I'm 3d printing a front vent plate for my new machine to keep my toddler out

... again... because my printer is not calibrated properly and it was 3mm too wide :(

She's going to shove random toys in there or worse, grab a screw driver and "fix" the computer like daddy does :P

@StephaneHockenhull that is cute. My nephew comes in my room to steal my stuff or try to "play" my Xbox

@StephaneHockenhull Use ducktape?

(yeah, that, ducktape)

Ohh ducktape... fancy!

:P

@StephaneHockenhull That looks like you put way more work in it then you had to

It looks nice though

Not in gamedev I don't think

But in bigger companies then where I work sure

I have never seen anyone use UML for anything outside of uni.

There are plenty of people who make diagrams, but not in UML, because the relations are not useful and nobody knows what the things mean.

Shouldn't the background of the bounty notice have a different color, than the "closed" notice?

It feels strange how two different things have the same look to them

@Tyyppi_77 whenever i have stuff to explain, I use UML. At work.

But my colleagues are like "wtf".

stackoverflow.blog/2018/01/15/thanks-million-jon-skeet/…

Wrong answer.

lol

at former employer, I made a markov chatbot for a few coworkers who had particular mannerisms in chat

On the off-chance that any of you are from or passing through Switzerland this weekend, come check out The Ludicious Festival and some awesome games from all over the globe

@AlexandreVaillancourt I get the arrows backward and it looks like the base class inherits from the other one

It would be nice if there was a "Bitcoin" for donating your computer for science.

@Bálint It was a bit of work to model but most of the time was waiting for the software to generate the geometry and then waiting for the printer to print (2h07m) each attempts. Took three prints to get the calibration values good enough that it'd fit but the big grille was a bit too small and didn't snap in place. I had to use a bit of hot glue.

As long as it keeps my 2 year old from looking for toys that would neatly fit in there :D

She has lots of "project" ideas as we call them. There's still a chance I'll come back from work and there's going to be one crayon neatly placed in each hole sticking out at exactly the same depth.

The problem with a science based project would be keeping the right flow of coins coming in, while accepting new projects and weighing them by complexity.

Hmmm....

like SETI but more general ?

Yeah.

Also, there would need to be some work done to get a project to work.

And that could be dangerous, to say the least...

simple, write it in Javascript and ask people to open the web page to crunch numbers.

But something to reward them for actually doing it.

I mean, how many millions of dollars of computing power (Probably billions) are going to mining bitcoins?

What if there was a way that something useful could be done with that computing power, and reward the user?

Just write an idle-clicker "game" like all those idle-clickers but that actually produces something..

I've heard of some free Android games that actually mine bitcoins in the background when running.

fossbytes.com/block-cryptocurrency-mining-in-browser

Well, I got a per-core CPU meter in my task bars on all my machine on every screen so it's easy for me to spot when a site has one of those.

@PearsonArtPhoto GRC

Gridcoin Research

Also... FoldCoin I think

Figured there had to be something

yeah :)

I don't know if GRC is still around though

poloniex.com/exchange#btc_grc

yup, still there

huge hassle to get it set up and running though

It's kind of obvious if I suddenly see one core pegged at 100% after opening a browser tab and nothing else is running.

Current estimate is that as of mid 2019, bitcoin mining will take as much power as the entire US

( and nobody posted a GIF on GDSE's chat :D )

something has to be done about it

That's the reason why I suspect that Bitcoin is going to continue to go up in value, at least as long as it can be mined.

It'll peter-out on it's own

Bitcoins right now are tied to an actual resource, electricity.

Like how online poker was the rage a while ago.

Its value is probably going to crash hard soon and then people will be turned off once they see how much they spent on hardware for nothing.

There is a role for digital currency.

But probably not as currency.

Just not at the current level...

And not without government backing, which cryptocurrency tries to avoid, so it can't really work as is.

And it's also a forever-deflating currency which is a big problem.

It could work without government backing.

But Bitcoin isn't the winner.

BTC is a horrid currency

buy ice cream cone. gotta wait 2 hours for the transaction to clear.

There's nothing to anchor its value so it'll oscillate more and more until people abandon it.

provided you pay ton in transaction fee

otherwise it's much longer

Someone need to do a video of this with the ice cream truck guy holding the ice cone melting while waiting for the transaction to clear.

hahaha yeah

And hand out the soggy cone bent in two to the kid 4 hours later.

All that said, I think BTC is a great experiment on many aspect.

I'll agree with that, though I question the guy's motives.

I think it was both a brilliant realization about how to make such a decentralized system work, and the cynical realization that "decentralized" and "tied to proof of work" would appeal to certain people who would raise its value so he could make a killing

We got a lot of technology improvements on distributed computing and I love all the PCIe extender options out there now.

PCIe extenders used to cost $2000-$3000 for a 1-to-4 extender. Now you can get one for about $100.

And I'm stuck with a 1050 until the mining craze ends.

@Almo have you looked at stocks for 1070s and 1080s? Even the 1060s are out of stock.

Oh, newegg got a few 1060s in stock now... yay...

They're now around $500 instead of $300. geez.

Hmm, it's been a couple of days since spectre and meltdown were written about and still no major data thefts

It seems like reading the memory with a 0.2B/sec speed is pretty inconvenient

But the big reason is theres a lot more ways that are much easier to exploit for data theft :P

I've never seen meltdown examples in action, how fast is that?

Spectre requires training the branch prediction which takes a while and then do statistical analysis of the timing over multiple attempts.

Spectre is super slow and can be pretty trivially slowed down to impractical levels. which in theory is still an issue.

@Bálint The Tempest spying hack is almost more worrysome than Spectre

it's possible to spy on hdmi,vga,displayport monitor signals and/or the LCD update pulses using a SDR receivers.

Or you know, just steal the hardware

Yeah, but that's more conspicuous.

People steal more protected stuff every day

wait, which one is the obious meaning...hrm.. my English fail

Researchers were also able to extract AES keys by just analysing the power going into a device.

but that only works if the device is single-core or doing nothing else and there isnt another computer on the same power feed

And you can get close enough to the computer to attach a measuring device to it

yeah. still facinating.

It is

like, the bug could maybe be hidden in a wall socket

It's like the time, when someone was able to figure out your PIN code from the way you tilted your device

I'm not sure what he's doing wrong, it works for me

Maybe his sphere collider is too large or he checked the convex box :P

YES! I got an entire USB root controller pass-through to the Win10 VM

But does it work, if you start another VM inside the VM?

The Win10 VM got its own dedicated 4x USB 3.1 ports now.

VMception!

Nah, I don't think I can double-passthrough :P

What OS does the physical computer use?

So I don't need to hand-assign USB devices as passthrough one by one, I can plug anything into those port and sub-hubs and it's in the VM

Next: the SATA controller.

I'm gonna have to figure out which SATA port corresponds to which one of the 3 SATA controllers.

That didn't go so well.. My entire machine hard-rebooted.

It's either the SATA passthrough that crashed it or the USB passthrough on its 2nd run... removing SATA and trying again...

Ooooooh. Looks like the NVMe drive could multiplexed with the same PCIe lanes as that SATA controller and that might have crashed the whole system when I passed the multiplexer device on to Win10 and it got reset, killing Linux's root drive.

Okay, passing any of the SATA controllers cause the machine to reboot. Oh well.