off-topic-tech

[tribal kraken]

Some early engine control units from 80-90s had also all the components glued/covered with some sort of elastic goo on the motherboard. Guess anti vibration. Still there were cold solders to found that were cracked. It was a Kawasaki GPZ750 turbo bike. Local electronics shop was still able to find the bad solders and dig in to them for a fix

[twin dew]

Also water proofing.

[minor pawn]

If i was to connect another ssd with windows on it to a laptop with an ssd that already has windows would it then read the new ssd as a storage device or would it not work?

[twin dew]

Would work in most cases, the one exception that can cause issues is bit-level clone where partitions have exact same "random ID"s.

[minor pawn]

One is windows 10 and the other windows 11

Would it also work with an external enclosure?

[twin dew]

The contents don't matter, the connection doesn't matter.

[minor pawn]

Okay thank you

[tribal kraken]

Simracing latest rigs

[twin dew]

Having exactly the same unique ID can cause issues, but that is only issue for bit-level drive to drive clones.
And if course to actually see the partition contents, it must be a type the OS can read.
So for Windows, just few types without extra software.

[maiden coyote]

https://pcpartpicker.com/list/MKF6GP I'm a little iffy on the cpu cooler.

Case is 88mm of height total.

[edgy hazel]

not a bomb™️

[dire igloo]

AXP120?
Also, overpriced SSD and MB

Could also consider 12400F at $110

[maiden coyote]

I've already got the motherboard cpu, and memory.

[sharp oasis]

Buh

[dire igloo]

aight then, still an overpriced SSD - and consider AXP120 (check if it fits the RAM)

as much as i like NM790, $80+ is too much for 1TB

MP44L at $59 is my recommendation.
for highend, MSI M480 at $68

[maiden coyote]

The axp120 is too tall I think

It'll only have like 10mm to pull air into the fan because the case is 88mm high, minus motherboard standoffs, and cpu height. So in reality I've probably got 70mm of clearance

[visual tree]

I was thinking of getting Cloud Gateway Ultra when I get fiber but looks like Ubiquiti released a new product: https://www.youtube.com/watch?v=FqMwPRiNyok

Although 2,5 Gbit would be overkill in my case, seems like Cloud Gateway Ultra should still be enough for my network

[maiden coyote]

It's a cheaper dream machine lol

In a nutshell.. you can use it with protect. Vs the other one is network only.

[dire igloo]

no mesh on the side?

[visual tree]

I could buy Unifi Intercom and connect it to cloud gateway ultra but my neighbours are jerks and they would smash a new intercom the moment they notice it

I had a very old intercom before and had to buy a new one because someone set it on fire....

So, only vandal-proof intercom can be installed

[maiden coyote]

No, the top is solid. flow through 2u chassis... similar to this one

This one has a vent for the psu, but that's it

[maiden coyote]

Wtf... it sounds like you need cameras lol

[visual tree]

Already have, my neighbors are getting worse lol

I could move out but bad neighbors are everywhere no matter where you go

[maiden coyote]

There are standards. Setting your shit on fire is above and beyond.

[cosmic imp]

Tfw I ordered a 1tb barracuda and Amazon sent me a 4tb ironwolf

How much did that rack set you back ?

[maiden coyote]

6x 12tb ironwolves set me back about 1200.

[winged valley]

That should be fine

[winged valley]

this

Noctua has an ultra sff cooler too

[winged valley]

Damn

Wtf do you need that much storage for tho

My unraid has 20tb usable space

[tough owl]

im so glad I have a UPS

I lost power twice and my cheap powersupply probably would've blown because I don't trust the power of where im living

@prisma cradle what you want?

[prisma cradle]

hi @tough owl

nice to connect with you

[tough owl]

https://tenor.com/view/meme-horrors-beyond-our-comprehension-low-tier-god-mods-banned-gif-8530537273735940092

[prisma cradle]

lol

hey kade, where r u from?

[tough owl]

theres no way this isnt a bot

[prisma cradle]

what?

[maiden coyote]

I'm iffy if that cooler, the one I've picked out will keep up with the 5600gt

[winged valley]

nobody randomly friends a random person theyve never met, only to then ask them "where are you from" in a public discord server

So either you're a bot or developmentally challenged

[prisma cradle]

got it
thanks for your explanation
btw, where r u from?

lol

[winged valley]

Gargle my balls

[maiden coyote]

Would you like to join me on what's app to hear about a remote work opportunity for up to $150 a day?

Maybe an extended car warrenty. lol

[languid gulch]

not enough emojis & shortened links

[twin dew]

*ooh* Asus has enabled Curve Shaper for 7000-series in the latest stable UEFI versions.

No idea if it works properly, but it was accessible in the BIOS.

[sour gate]

Whats the theoretical molecular multi baring crystal lattice structure of my geometrically challenged noodles I had for lunch today

I always get the wrong guy

[twin dew]

Seems that by AMD side official line is that it will only work with 9000-series.
But not for technical reasons, just because AMD isn't bothering to validate and tune it for 7000-series.

The company confirmed that there’s nothing in Ryzen 9000’s hardware that enables Curve Shaper. The feature could work on Ryzen 7000 CPUs, at the very least. Instead, it pointed to firmware development costs.
But seems that Asus firmware people want to enable it for 7000-series too.

Shamino wants to play with 7000-series too, not just 9000-series 😁

[twin dew]

And it does work at least somewhat, someone is able to boot into Windows with -45 instead of -20 with it.
Not stable, but didn't check what exactly his tune is.

[languid gulch]

AMD better grab a DBZ IP license for their next gen CPUs

[twin dew]

?

[languid gulch]

because they'll be over 9000

[twin dew]

Ah, Over 9000 thing?
They are already over 9000 with the actual models.

9950X is way over 9000.

[languid gulch]

yea true

[twin dew]

And Intel was there already in 2018.

[languid gulch]

yea but intel wouldn't spend money on that, i feel like AMD would

[twin dew]

But that makes it worthless at this time.

[willow pike]

intel will soon be over 9000W

[dire igloo]

www.pcworld.com/article/2413156/hackers-can-wirelessly-watch-your-screen-via-hdmi-radiation.html

[twin dew]

Yes?
https://en.wikipedia.org/wiki/Tempest_(codename)

And that has been taken into account for specific use cases in room construction for long time.

