eslaught

27 ms·

271.
by eslaught 4y ago
Computer Science Research Dept., SLAC National Accelerator Laboratory | Research Scientist / Engineer | Full Time We're a research group within SLAC, headed by Alex Aiken ( https://theory.stanford.edu/~aiken/ )
272.
by eslaught 4y ago
SELinux?
273.
by eslaught 4y ago
But that was the entire point of SIP, wasn't it? To mitigate the impact of a compromised root account. If it doesn't work, then what's the point of going to all this effort in the first place?
274.
by eslaught 4y ago
This phrase bothers me. It's sort of a contradiction in terms: what does it mean to hold a "strong" opinion "loosely"? In my experience, those tend to be highly correlated. The people who have strong opinions hold t
275.
by eslaught 4y ago
I went to UCSD (a public university) for undergrad and then Stanford for grad school (PhD). Both programs were CS. In my opinion, the average CS undergrad at UCSD is honestly not that different from Stanford. At Stanford, there are a few tr
276.
by eslaught 4y ago
The thing that's crazy is that even the "regular" items sometimes have artificial sweeteners now. You basically cannot assume any product category, with any labeling, with not have it. You have to check the ingredient list yo
277.
by eslaught 4y ago
It's true (in some memory spaces) in AMD GPU too: https://llvm.org/docs/AMDGPUUsage.html#memory-spaces
278.
by eslaught 4y ago
Is it correct to say that when you get something "shipped and sold by Amazon.com", it really comes from them? It's the "fulfilled by Amazon" (the infamous FBA) that you need to watch out for. It's a bit screwy,
279.
by eslaught 4y ago
1. No one updates supercomputers automatically. Instead you firewall the compute nodes and have a team dedicated to maintenance and support to do periodic upgrades. 2. I'm sure they had a support contract, which may have included runni
280.
by eslaught 4y ago
Thanks, I appreciate this. I don't think this practice is widespread, which is exactly why I made a point about human engineering in my original post. But I do appreciate that solutions like this exist, and I should look into driving m
281.
by eslaught 4y ago
Yes, exactly. And how many of these projects go to any effort to pin their transitive dependencies? In my experience, this is basically no one; in fact I've seen many (most?) projects not even pin the versions on their top-level depend
282.
by eslaught 4y ago
To be honest, I don't see how you can give this advice. A couple of the most popular package management systems for Python are very obviously deficient (and sigh, these are usually the ones I get stuck working with, due to outside cons
283.
by eslaught 4y ago
Disclosures: I've written a book I'm currently trying to get published. Also, almost all of my professionally-written (as well as personal) code is open source. And I have used the Wayback Machine extensively. To be honest, I&#x27
284.
by eslaught 4y ago
"No trust required" I think that's what's intended by "zero trust". It's not that trust is bad, it's that it's not necessary when you architect the system in a certain way.
285.
by eslaught 4y ago
Those attacks don't scale. They're also tied to a person or people who need to be physically present, making it possible to arrest them. The consequences of potential attacks on centrally-orchestrated traffic are a lot more severe
286.
by eslaught 4y ago
I haven't read the book, but... Aren't the digits of pi randomly distributed? So then, you can encode literally any message you want by picking an offset into them. Could even design a "compression" algorithm where a mes
287.
by eslaught 4y ago
HIP is basically CUDA with s/cuda/hip/g. My experience is that the stack is pretty rough around the edges. But when it works, you (almost) literally find-and-replace, and it pretty much works as advertised. However, just beca
288.
by eslaught 4y ago
There's a danger, while steelmanning, that you might come up with an argument that sounds good to you , but it's actually weaker to the opposition. Therefore, while trying to steelman their argument, you actually fill it with hol
289.
by eslaught 4y ago
I should have checked this before posting. I'm actually on Gboard, so all of my comments apply to Gboard. What customizations do you use?
290.
Ask HN: Why is mobile keyboard completion so bad?
5 points by eslaught 4y ago|9 comments
291.
by eslaught 4y ago
Ye olde Lisp dump image. A neat approach, but results in an absolutely horrific cross compilation experience. As someone who worked briefly on a Common Lisp implementation, including trying to get it to cross-compile, I'm very glad thi
292.
by eslaught 4y ago
How is this different from the old privacy.firstparty.isolate, and do I still need that/should I keep that enabled?
293.
by eslaught 4y ago
Just to follow up on my own question (though I appreciate the siblings): I found this article with some overview and example command lines: https://www.docker.com/blog/multi-arch-images/ . As best I can tell, you
294.
by eslaught 4y ago
Are there more details on "docker buildx build" that you can point us to? The command line reference doesn't seem especially helpful: https://docs.docker.com/engine/reference/commandline/buildx_
295.
by eslaught 4y ago
Terra is a bit unusual in that it consumes LLVM (another large project), and in ways that LLVM doesn't seem to have been quite intended for. So the Makefile build was always a bit of a chore because we had to reverse engineer a bunch o
296.
Terra 1.0.0(elliottslaughter.com)
4 points by eslaught 4y ago|2 comments
297.
by eslaught 4y ago
This does sort of beg the question though: how do you draw things that don't exist? Everything (or, almost everything) I personally care about drawing is sci-fi or fantasy, and sort of the whole point of it is that it doesn't actu
298.
by eslaught 4y ago
They could have posted from a non-throwaway (or at least non-anonymous) account and identified a clear point of contact. As it is with such vague instructions, it would be far too easy for the issue to get lost in the Rube Goldberg machine
299.
Terra 1.0.0-Rc1(github.com)
1 points by eslaught 4y ago|0 comments
300.
by eslaught 4y ago
I'm surprised you're not using HIP? At least in my experience it seems like HIP is the go-to system for programming the AMD GPUs, in large part because of CUDA compatibility. You can mostly get things to work with a one-line heade

More ›