paulmd

78 ms·

781.
by paulmd 3y ago
what about the capitalist revolutions where they started murderous authoritarian “helicopter rides” and burning people alive inside oil barrels and stuff, or generally just starving or mistreating the populace? funny how that always gets le
782.
by paulmd 3y ago
the fact that it probably will not is actually the problem
783.
by paulmd 3y ago
there is an entire hobby for adults putting together random dumpstered parts/whatever they could scrounge up cheap too, google 'homelab' and thank me later ;) (I get your point though, I agree, it's very important for ki
784.
by paulmd 3y ago
oh no, I just bought a refurbed A770 16GB for tinkering with GPGPU lol. It was $220, return?
785.
by paulmd 3y ago
thinking this through a little bit, you are launching a series of gradient-descent work tasks, right? taskId is your counter value, weightIdx is your key value (RNG stream). That's how I'd port that. Ideally you want to define
786.
by paulmd 3y ago
for GPGPU, the better approach is CBRNG like random123. https://github.com/DEShawResearch/random123 if you accept the principles of encryption, then the bits of the output of crypt(key, message) should be totally uncor
787.
by paulmd 3y ago
absolutely wasn't saying sony was good, but, in a practical sense they are going to respond in certain ways to protect their ecosystem/walled-garden like any other. companies are absolutely going to respond if someone puts out an
788.
by paulmd 3y ago
Discord is actually an interesting analogy because they share apple’s overall philosophy here. Discord doesn’t absolutely firewall themselves with attestation/etc but they will absolutely ban you for using alternative clients like Disc
789.
by paulmd 3y ago
this isn't just limited to epyc either, it's in the consumer platform as well, and it applies to non-pro cpus that have been used in branded systems. https://www.youtube.com/watch?v=bFNJVaO9E-o&t=215 not only
790.
by paulmd 3y ago
anecdotally, probably at least 25% of enthusiasts have had a cpu that "randomly died" or "started getting unstable over time" even though they supposedly never overclocked it, and the leading cause there is memory contro
791.
by paulmd 3y ago
Frankly this is aimed more at XMP, where vendors routinely bump IMC voltage to get those swanky binned gaming kits onto the QVL. A ddr4-4800 kit obviously isn’t going to work on stock voltage with a memory controller rated at 3200, the gap
792.
by paulmd 3y ago
however shocking you may find it, typing “epyc cpu” into eBay will nonetheless produce quite a few results. this used to be even bigger in the Xeon days. Being able to drop in a $50 upgrade to x99 and get the 2-3 best chip in socket owns ac
793.
by paulmd 3y ago
B2B customers often value consistency and quality over price. Internal cost of switching is not zero, we have working systems on the old thing and if it’s making us super happy why would we spend money to switch to an unknown? In contrast
794.
by paulmd 3y ago
https://en.m.wikipedia.org/wiki/Mail_cover > Since 2001, the Postal Service has been effectively conducting mail covers on all American postal mail as part of the Mail Isolation Control and Tracking program.[1] htt
795.
by paulmd 3y ago
At this point the secret sauce is that the EAA isn’t tolerating the same degree of certification fucking and laxity from airbus, and that they generally seem to have their act together. Like what’s the secret sauce of nvidia vs radeon or AM
796.
by paulmd 3y ago
of course they can't be, because google's RCS encryption extensions are proprietary and they've refused interop. https://arstechnica.com/gadgets/2021/06/google-enables-end-t... people forget th
797.
by paulmd 3y ago
have them send an icloud link
798.
by paulmd 3y ago
eponymous laws often tend to function as thought-terminating cliches especially if taken to extremes. hanlon's razor is another one: malice is a very real thing in the world, it's not at all the same thing as occam's "do
799.
by paulmd 3y ago
> How did you manage to take this as a personal insult? years and years of "apple sheeple" variants tend to take their toll, you're just the latest in an endless parade of microaggressions even if you don't think your
800.
by paulmd 3y ago
this sounds like proof of stake to me yes, you can indeed build a secure system on the basis of increasing the economic cost of attack beyond reasonable levels and by forcing attackers to repeatedly slash their stake to perform an attack
801.
by paulmd 3y ago
you’re talking to a forum that is probably 50% iPhone and has very good technical reasons to do so, this is insulting and it’s absurd that it’s so casually normalized to directly insult people in this fashion
802.
by paulmd 3y ago
isn’t googles RCS encryption a proprietary non-standard that other companies have requested to interop against and been ignored? yeah can’t imagine why apple doesn’t use it
803.
by paulmd 3y ago
apple
804.
by paulmd 3y ago
> People might want to think about how AirTags and Find My Phone work... rotating BTLE identifiers controlled by a pseudorandom sequence derived from a key, and tunneled over end to end encryption?
805.
by paulmd 3y ago
Nah, assholes ruining access to the beach is a very real phenomenon
806.
by paulmd 3y ago
It’s also at severe risk of ruining the fun for numerous other hacker-spirit communities like hackintosh or opencore. Apple can come down on this in ways that potentially make it much more difficult for hackintosh to operate, or for peopl
807.
by paulmd 3y ago
This thread is a magnet for people who not only didn’t read the article but didn’t even read the headline, or are just continuing to generally repeat the same arguments despite the article contradicting them, without actually making an argu
808.
by paulmd 3y ago
That’s what happens when you gamify “moderation” (pushing buttons at people) instead of providing answers. You get what you incentivize. Honestly, another major deficiency of their approach is the lack of any meta-moderation. If you gamify
809.
by paulmd 3y ago
AMD’s opencl runtime also historically has an incredible number of bugs and paper features that make any sort of portability between nvidia and AMD cards quite difficult - you are running a special build and a different codepath for AMD any
810.
by paulmd 3y ago
In general the problem is not that the documents are not readable or comprehensible - I understand perfectly well that in legalese it says that the situation will favor the business in every possible legal fashion and if some of those are n

More ›