perspective. the point is that starting from full isolation and straining to get things to talk to each other in the hope of achieving productivity makes sense. starting from full interactivity and straining to isolate but just when it's necessary "for security" is a battle we will all eventually lose, and we should know it already.

anaphor 12y ago
Doesn't djb espouse a similar viewpoint wrt sandboxing? Not to say she doesn't have any original ideas for how to achieve it, but it's hardly a novel concept.
- simondedalus 12y ago
  yes. i'm not saying the concept is novel, i'm saying that it's the only viable option for the systems of today. related to that, i think the qubes approach will be adopted eventually (and that articles of hers, specifically, that i've read will likely come into fashion, despite the--dare i say it--borderline sexist reaction many on hn seem to have to "her tone"). regarding outright original thought, i'm sure both she and he are guilty of it. probably you and i as well; it's easier than a lot of people think. the world is large and complex, and really most of the "oh my god i can't believe [person] thought of [amazing idea]!" really ought to be marveling at the fact that [person] implemented [amazing idea] as a practical tool. sure, we rehash over and over in our thought, but the world of ideas is so specialized that given even a few months of serious study, novelty is not particularly impressive.
  - jnbiche 12y ago
    Oh come on, sexist? Did you read her exchanges in the comments? She's just another Torvalds, with respective to her attitude and people skills. Why shouldn't we call her on it, if we criticize Linus? She may be a brilliant researcher, but being aggressive and insulting toward others will only hinder her attempts to redefine the security landscape.
    - simondedalus 12y ago
      if someone said to linus "given that kernels are your research project, i expected that you'd already tested whatever you're talking about on wayland and contacted its developers about security holes," and he replied curtly, i don't think anyone on hacker news would be "calling him out." the idea that her tone is "too aggressive" is highly suspect. it is not unusual to express ideas as if you believe them, nor to conduct yourself as if your blog posts needn't be written in the tone of My-First-Job-Interview or Visiting-My-Conservative-Grandmother-In-The-Hospital.
      - rjknight 12y ago
        I think the kind of people who are bothered about Linus' tone are likely to be bothered about similar things in other people. For my part I've never found Linus particularly offensive and I'm equally puzzled by people who disagree with Joanna's style too. Perhaps some people are just too easily offended.
      - vacri 12y ago
        You would be wrong. Torvalds gets some regular hate-machine action here every couple of months. Someone finds a frustrated mailing list message of his and posts it out of context, and half of the resulting comments are about how Torvalds really shouldn't be a project lead with an attitude like that, or how he's damaging open source, or all sorts of pronouncements on his behaviour.

anaphor 12y ago

Doesn't djb espouse a similar viewpoint wrt sandboxing? Not to say she doesn't have any original ideas for how to achieve it, but it's hardly a novel concept.

simondedalus 12y ago
yes. i'm not saying the concept is novel, i'm saying that it's the only viable option for the systems of today. related to that, i think the qubes approach will be adopted eventually (and that articles of hers, specifically, that i've read will likely come into fashion, despite the--dare i say it--borderline sexist reaction many on hn seem to have to "her tone"). regarding outright original thought, i'm sure both she and he are guilty of it. probably you and i as well; it's easier than a lot of people think. the world is large and complex, and really most of the "oh my god i can't believe [person] thought of [amazing idea]!" really ought to be marveling at the fact that [person] implemented [amazing idea] as a practical tool. sure, we rehash over and over in our thought, but the world of ideas is so specialized that given even a few months of serious study, novelty is not particularly impressive.
- jnbiche 12y ago
  Oh come on, sexist? Did you read her exchanges in the comments? She's just another Torvalds, with respective to her attitude and people skills. Why shouldn't we call her on it, if we criticize Linus? She may be a brilliant researcher, but being aggressive and insulting toward others will only hinder her attempts to redefine the security landscape.
  - simondedalus 12y ago
    if someone said to linus "given that kernels are your research project, i expected that you'd already tested whatever you're talking about on wayland and contacted its developers about security holes," and he replied curtly, i don't think anyone on hacker news would be "calling him out." the idea that her tone is "too aggressive" is highly suspect. it is not unusual to express ideas as if you believe them, nor to conduct yourself as if your blog posts needn't be written in the tone of My-First-Job-Interview or Visiting-My-Conservative-Grandmother-In-The-Hospital.
    - rjknight 12y ago
      I think the kind of people who are bothered about Linus' tone are likely to be bothered about similar things in other people. For my part I've never found Linus particularly offensive and I'm equally puzzled by people who disagree with Joanna's style too. Perhaps some people are just too easily offended.
    - vacri 12y ago
      You would be wrong. Torvalds gets some regular hate-machine action here every couple of months. Someone finds a frustrated mailing list message of his and posts it out of context, and half of the resulting comments are about how Torvalds really shouldn't be a project lead with an attitude like that, or how he's damaging open source, or all sorts of pronouncements on his behaviour.