So that has been done with various computer to display cable techs for long time, with no AI in the mix...

So even the clickbait article both says it is new, but then says it isn't new...

Just more reliable than the older methods for HDMI that the writers knew about.

[willow pike]

this is the kind of shit that the cia would use, and if you're threat modelling against the cia, it's already over

[twin dew]

And I would like to see someone do the same for multiple displays with cables going next to each other...

[willow pike]

your average rats nest under a pc nerd's desk would give them grey hairs trying to figure out

[twin dew]

Mostly theoretical stuff in real world, but it is taken into account in specific cases by including faraday cage type stuff in the walls etc.

And with specific testing for the HW to see that you don't install malfuctioning or too much radiating devices in such rooms.

And so on.

[willow pike]

Intel Corp. plans to eliminate thousands of jobs to reduce costs and fund an ambitious effort to rebound from an earnings slump and market share losses.

The workforce reduction may be announced as early as this week, according to people familiar with the company’s plans, who asked not to be identified because the information isn’t public.

[twin dew]

So again trying to cut to success?

[willow pike]

can't link but it's bloomberg

[twin dew]

Which is the cause for the whole mess they are currently in 🤣

[dire igloo]

remove https from link

[twin dew]

Probably paywall?

[twin dew]

But anyways, Intel long time ago:
Lets get rid of the most expensive engineers to cut costs! (as in the most experienced ones)
And then in less than 5 years they were in major problems with 10nm development etc.

So of course when you are still having problems stemming from bad layoffs & early retirement programs, of course you need to cut costs by firing people 🤣

Because I really don't expect the cuts to touch the upper management at all, and very little work on the middle management either.

When the most important thing is that the Exec level gets their full performance bonuses.

Still being in operation few years from now is much lower priority to that.

[willow pike]

don't forget when they cut $4 billion from fab investment to fund shareholder dividends!

exactly the same time as they cut bonuses for, uh, fucking everyone

whole company has been fucked since 2015

[twin dew]

That was probably when the effects of those layoffs started to show up in major ways, but the actual layoffs I'm talking about would seem to have happened somewhere around 2010 or so.

Cannot find the actual year-range, but 2010-2012 would seem most likely.

[willow pike]

that tracks

[twin dew]

Ah, might even have been already in 2006.
Otellini took over in 2005, and then:

In 2006 he oversaw the largest round of layoffs in Intel history when 10,500 (10% of the workforce) were laid-off in an effort to save $3 billion per year in costs.

[willow pike]

10nm failed in every way and they didn't know how to fix it

14nm had a rocky start, hence no 5th gen big desktop part

hubris combined with lack of expertise

[twin dew]

And then Krzanich:

In 2015-2016 Intel eliminated more than 15,000 jobs companywide which is now the largest downsizing in the company’s history.

[willow pike]

the cut I was talking about was 2022

it's a wonder they have any employees left

[twin dew]

In some US companies, 10% layoffs each year is the norm...
Each manager has to fire 10% of their underlings, even if all are top-notch.

[willow pike]

and they wonder why they bleed expertise

[twin dew]

Or why the internal politics get to be the most important thing for the employees, instead of actually doing something.

[willow pike]

like how the emerald rapids team designed a chip that the packaging team couldn't get to work, but doing a whole ass tapeout was seen as easier than designing it better in advance

or was that sapphire rapids

the 4 chip one

[wanton orchid]

I mean closing company is basically firing everyone

so they are doing it fairplay at least 🤡

[twin dew]

But Otellini was the finance guy as CEO who started the fuckup, no idea how much the next few CEOs made it worse.
And seems that even the current one isn't really trying to restore the company as first priority.

[willow pike]

there's only so much a new leader can do when every layer of management is rotten

[wanton orchid]

it's the new way of defacing companies, keep the face, swap everything behind with some cheap shit

[twin dew]

But the newest layoffs thing to stay to short term financial guidance when old decisions bite in the ass to cause more expenses is on him.

Instead of just taking the hit and specify the reason.

[willow pike]

sapphire rapids went through 12 settings before it launched

lunacy

though calling it 12 is misleading as they up and redesigned the floor plan almost from scratch halfway through

[wanton orchid]

quick question, any advice on the cnc support vs 3d printed support argument ?
cnc should still be relatively cheap to prototype right ?

I tend to think 3d printing is more robust

because no broken material bounds

[twin dew]

Not really.
Plastic 3D printing with the cheap printers is fragile, as the layers don't bond together that well.

With CNC you have a base ingot that you cut stuff out.

Or other base form, might be partly formed already, a forging, casting etc.
But the base form you start with will be much higher quality than most 3D printers make as finished product.
And the cutting done with the CNC doesn't really affect the base material properties.

[wanton orchid]

ok

I would go with ABS, which iirc is known to be unsensitive to cuts

I know cheap 3d printing is cheaper than cnc
but I wonder on the higher end of quality, how does the cost compare actually

[twin dew]

Cheaper end 3D-printing is mostly about getting the shape right, then the actual production is with something else.
Except for hobby-level stuff and very small runs.

[wanton orchid]

I'm looking for small runs

and starting as hobby

I know cnc have way higher yielding

but I'm not intersted about that rn

I was only looking for comparing build quality, with quality 3d printing, with 3d permitted mechanical optimizations (honeycomb, etc..)

at similar price points (high)

[twin dew]

You really cannot do honeycombing etc. with CNC in one pass, need to do multiple parts and attach somehow.
And 3D printing is additive, CNC is reductive with the head of the CNC needing access to all the places to make the cuts, so more limiting and lot more lossy in terms of material.

But higher quality end-part, if it can be made with CNC.

[wanton orchid]

I know for the complex 3d part, that's why I specified I'm talking to compare with 3d printing optimization included (which means the cnc version would not be that optimized)

[twin dew]

Those are about saving material, time and weight with the 3D print side.
Not getting higher quality result in other ways.

