5 ms·

Seems like the thousands of people who bought it care, myself included. My role has significant on-call responsibility. It's nice to be able to take it charged

Seems like the thousands of people who bought it care, myself included. My role has significant on-call responsibility. It's nice to be able to take it charged anywhere I could need it and know that for several hours it can handle anything I will throw at it, reliably. From a security perspective, I prefer the separation from the x86 architecture if for no other reason than it imposes extra cost for exploit development.

pjmlp 2y ago
It is only a matter of time. https://www.darkreading.com/application-security/patchless-apple-m-chip-vulnerability-cryptography-bypass https://www.darkreading.com/application-security/patchless-a...
- wyldberry 2y ago
  That wasn't the point. On a long enough timeline, the security of all systems drops to zero. If the entire world is running x86, then the threat model for attackers revolves around abusing x86. If things are homogeneous, it raises the bar and resourcing required to make attacks. As always with security, everything is a tradeoff.
  - namaria 2y ago
    Your argument is basically security by obscurity. You're better off in an ecosystem where a lot of attention is paid to exploits and patches then in another where it might be a long time before a zero day becomes known and fixes are issued.
    - wyldberry 2y ago
      None of what i said is security by obscurity (which is also can be an effective tactic,but obviously not the only tactic). There are only so many human hours and minds interested or allocated to exploitation and offensive security. If everyone used the same architecture for everything, the economies of scale on the offensive side (due to state funded actors) would blow everyone else out of the water. From a software perspective, Windows has an incredible amount of skilled eyes on each patch release, but we still see new exploits. Same for Linux. Likely same for MacOS. All i'm advocating for is that having separate hardware architectures is good because it raises the barrier to entry, even if it's only the next marginal step.
      - dannyw 2y ago
        Security by obscurity isn’t even bad. It’s only bad if it’s your sole defence. I am confident that my non-default SSH ports, that only accept connections after a sequence of port knocking, adds a slight bit of security to nothing. For example: xz backdoor.
justsomehnguy 2y ago
> My role has significant on-call responsibility. It's nice to be able to take it charged anywhere I could need it and know that for several hours it can handle anything I will throw at it, reliably. You could do that for years with an extended battery in ThinkPads (and Dell I think, never used them). More so, you could just bring additional charged batteries with you, in case you really needed. But yes, it took a little more space and weight than M1. So M1 didn't allowed people on-call to be able "to throw anything at it for several hours", the tech was already there. The only thing M1 did bring is a small savings in the weight and space for the people who needs a glorified terminal with them. Hint: ultrabooks were a thing long before M1.
- kaba0 2y ago
  The M-serious brought an absolutely insanely big jump on the performance-energy efficiency graph, and that’s without doubt. Nothing came even remotely close to that before. Those ultrabooks were insanely expensive and had like 3 hours of battery life.
  - justsomehnguy 2y ago
    If you are on-call then you diagnose and repair some remote system. You don't need the performance for it. > Those ultrabooks were insanely expensive and had like 3 hours of battery life. Ah, yes M1 for $300 is surely better! And no, they didn't had 3 hours of the battery life. I had 4 hours+ just a couple of days ago on a totally not tuned ASUS TM420 with a 3-cell battery. > Nothing came even remotely close to that before Except you needed to adapt your whole workflow for it, but as usual this part is ignored. Sure, today you can do that way more easily because most (read: webdev, lol) things are done, but...
    - kaba0 2y ago
      Ok, 4 hours vs 10. It is still the difference between not powering off between bringing it from one power source to another and an actually portable use-case.

pjmlp 2y ago

It is only a matter of time. https://www.darkreading.com/application-security/patchless-apple-m-chip-vulnerability-cryptography-bypass https://www.darkreading.com/application-security/patchless-a...

wyldberry 2y ago
That wasn't the point. On a long enough timeline, the security of all systems drops to zero. If the entire world is running x86, then the threat model for attackers revolves around abusing x86. If things are homogeneous, it raises the bar and resourcing required to make attacks. As always with security, everything is a tradeoff.
- namaria 2y ago
  Your argument is basically security by obscurity. You're better off in an ecosystem where a lot of attention is paid to exploits and patches then in another where it might be a long time before a zero day becomes known and fixes are issued.
  - wyldberry 2y ago
    None of what i said is security by obscurity (which is also can be an effective tactic,but obviously not the only tactic). There are only so many human hours and minds interested or allocated to exploitation and offensive security. If everyone used the same architecture for everything, the economies of scale on the offensive side (due to state funded actors) would blow everyone else out of the water. From a software perspective, Windows has an incredible amount of skilled eyes on each patch release, but we still see new exploits. Same for Linux. Likely same for MacOS. All i'm advocating for is that having separate hardware architectures is good because it raises the barrier to entry, even if it's only the next marginal step.
    - dannyw 2y ago
      Security by obscurity isn’t even bad. It’s only bad if it’s your sole defence. I am confident that my non-default SSH ports, that only accept connections after a sequence of port knocking, adds a slight bit of security to nothing. For example: xz backdoor.

justsomehnguy 2y ago

> My role has significant on-call responsibility. It's nice to be able to take it charged anywhere I could need it and know that for several hours it can handle anything I will throw at it, reliably. You could do that for years with an extended battery in ThinkPads (and Dell I think, never used them). More so, you could just bring additional charged batteries with you, in case you really needed. But yes, it took a little more space and weight than M1. So M1 didn't allowed people on-call to be able "to throw anything at it for several hours", the tech was already there. The only thing M1 did bring is a small savings in the weight and space for the people who needs a glorified terminal with them. Hint: ultrabooks were a thing long before M1.

kaba0 2y ago
The M-serious brought an absolutely insanely big jump on the performance-energy efficiency graph, and that’s without doubt. Nothing came even remotely close to that before. Those ultrabooks were insanely expensive and had like 3 hours of battery life.
- justsomehnguy 2y ago
  If you are on-call then you diagnose and repair some remote system. You don't need the performance for it. > Those ultrabooks were insanely expensive and had like 3 hours of battery life. Ah, yes M1 for $300 is surely better! And no, they didn't had 3 hours of the battery life. I had 4 hours+ just a couple of days ago on a totally not tuned ASUS TM420 with a 3-cell battery. > Nothing came even remotely close to that before Except you needed to adapt your whole workflow for it, but as usual this part is ignored. Sure, today you can do that way more easily because most (read: webdev, lol) things are done, but...
  - kaba0 2y ago
    Ok, 4 hours vs 10. It is still the difference between not powering off between bringing it from one power source to another and an actually portable use-case.