mjb

24 ms·

211.
by mjb 4y ago
Introductory material is always welcome, but I suspect this isn't going to hit the target for most people. For example: > Therefore, if the price isn’t an issue, SSDs are a better option — especially since modern SSDs are just about
212.
by mjb 4y ago
It's fascinating to think about what the internet may have looked like if the early internet adopters had thought about how to welcome and help people, rather than engaging in salty gatekeeping. Specifically, thought about how to scale
213.
Formal methods only solve half my problems(brooker.co.za)
69 points by mjb 4y ago|29 comments
214.
by mjb 4y ago
Emily Oster has a useful post about why, despite the effort, its is difficult to control for everything that matters in observational studies: https://emilyoster.substack.com/p/coffee-definitely-either-k... Specificall
215.
by mjb 4y ago
Yes, it's an active area of research. One very recent example, focusing on Dafny: https://www.amazon.science/publications/testing-dafny-experi...
216.
by mjb 4y ago
> I think the biggest takeaway here is system-wide operations are bad for system-wide reliability. Yes! System-wide operations, whether they are human-driven operations ("ssh onto that box"), control-plane operations ("re
217.
by mjb 4y ago
You can't compare TW and TWh in that way (one is power, the other energy). With solar the way that TW translates to TWh is a little complicated, but it sure isn't 1:1 (probably closer to 1:2000).
218.
by mjb 4y ago
It is tough. My approach when learning new protocols like Raft or Paxos is to implement them in Pluscal (TLA+'s higher-level language) or P ( https://github.com/p-org/P ). I've found that helps separate the pro
219.
by mjb 4y ago
And some folks skiing down Lhotse (the mountain connected to Everest via the famous South Col): https://youtu.be/wPXSFVruIHI
220.
by mjb 4y ago
> It's build-a-bear for VMMs :) Yeah! Nearly everybody says they want a simple and minimal solution in this space, but everybody means different things when they say that. For example, in Lambda we do everything at the block level,
221.
by mjb 4y ago
(I'm the guy in the video on the linked page, but this is just my opinion, nothing official) > Looking at the dismal state of sandboxing APIs in the major operating systems, and the advent of KVM on Linux, WHP on Windows (including
222.
by mjb 4y ago
I enjoyed reading that a lot. > The main advantage of distributed transactions is that they make distributed systems look less distributed by choosing CP, but that inherently trades off availability! This is true, but I suspect that its
223.
Distilling the Real Cost of Production Garbage Collectors [pdf](users.cecs.anu.edu.au)
119 points by mjb 4y ago|156 comments
224.
by mjb 4y ago
That's a good question! Accelerating cold starts with checkpoint and restore is a good idea. There's been a lot of research in academia around it, and some progress in industry too. It's one of those things, though, that work
225.
by mjb 4y ago
Yeah, that's right. I talk about some of the details here: https://www.youtube.com/watch?v=A-7j0QlGwFk We're working on some more formal documentation of this, both to help customers understand how to optimize thi
226.
by mjb 4y ago
It's not entirely accurate that Lambda pulls container images from ECR at start-up time. Here's me talking about what happens behind the scenes (which, in the real world, often makes things orders of magnitude faster than a full c
227.
by mjb 4y ago
In a way, sure. But Lambda also does things that CGI didn't do: dynamic auto-scaling, integration with queues and streams for async processing, strong security isolation with fine-grained authorization, host failure detection and recov
228.
by mjb 4y ago
I do think people are a little silly about knife steel, but in this case I'd love to see the company's test results backing up their strong statement. At 60HRC, S30V should have (according to the data sheet) similar toughness and
229.
by mjb 4y ago
This is a really interesting example of half of a dynamic that constantly plays out in industry. One half is that first-order reasoning seldom leads to the "right" answer, and that optimization opportunities don't tend to be
230.
Scaling Containers on AWS in 2022(vladionescu.me)
15 points by mjb 5y ago|0 comments
231.
by mjb 5y ago
This Old Tony, Blondihacks, Wintergatan, and others have done end-to-end videos from CAD to machining to assembly. I particularly recommend Blondihacks' beginner videos: well produced, fun to watch, and cover nearly everything you need
232.
by mjb 5y ago
That's very cool, and a beautiful piece in the end. Good choice to do it in 303. 316 really is a pain. The ingenuity required to get things like this one on manual tools (especially pre-DRO, but even now) is amazing. Starting out machi
233.
by mjb 5y ago
It's a great book. Short and to-the-point. Highly recommended.
234.
Will circuit breakers solve my problems?(brooker.co.za)
2 points by mjb 5y ago|0 comments
235.
by mjb 5y ago
I absolutely love the The Datasaurus Dozen. Like Anscombe's Quartet, it reminds us that summary statistics are all misleading in one way or another, and there's no substitute for looking at our data from multiple angles, and often
236.
by mjb 5y ago
> Narrowing the interface of a platform will not necessarily improve security, unless it reduces the footprint of host code likely to contain vulnerabilities, so we have focused on understanding the kernel footprint underneath each virtu
237.
by mjb 5y ago
It seems like a good first goal would be to dramatically reduce the size of the code base you need to trust (especially complex code with high privilege), and focus on verifying the correctness of that code base. Then, have fine-grained enc
238.
by mjb 5y ago
This is a great write-up, and also a great illustration of something that's fundamentally broken in the way Linux and Unix think about security. `suid` binaries exist for a good reason: they allow users to escalate their privilege in l
239.
by mjb 5y ago
LWN and Usenix are the two subscriptions I happily pay for, and would be happy to pay more for. The quality of the content, and flexibility around how to share it, is well worth supporting. I also pay the ACM, and that one physically hurts
240.
by mjb 5y ago
We don't have a paper on DynamoDB's internals (yet?), but here's a talk you might find interesting from one of the folks who built and ran DDB for a long time: https://www.youtube.com/watch?v=yvBR71D0nAQ And

More ›