[wanton orchid]

it's mainly for things like mini case components (3d)

my take is if the 3d printing is made so strain goes alongside the fibers you are not loosing much physical properties

[edgy hazel]

very depending on material and settings

[edgy hazel]

ABS is old 3d printing. Nowadays you can get pretty solid stuff with other materials which don't want to kill you. GF-PTEG is pretty solid and isn't as brittle as ABS

although you should print glassfiber with good ventilation and any post processing (like cutting/sanding) should be done with a mask

[languid gulch]

i feel like being a corporate spy at these places at these times would either be hilarious or terrifying

you'd get to see the shitshow up close & personal

[night girder]

Huh, I assummed this was said by a scientist. But apparantly, it was Donald Rumsfeld the States Secretary of Defense. Makes also sense 😄

The idea of unknown unknowns was created in 1955 by two American psychologists, Joseph Luft and Harrington Ingham in their development of the Johari window. They used it as a technique to help people better understand their relationship with themselves as well as others.

[languid gulch]

the problem is that it's a decent explanation of what incompetence is, it just sounds dumb until you really dig into it

[night girder]

How is recognizing "know knowns, known unknowns and unknown unknown" a sign of incompetetence, I am confused/lost?

[languid gulch]

the known unknowns & unknown unknowns are where incompetence is

recognizing it isn't incompetence

[night girder]

When you look at it from leadership, maybe? From a scientific point of view it makes a lot of sense.

In science, they known what they don't know, but they also know, they don't know about unknowns either.

[languid gulch]

oh yea he was trying to take a scientific concept & apply it to policy, which oftentimes doesn't work

people like to use it to justify stupid actions in order to get what they want

i think that's called "weaponized incompetence"

[night girder]

Ok, now that I read the context. Yeah, that's stupid 😂

Reports that say that something hasn't happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know.

[languid gulch]

also at the time, he explained it badly, which the media ran with & memed to death, which parodied the actual concept

[night girder]

So because a report said something didn't happen, in his logic, there is a chance it did happen. The unknown unknown.

You can justify about anything with that logic.

[languid gulch]

(yup)

[night girder]

An unexpected usage spike resulted in Azure Front Door (AFD) and Azure Content Delivery Network (CDN) components performing below acceptable thresholds, leading to intermittent errors, timeout, and latency spikes. **While the initial trigger event was a Distributed Denial-of-Service (DDoS) attack, which activated our DDoS protection mechanisms, initial investigations suggest that an error in the implementation of our defenses amplified the impact of the attack rather than mitigating it. ** 🤣 - Microsoft Azure got DDOS attacked. And then their own defense, made it worse. (azure.status.microsoft/en-us/status/history/#incident-history-collapse-KTY1-HW8)

[languid gulch]

that capitalized F in AFD is doing soooooo much work 🤣

[edgy hazel]

AFD is also a far right party in germany

[languid gulch]

yup

but it's AfD

[edgy hazel]

my stickers spell it differently 😂

[languid gulch]

this was on reddit today

[night girder]

They are making a little house for their kids.

[languid gulch]

oh i know, but also 👀

[edgy hazel]

(while unsurprisingly also having the highest rate of child abuse cases)

[night girder]

Yeah, I know ...

[languid gulch]

like, they could have had their hands tilted sideways & had the same effect. nope, couldn't do that

[night girder]

What does schutzen mean? Shelter?

[edgy hazel]

more "saving"

standard anti-lgbt rhetoric

[night girder]

this took an unexpected turn ...

I just wanted to share some news about Azure getting ddos attacked and making it worse. 😂

[edgy hazel]

yeah lol. Microsoft's incompetence leads everyone to the worst of places

[languid gulch]

lol sorry, saw that meme this morning & the AFD acronym made me immediately think of it

i do love the idea that even microsoft does the "everything we're doing is making it worse!"

[night girder]

I try to avoid reddit, especially since their latest move to block DuckDuckGo. www.msn.com/en-us/news/technology/reddit-is-now-blocking-major-search-engines-and-ai-bots-except-the-ones-that-pay/ar-BB1qySUK

[night girder]

Yeah, they will investigate. Atleast that's what they say.

[edgy hazel]

filament dryer should arrive tmrw. Maybe then I can print with my brand new roll of TPU

love paying 35€ for a 500g spool full of WAP

(wet ass plastic)

[languid gulch]

i've been tempted to buy a 3d printer to make little things that i need around the house

[jagged snow]

THey're handy

[edgy hazel]

I've been doing a lot mounting and holding stuff. Also phone cases

and ofc, a shitton of DnD

[gilded helm]

Bloody hell, when is UPS going to stop going to businesses when they're closed? What thee hell is the point of them adding a whole set of form fields for businesses hours?

Big data has been good enough to optimize their routes to avoid left turns in the USA for at least 15 years. But not avoid second trips due to going to businesses before they open, apparently. -_-

[languid gulch]

that kind of message from GN usually also indicates they're contacting govt officials

[wanton orchid]

why would ABS want to kill you ?

[edgy hazel]

fumes

[wanton orchid]

but that's only when printing right ?

and you said glass fibers have problems with airborns too

[edgy hazel]

yea

[edgy hazel]

much easier to filter than VOC

and after looking through some tests myself, ABS strength in prints is highly overrated

[wanton orchid]

PETG have better strips tie ?

[edgy hazel]

PETG might be a bit harder to print (normal petg is a bit gloopy), but it tends to flex instead of break like ABS does. Combine that with some fibers like carbon or glass (chopped fibers, not ground) and you got a pretty hard and strong print, with better layer adhesion too

The printed stuff is all normal PETG

Fan shroud, Direct drive mount, Spider hotend mount

I still need to do some cable management so it doesn't look like a bomb

[wanton orchid]

https://www.youtube.com/watch?v=HMXTeuhShaA thank this guy very much, petg indeed looks nice

[edgy hazel]

PETG is what you print for high durability until you are able to print nylon

[willow pike]

oh no

[languid gulch]

oh i can't wait to see the evidence

[tough owl]

I wish GNs merch wasn't so flashy

Because I'd buy a shirt

[willow pike]

i got their screwdriver set, it's pretty good

but i prefer the LTT screwdriver

[edgy hazel]

☹️

Rip balanced fan

Noctua replacement has already been ordered. Sometimes you gotta be forced to upgrade

BTW thank you noctua for making a 24V fan

[twin dew]

Industrial fans are very large part of Noctuas business.
Both generic and made-to-order ones.

