Have my main server back home, while I’m traveling I have a script to reencode to av1 onto my local machine, works beautifully and the quality drop isn’t too bad (colors look weird but think that’s the Intel xe encoder)

Posted from my iphone…

Used to have miredo, which worked pretty well, but think they killed that.

Should still have old 6to4 protocols, they use ipv4 address tricks to get everything working.

They died when inkjet ink became their core business the rest of the company revolved around.

Also Carly fiorina, she ruined it for women ceos for a while.

Cups was due, too much functionality on too many systems, it needed to be more limited and secure by default.

For larger applications you don’t use agm, you often go back to flooded batteries with even replaceable, high cycle cqthodes: https://www.sciencedirect.com/science/article/abs/pii/B9781782420132000030

This is something I’ve felt we’ve needed for a long time, but cloudscalers have their own environments that include resource management and beyond dev, if anything goes wrong they either reboot the net image or offline it for maintenance.

This is something I’ve wanted to throw together, will give it a try soon, could even be useful for development.

It won’t, the market share is generally complementary, not competitive, the sectors tend to be different (more or less until recently).

Mostly, if people are really scared into might fold (unlikely, but we don’t know everything) then the ftc will roll out the red carpet for a player like Intel.

Doubt it will happen, qcomm is too smart, but it’s not unthinkable, and it would give qcomm domination over US cpus, save hyperscalers.

It only happens if people are truly terrified.

Molten salt sounds like a terrible design for modular, the whole problem is if it loses power it freezes solid, you’d want a huge one with tons of backup imho.

I’d imagine a tiny pebble bed or traveling wave, something fairly inert and safe.

Edit: I guess that’s the point, give someone a reactor, if they screw it up it safely freezes dead. Problem solved.

Basically you save money on tech/support because of scale.

So you triple and quadruple your sales and marketing spend to get more business.

In the end it just doesn’t work, except the smaller guys and a lot of them are just hanging on as the stacks get more complicated.

Aws and gcloud are thickening the stack and driving everyone else out of business.

We really need them for boomers, for the world’s sake.

Nobody wants it, they just want brighter lights for themselves to compensate for being blinded by the brighter lights of others, but actually to retaliate, nobody can have brighter brights than me!

We’d need regulations for this, which we’d never get, especially after the Chevron doctrine was reversed.

Poles: “Don’t invite them over for beer…”

Yeah, so we’re still what, 10k behind China this year?

Hello.

As someone who’s in the space and has been around Qcomm and their deals before.

It won’t happen.

They will flirt like you can’t imagine, they will propose, make offers, etc.

But closing the deal? No.

They are very smart, and Intel is too big for them to dismantle and exploit with value.

Their interest is not in Intel belonging to them, but in a large, Intel shaped hole in the market that they can attack, and their discussions are more likely about Intel’s roadmaps so they can understand how they could best exploit Intel’s fall.

They are unlikely to even hire some of Intel’s spoils, maybe a few strategic VPs, but… they’re just smart and ruthless and Intel is the dregs and bloated nowl.

The only way they’d do it is if the government sweetened it such that Intel was basically free, and they could fire as many as they want in a reasonable period, basically letting them own Intel without any cost at all. That is possible depending on how desperate the government is to prevent their fall, but I don’t think anyone can make the right promises in time.

Went back to BF4, felt like coming home.

We lost so much in the last decade, maybe a few new shooters can help bring things back.

They also have 100x we many executions as we have, probably closer to 1000x.

To absolute morons who make sure it comes with a bed cover they never take off.

2 things:

1. This seems to be a specific attack for their IM protocol if the entry node was compromised, and could be placed nearby the client. To make this much easier, you’d want to compromise both the entry and exit nodes (in this case exit node is TOR native, so it’s more like internal node).

This has never been unknown, this is one of the fundamental attack vectors against TOR, the IM protocol seemed to make correlation easier due to its real time nature.

They added a protection layer called Vanguard, to ensure the internal exit nodes were fixed to reduce the likelihood that you could track a circuit with a small number of compromised internal exit nodes. This seems like it would help due to reducing likelihood of sampling.

2. TOR has always been vulnerable, the issue is the resources needed are large, and specifically, the more competition for compromising nodes the more secure it is. Basically now the NSA is probably able to compromise most connections, and they wouldn’t announce this and risk their intelligence advantage unless there was an extremely valuable reason. They definitely wouldn’t do so because a drug dealer was trying to make a sale. Telling normal law enforcement basically ends their advantage, so they won’t.

Other state actors might try, but they’re not in the same league in terms of resources, IIRC there are a LOT of exit nodes in Virginia.

tl;dr - The protocol is mostly safe, it doesn’t matter if people try to compromise it, the nature of TOR means multiple parties trying to compromise nodes make the network more secure as each faction hides a portion of data from the others, and only by sharing can the network be truly broken. Good luck with that.

Well, it can’t run multithreaded jobs at full speed.

Exhibit A: The latest AMD patch for multicore scheduling across NUMA.