[charred pewter]

how did you break your fan blade

[twin dew]

Stuff broke earlier, and the fan was probably still spinning.
And one of the blades took enough damage to start a crack.

[edgy hazel]

That was a joke, I took that pic during disassembly.

[twin dew]

Ok.

[edgy hazel]

But yeah I accidentally touched the spinning fan with pliers

[twin dew]

But I made a mistake and ordered Ryzen 5600X for 142e.
When I should have ordered 5700X for 180e.
And I'm not going to fuck the local store over by trying to swap the order.

The 4 extra threads would just have given more headroom for more VMs before amount of vCPUs goes over the amount of HW threads.

Even when most of the VMs shouldn't be anywhere near saturating even one half-core.

But I will also have extra AsRock B450 board and extra 2x8GB of RAM when this is over, so will have to see if I will get better CPU too at some point before the 5000-series ones go out of stock.

[jagged snow]

I don't understand how a company as big as apple can provide such a bad experience

Trying to create an apple id through appleid.apple.com does not work on windows

Their support page says that the only supported way to create an apple id on windows is thorugh the APP

You have to use apple tv, apple music, or itunes to do it

How annoying is that

[charred pewter]

oh, it doesnt stop there 😄

[twin dew]

Any guesses on to what the difference is between the two 65W options given?

[charred pewter]

we publish games for small devs on app store, and ... there is a massive amount of "wtf apple?" with the entire process

[twin dew]

(As in I know at least one major difference between those options, but does anyone else?)

[charred pewter]

another popped up last week when we looked into "Why can app titles only be 30 chars max?" ... turns out, they used to allow 255 chars... then dropped it to 50 ... and the reason was people were changing titles to include 'sale 50% off now!' type shit in them. so now apple reduced them to just 30 chars, which is barely functional for many games/apps

on top of the reduction, they also now force that "once you set a title, you can never change it again" ........ so like, if they put that stipulation in, why not raise the char count again ?!

🤦 🤦

and dont get me started on "only the owner of the developer account can accept this new terms and policy change before anyone in the company can do anything in the portal" bs .... most imes the 'owner' of the account is like some joe who set it up before they left the company or something, or its the CEO of the company who doesnt even know how to sign in again

regardless of setting some competent people as 'full admin' ... those accounts cannot accept the new policy changes 🤦

i'll stop .... apple is so full of shit 😄 i could go on for days

[jagged snow]

My guess would be a difference in voltages and power delivery agressiveness

[twin dew]

In a way, 65W Commercial is actual 65W limit, not the 88W like the consumer.

[jagged snow]

Makes sense

[wanton orchid]

where is system config 4 and 5

[twin dew]

Would probably be for higher max power?
But I have 65W CPU installed.

[wanton orchid]

isn't this options like asrock rack tuned profiles ?

[twin dew]

This is in AGESA itself, so AMD code.

[wanton orchid]

for different kind of board use case ?

oh

So not asrock tuned

I would say kind of same as fight or flight
except I mean :
consumer one is probably a bit less safe but more sweetspot for performance

[twin dew]

Consumer is the "by the official spec" one, and the Commercial is basically 65W hard-limit ECO mode.
Just that those ECO modes are new terminology for AM5, but the same thing existed in AM4 in this AGESA setting.

So with 105W spec CPU, there would probably be around 90W ECO equivalent as config 4, 105W Commercial (105W ECO) as config 5, and then the 105W consumer as config 7 or 8.

Depending if that around 90W one had consumer equivalent or not.

[jagged snow]

After an hour of trying, I cannot create an apple ID

How on earth do people use their products

I thought they were supposed to be the "just works" software company

[twin dew]

"It just works", if you use their products, and don't try to do anything that is officially supported.

Like interfacing with non-apple products or software.

[edgy hazel]

To change the owner you'll need to call them and commit some identity theft by acting like you're to owner of the account.

Don't ask how I know

[charred pewter]

but then your account is the one that becomes inaccessible when you leave the company or are off on vacation and cant sign in to approve a stupid document 😄

[edgy hazel]

https://tenor.com/view/circle-marriage-love-like-a-circle-never-ends-gif-12894046

[jagged snow]

I suppose
As far as I can tell, it's not even currently possible to create an apple id from a windows device currently

[edgy hazel]

Password sharing is always an option

[twin dew]

But again against the ToS.

[charred pewter]

but the 2fa ...

[edgy hazel]

Bitwarden

[charred pewter]

goes to the owners phone .... but you can add multiple phones, but then it violates their rules

we have to maintain an account for another company, and we are always running into this issue.... we could try to sign in to the account ourselves, but then we have to still get the owner to send us the 2fa code that ends up on his phone LOL, and the guy is like ALWAYS on vacation

[edgy hazel]

• Create an account like [e-mail]
• add login info and 2fa to company bitwarden
• done

[charred pewter]

and apples stupid system asks for that damned code like every 24 hours

you cant use bitwarden for a PHONE DELIVERED CODE

they dont use TOTP

[edgy hazel]

Oh true sry

[charred pewter]

😦

i WISH they did

[edgy hazel]

• Add IT hotline number for 2fa

And hope you're the first one that picks up

[charred pewter]

heh! yeah, we tried to change the phone number to a shared google voice phone number, but apple wouldnt even accept the phone number change to it lol

"not a valid phone number"... like wat?

[twin dew]

Because Google Voice has specific phone number range that is blacklisted for lot of stuff?

[charred pewter]

appears so

funny how i can use my google voice number to 'validate' my youtube channels i set up 😄

[edgy hazel]

Bro printing TPU is so STUPID
I'm at 130% extrusion and it's STILL underextruding

[charred pewter]

😦

[twin dew]

What temp?

[charred pewter]

herg, did i show you something fun my father 3d printed? https://youtu.be/UibrEBnPHJg

[edgy hazel]

220 currently. But temp isn't the issue. I've tried all between 215 and 240

[twin dew]

Seems to need pretty high temperature (230C or so?) and slow head movement (15-20mm/s).
Ok, so the temp is pretty ok at least.

[edgy hazel]

And over 228 I'm getting terrible popping. Maybe it just needs a night in the dryer.

[twin dew]

But I'm not even novice with 3D printers.

[edgy hazel]

This is the wettest filament I've ever had

[edgy hazel]

Its so wet I've glued a pack of silica into my dryer

[charred pewter]

nice, you made one too!

my father has gone through a few printers ... he is using a bambu right now that he is impressed with

i dont have one ....... i dont have the patience 😄 and seeing you frustrated at times, i ... yeah, i dont have the patience

[edgy hazel]

I'm just very loud when trying to solve problems 😂

[charred pewter]

the silica packet is hilarious

[edgy hazel]

Tbh if I were running my printer with stock settings and normal PLA I probably wouldn't have any issues

I could've also not turned this:

Into this

[charred pewter]

gravity feed maybe?

[edgy hazel]

but alas, I need something to play with

[edgy hazel]

nah I just moved the extruder motor from the side directly ontop of the hotend

it allows me to print with flexible materials, which would normally get stuck or bunch up in the tube

[charred pewter]

ahhh

[twin dew]

But having the roll above would still mean much less pull-force needed for that motor.
Which can be problem if the material is too hard and smooth.

[edgy hazel]

which is why I went for a dual gear extruder. It's hard to pull stuff through there even with the stepper motors not locked

But the side thingy is only for stuff I print directly from the dryer anyways. I still have a top mounted spool for my not as thirsty filaments

BTW @charred pewter do you wanna know how you make 3D printing much more enjoyable?

[charred pewter]

ask someone else to print for you ?

[edgy hazel]

No...

[charred pewter]

😄

[edgy hazel]

Like this

[charred pewter]

ive got my father making a bunch of car parts for me

grommets, clips, fasteners, n shit

thats how you make everything much more enjoyable

[edgy hazel]

Groundbreaking tech

[twin dew]

Tress-FX remade more shittily?

(The AMD tech released as open source in 2013, and the name was just TressFX)

[stray badger]

why teapots nvidia. why

[twin dew]

https://en.wikipedia.org/wiki/Utah_teapot

[twin dew]

But basically both 3D-community injoke and free "select your detail level" model.

[willow pike]

https://twitter.com/KobeissiLetter/status/1819104168919158965

The Kobeissi Letter (@KobeissiLetter) on X

BREAKING: Intel stock, $INTC, crashes over 12% after reporting Q2 2024 earnings and suspending their dividend.

They reported EPS of $0.02, below expectations of $0.10, on revenue of $12.8 billion, below expectations of $12.9 billion.

Intel also announced that they will be

Twitter · Aug 1, 2024, 8:13 PM

[languid gulch]

steve utters & stocks quake

[edgy hazel]

The stock market was a mistake

[stray badger]

welp it looks like we may be down to 2 GPU makers again

[edgy hazel]

yeah they shat the bed with their 13&14th gen

[languid gulch]

and 15th

also we're yet to see anything definitive out of the oxidation issue except rumors that it came out of the Chandler fab & that "it's definitely fixed guys, trust us"

[languid gulch]

haven't seen any info as to if this is going to affect ARC development

[soft bloom]

Why is it so hard to find a way to log temperature over time?
Like i can't even google it in store cause they have some weird stupid ass name (not in english)
And it's not like i want to write code and find adapters to make it work on raspberry

Also, why is it that a model of a multimeter without thermal couple actually lacks the functionality for it, and not just the probe itself? I mean, i can't upgrade it 😦

The name itself is stupid

Thermometer with data recording

Like calling a car "small family vehicle for everyday use not racing no truck"

Also when it records data it is immediately separate category from all other electronic thermometers...

Shit i might as well use my pi3 since i have no desire to pay that much for something that has driver on cd disk and it only supports windows 7

But then i need to figure out how to connect standard thermal couples to pins on raspberry

Well, a mediocre solution would be to buy thermometer with a nice display and probe. Or just a simpler (non electric one). And log by hand

NightHawk has so why i can't? Also technology connections has a couple of them

I want to understand if my variation of pcm stabilization is any better (while being cheaper)

So i need to log data over time and see how fast it heats up and at what point it transitions phase

I wish there was a clever way but i don't think that's the case

He had these usb thermometer dataloggers

Note: measuring temp of liquid inside double plastic bag is more like guessing game: even at the same point readings jump 2C each time
Note: that is not normal behavious for ir measuring

I can't wait.
Also i found one locally and it has driver on cd disk and it supports win xp an 7 and costs too much for me to care

Really, 6 years ago i brought rasp3 for almost same money

(note: that was before dollar gone lose)

[edgy hazel]

for just temp logging a raspi1 would be enough. But you could prolly go cheaper with an esp32 and some tinkering

[languid gulch]

aaaaaand youtube ads felt like showing up again

all of them unskippable

it should be like cable or satellite, your ISP bill should pay to not have ads on the internet

good point, internet ads should just be banned then 🤣

in-video sponsors are fine

getting spammed with bullshit is not

nope

hence my complaint

oh i'm just waiting until there's literally no other option before doing that

i think the big problem is coming up with an alternative to switch to if things go bad

google could just choose to pull the plug on anyone suing them

the only thing that'd work without losing a bunch of communication ability would be making some kind of globally public utility

think i figured it out

migrating over to firefox again 🤣

sometimes stuff that works on chrome won't on firefox, & vice versa

i just hate the migration

such a time waster

for a long time a bunch of stuff was just broken on ff

none of them are great

just less shit at the moment

yea i haven't used safari in close to 20 years

i think netscape navigator was still around back then

🤣

or if it's gonna be a monopoly, it should be an international public utility

🤣 hey i'm the one who suggested it

there's good, bad, & crazy everywhere

and plenty of stupid

[stoic onyx]

Opera is still my preferred browser, but I think I’m very much a minority.

I originally picked Opera coz it was really fast, but this was way way back. Just stuck with it ever since.

[languid gulch]

yea but it can be cheaper

[pure karma]

freedom is nerver free its just a whole lot of dom(ination)

[soft bloom]

The what?

[soft bloom]

And then 80% will go down since no easy way to maintain. It will be get occupied by tech guants that will insert ads in your toliet's flush since there's no one to compete against them

[edgy hazel]

Its the best browser for macbooks

Saves a lot of battery

[tough owl]

Webkit has gotten really good recently

[languid gulch]

ah crap, firefox doesn't have tab groups. that sucks

[charred pewter]

tab groups?

[languid gulch]

like on chrome & edge

[charred pewter]

i use chrome ..... tab groups?

trying to see what that is lol

[languid gulch]

you can cluster groups of tabs into a folder looking thing

where the tabs are, instead of a bookmarks folder

[charred pewter]

you mean the favs bar with buttons ?

[languid gulch]

no

[charred pewter]

hm

[languid gulch]

tab groups on the right

[charred pewter]

if i try to drag a tab onto another tab, they just shift places

[languid gulch]

i click on the blue one, & there's a dozen "sub tabs" i guess you'd call them

all on that subject

[charred pewter]

weird

[languid gulch]

great for organizing

[charred pewter]

although i dont keep a dozen websites open, cause ... chrome ...

[languid gulch]

and absent from firefox

i've got plenty of ram to spare, so i'm ok

[charred pewter]

but i do have three to four 'profile' windows open, all their their own tabs and sessions

firefox didnt support multiple profiles until recently ... which was the leading reason i never switched

but with chrome being more and more enshittified ... i may be switching to firefox soon 😉

[languid gulch]

i'm in the process of switching since chrome nuked youtube ad blocking

i refuse to watch 2 minutes of ads on a 2 minute video

[charred pewter]

i still dont get that ... i go to youtube all the time and never see "these ads" people complain about ...

[languid gulch]

depends on where you live

[charred pewter]

i just have ublock origin installed

[languid gulch]

yea, mine stopped working for youtube ads

[charred pewter]

i live in the capital of capitlism !

[languid gulch]

even extra filters aren't working now

[charred pewter]

that sucks

[languid gulch]

no idea why

still working on firefox tho

same settings as on edge

only difference is ff isn't chrome based

[charred pewter]

yeah google is so full of shit lately

just today i found out that google has decided to stop sending ME some emails to a 'group' ... yet everyone else in the group is getting the emails ... but nope, google decides 'the owner of the group should not get any of these emails from this important service provider'

[languid gulch]

wtf

[charred pewter]

but im getting emails from OTHER providers

yeah, i checked everywhere, spam, settings, everywhere. google is just flat out not sending ME the emails from one provider we use

and im like the top technical contact too and NEED those emails to react on ! 🤦

[languid gulch]

shoot them an email or a phone call letting them know that they'll be billed for lost hours for failing to comply

[charred pewter]

heh! right, ive been down rabbit holes with trying to find someone at google to talk to about things

[languid gulch]

or just send em a bill, they might just pay it to make it go away 🤣

[charred pewter]

OLOL

Come on Intel .... get it together, pull things around!!! I dont wanna have to buy AMD for my next build!!!

[languid gulch]

oh 15th gen is equally fucked

[charred pewter]

😢

dejavu

[edgy hazel]

come team red

we have cookies

[charred pewter]

but cookies are BAD according to the internetmob

[willow pike]

intel don't love you

don't love them back

(neither do AMD, companies cannot love you and do not deserve allegiance)

[languid gulch]

"buy ryzen 9000 series, at least it doesn't melt!"

[twin dew]

And after long and hard road, the first VM finally boots to login prompt \o/
Combination of stuff like disabling graphics so needing serial console.
Also that not including a virtual disk in boot order caused it to not show up for Grub2.
Etc.

[languid gulch]

at least intel has finally admitted that any 13th/14th gen that's drawn over 65W is screwed

[willow pike]

isn't that all of them

[languid gulch]

how observant of you

[edgy hazel]

wait gimme sauce

I wanna see

[languid gulch]

https://arstechnica.com/gadgets/2024/08/intel-extends-warranties-by-two-years-for-crash-prone-core-desktop-cpus/

[willow pike]

god only knows what the gamers nexus exposé will be about

[edgy hazel]

"like the Core i5-13400" jesus

[languid gulch]

i wonder what the delay between that tweet & the post is about

[twin dew]

https://www.reddit.com/r/hardware/comments/1ei1zvm

[willow pike]

video editing and fact checking probably, plus lawyer consulting

[willow pike]

oh

[twin dew]

Then there is stuff like this that points that Intel is in full on "Deny all RMA:s we can" mode.

[willow pike]

oh

[languid gulch]

i'm wondering if they're also getting lots of interesting phone calls from some people

[willow pike]

oh no

[edgy hazel]

man they fucked their shit up

[willow pike]

oh dear god

[languid gulch]

yea this screams potential bankruptcy to me

if they actually do what they're supposed to do

[edgy hazel]

I think that's why they're doing this shit now

[languid gulch]

yup

ok, i need to run this past someone who might know wtf could be going on, stay with me for a minute

the overvolt issue

[willow pike]

there are multiple possible causes even if we take intel at their word

which I am not inclined to do

[edgy hazel]

nvidia CPUs coming soon

[willow pike]

they already make CPUs

[languid gulch]

could it be that there are just too many cores for the "entry voltage" going into the CPU if it's only something like 1.35-1.5v, and by the time it gets to the cores "at the end of the line" the voltage has dropped too much to be useful?

[willow pike]

they don't have an x86 license but the recent ARM on windows push opens up an nvidia CPU possibility

[willow pike]

if it got too little voltage it would not degrade

it wouldn't be stable but it wouldn't degrade

[languid gulch]

right, but that would explain why it'd ask for more

if the "end cores" are asking for enough voltage to function, but that'd mean it'd have to start out way higher, enough to degrade

[willow pike]

intel has made single-die 28 core CPUs without having such a problem

[languid gulch]

idk cpu architecture (obviously)

[languid gulch]

right, but maybe it's a design flaw inherent to LGA1700

[willow pike]

it probably helps to understand that voltage is transmitted at a much higher layer than the transistors are, it's not flat

[languid gulch]

it'd be interesting to see a schematic plus where the failures are occurring

[charred pewter]

sigh

[languid gulch]

i mean, could that kind of weird issue be a possiblity?

[willow pike]

raptor lake, 13/14 gen

(I googled "raptor lake die" and got posts about it dying lmao)

a side-on view of metal layers

the voltage will be well above the cores before it goes down to them

[languid gulch]

i'm just trying to think of why it'd ask for that much voltage to begin with, & my only non-electronic brainicles could think of is "if it's DC, then it's fighting resistance, & maybe the distance it has to go requires it starting off at a higher voltage"

[willow pike]

they say it's asking for too much because of a bug

many people think they told it to ask for too much on purpose to get a few extra MHz out of it

[languid gulch]

that's one hell of a fuckin bug

[edgy hazel]

expensive bug

[languid gulch]

trillion dollar bug

[willow pike]

it's completely unbelievable they had no idea this would happen, or that they didn't catch it for two years

[languid gulch]

it's literally unbelievable

i can't believe that the one thing that they do, & have ever done, they'd screw up this badly

[willow pike]

intel manufacturing used to be the envy of the entire world

[languid gulch]

now it's a Yugo dying at 40k miles

[willow pike]

about the time i ran intel, core 2 duo, core 2 quad, the first core i7

[edgy hazel]

give it up for capitalism

[willow pike]

(I am old)

I've been eating my popcorn while intel slowly recedes for a while now but this

this is beyond the pale

[languid gulch]

i think the US should take it over for a bit like they did with GM & Ford during the 2008 crisis

plus the hints about them pumping the voltage while binning to make them seem like they're a higher bin when they're not? that's like selling a Mercury as a Lincoln because you're using the same paint

[night girder]

I don't get microsoft (source: answers.microsoft.com/en-us/skype/forum/all/skype-insider-release-notes-for-skype-812576201/70072f8d-3a54-4739-ad2f-dd0cdc665a6f)]

They bake ads into Windows while removing it from other products. What idiot is leading Microsoft marketing?

[languid gulch]

if i had to bet, an intel or volkswagen transfer

[night girder]

The only logic I can think of, they put ads into a product, monitor it for profitability. If not desired profitability, go more hardcore and put it into Windows and stuff. OR they realize Skype is a failing product (discord, slack, ... competition is better) and this a last attempt before throwing in the towel.

[languid gulch]

i do wanna know who the geniuses are that have convinced someone to pay them a lot of money to tell others that all they have to do is push for more ads

[charred pewter]

SEO peeps

[night girder]

Right? Like, sometimes it feels that Microsoft news is more about ads than AI this year.

Something is going on for sure.

"2024, the year Microsoft tries to refocus their ads"

[charred pewter]

microsoft has ads? maybe they should get that looked at

[night girder]

they are?

[languid gulch]

they are, then they check their bank accounts, & suddenly everything's fine

[night girder]

They are always fine. They just want more.

But somebody said they needed to focus on ads this year, I am sure of it.

[charred pewter]

im not sure if i said that, but we do ... lol

[languid gulch]

https://www.tomshardware.com/pc-components/cpus/intel-says-13th-and-14th-gen-mobile-cpus-are-crashing-but-not-due-to-the-same-bug-as-desktop-chips-chipmaker-blames-common-software-and-hardware-issues

same crashing as desktop, but intel's claiming it's a different issue

[night girder]

lol

[languid gulch]

i'll be calling BS

[charred pewter]

"its all the motherboard mfg's fault"

[languid gulch]

i feel like the mobos should sue for defamation at this point

[safe trench]

for doggos sake just get 12th gen

you guys are forgetting the whole fact that amd cpus were destroying the mobos socket

and then a bios update fixed it all

if intel does something just slightly different

everyone begins hating on them

but if amd does the same thing but worse no one will care

its like you guys are fanning over amd like intel has no tomorrow

"oooohh guys look at this ryzen 7800x3ddddddddd!!!!!"

as if anyone has the money to afford that

you guys honestly forget the budget systems

and howmany there are out there

[cyan crescent]

Pressure testing the new build

[stray badger]

7800x3d is not that expensive, if you are on a budget, 5800x3d is also good

[cyan crescent]

Nah. Get the 7950X3D

[stray badger]

Twas asus problem with x3d voltage

I did get the 7950x3d

[safe trench]

same thing with intel

the mobos are overvolting the cpus

and since its intel its like its going down for intel

[stray badger]

Its fixed, intels is not fixed

[safe trench]

EXACTLY

and you know how amds was fixed?

[stray badger]

And it was only asus

[safe trench]

A BIOS UPDATE

[safe trench]

its only high end mobos

[stray badger]

My 7950x3d has not exploded

[safe trench]

intel is just having the same problem of amd

just super similar

[stray badger]

But much worse

[safe trench]

not quite

honestly

for doggos sake

[dire igloo]

Can't fix hardware OCP failures with software

[safe trench]

get an 12 gen, you dont need the extra 20 fps

[dire igloo]

You'd need a new board revision for that

[safe trench]

(not mine)

[dire igloo]

Speaking of the OCP failures that GN found during X3D failure testing btw

Not the X3D failures themselves

[safe trench]

well its similar to intel

• at least its not making your pc a fire hazard

i know its not like a pc can burst into flames like that

but still

[dire igloo]

PGM: am I a joke to you?

H1 v1

[cyan crescent]

Should i move the return to the bottom on my build?

[willow pike]

AMD were very specific about voltage limits for AM4, motherboard makers were ignoring these limits which only became a problem for X3D. AMD immediately mandated BIOS fixes to actually follow the fuckin' spec, and made customers whole with warranty claims.

motherboard makers were also fucking about on intel, but with intel's implicit blessing, and this problem is completely separate from the: oxidation, microcode bug, and completely different mobile bug they haven't defined. and they're being fucky with the warranty.

saying they're both the same is disingenuous at best

people have been sounding the alarm about this raptor lake thing since february and only now are they releasing an update which may fix it

[safe trench]

just dont use your pc till it comes out as long as its not work

[winged valley]

Okay so I have a DMG with the full 6GB installer for el capitan
The installer, not a .pkg
I use balena etcher to make a flash drive with it, and it doesn't warn about it not being bootable
So why does it not show as bootable
Im a windows guy not a mac guy, so what am I doing wrong here

Im on an m1 mac so that complicates things

[stray badger]

also, only a few mobos had a problem. EVERY SINGLE INTEL 13/14th GEN has this flaw

[languid gulch]

also intel trying to place the blame on mobo partners knowing that they themselves were the problem is extra slimy

also i'll shit on anyone who'se choosing to sell bad stuff to customers, i'm brand agnostic. roasted the fuck outta AMD when they started frying boards, & Intel deserves at least as much monkey shit thrown at them, if not more

[pure karma]

so far it is tho its 50/50 intel and mobo's

intel fucked up and the mobos fucked up making a 200% fuck up

[languid gulch]

i place it 100% on intel. the CPUs being bad from the start means that it wouldn't matter what the mobo manufacturers did

the bullshit about intel not giving specific parameters, & the mobo partners not requiring it, is a separate issue to me

actually, intel refusing to give parameters might have been a sales tactic. knowing their chips were shit, and that way when things started going wrong, they could first distract the market by deflecting blame to the mobos

that bought them 6 months

[pure karma]

honestly truth is everyone is in the wrong here and no one wants to just do the right thing and then move on

[languid gulch]

the big problem from intel's view rn is 15th gen. it has the exact same problem, only now everyone views them as shit, & rightly so

[pure karma]

honestly im not woried about them

[languid gulch]

oh i don't have to worry about them

[pure karma]

people are gonna buy them anyway

[languid gulch]

i care about the consumers & how they're getting screwed

[pure karma]

yea true the lack of a recall for thoes affected was too far

mine is still good

[languid gulch]

or it's degraded & it hasn't shown up yet

could go unstable tomorrow

[pure karma]

overclock prob making up for it

i havent done any prevention im just rawdoging it

[languid gulch]

swapped out a light fixture today, previous one was never grounded

[wanton orchid]

damn microshit still showing off their skill

iscsi booting into winpe the iso
makes winpe ask for driver when wanting the media
... yet the actual media iscsi is already explorable in file explorer

[languid gulch]

i heard 365 had brain farted

[wanton orchid]

this is quite a big joke

so basically winpe is like "where driver to see media"
showing the actual media already as a location to search in for driver for media

🤡

[languid gulch]

it's like people get too close to the work

gotta hire some noobs that are afraid to make simple mistakes

[wanton orchid]

like .. mf ?

[languid gulch]

lol

[wanton orchid]

this is what shows when you boot from usb too

[languid gulch]

i guess it'd be faster than booting from a tape drive

[wanton orchid]

except it does not try to make it look like it can't read even thought it obviously shows it can read very fine

[languid gulch]

at one point when i was building my system it kept trying to boot from my external storage drive

it jumped clear to the end of the list

[wanton orchid]

wait mb it seems it want to install

it was just it confused itself about cd driver instead of hdd driver

I removed the other cd and let it only have iscsi cd, and it want to install good, showing me all the install options and everything

wait it works too good now

it loads every LUN
(I set up an other iso on an other LUN)

it can be handy to map tool isos to multiple targets then, but it means better use multiple different targets for convenience

[twin dew]

General problem with VSOC voltage.
Asus MB:s just didn't have VRM current protection enabled during pre-CPU initialization, so the destroyed CPU with internal short could pull insane current and burn the socket (also melt the silicon and the substrate).
Every MB was destroying CPUs, but everyone else had VRM output short circuit protection enabled in early boot.

[twin dew]

No, AMD had communicated that 1.4V VSOC was the max, but that was wrong.
And limited it to 1.3V VSOC once the issue popped up.
But non-X3D CPUs had been dying in small numbers from the too high VSOC for months already when it went public with X3D:s burning Asus sockets.

It was some kind of voltage domain interaction thing.
The damage was happening at edge of the IGP inside the IOD.
The IGP in Raphael runs on VCore.
Most of the IOD runs on VSOC.
And then you have the IF voltages in the mix that is third voltage domain.
Was probably that too large difference in voltage between VCore and VSOC was causing the damage, and X3Ds use lower VCore most of the time.

So my expectation is that if the VCore (dynamic) was too far below VSOC (static), damage happened in the IGP and the IF transceivers in IGP.

And AMD just hadn't simulated anything like that when determining the max voltages.

In previous AMD generations, the IGP ran on VSOC, not VCore, but they changed that for AM5.

https://www.youtube.com/watch?v=b6vQlvefGxk

But back to the AMD, three things, two related:

Causes for dying CPUs, happened with all MB makers:
1: AMD determination on max "safe" VSOC as being 1.4V was wrong.
2: MB makers setting the VSOC automatically to near that max to make memory controller more likely to be able to do high RAM speeds.

Cause for Asus burning sockets:
3: Asus having the runtime VRM limits when the CPU wasn't yet operational and wasn't supposed to pull any current, allowing the shorted IODs to pull 300+A and melt themselves internally and destroying the socket.

Other MB makers boards just shut the VRM down and didn't proceed in the initialization with dead CPU.

But that destroyed socket only happened after the CPU had shorted, the computer had shut down, and then the person tried to start it again.
With most makers it just got stuck in post on CPU stage or shut down again.
With Asus it got stuck in post, but burned the socket.

[edgy hazel]

had to check I didn't have my playback speed accidentally set to 1.25x

RIP fruit baskets

[safe trench]

at this point just get a 12 gen

14th gen isnt even worth it, just dispose it

[twin dew]

So give more money to Intel and then give a gift of not having to replace the broken CPU?

[safe trench]

no

what do you want me to do

sell amd my kidney because of a bios update thats not launched yet

[twin dew]

Use the Intel system with current CPU limited to be stableish, until the update is out, then RMA the broken CPU?

Or start the RMA process now.

[safe trench]

at this point 3 gens of intel are a huge hit

then the next 3 are them trying to squeeze extra cash dollars

[twin dew]

If someone wants to throw the new CPU into trash at end of RMA, I'm all for that.
But if someone skips the RMA, that is just giving Intel money.

[safe trench]

true, but its kinda unfair for intel as somegames if they crash it directly blames intel which is true ill give you that but somethimes it wasnt even intels fault and they are the blame

[twin dew]

Intel not checking on MB makers work is also on Intel.
AMD has been monitoring and approving MB defaults, Intel hasn't.

[safe trench]

thats because after they overvolted their cpus it gotten real serious with amd

[twin dew]

AMD has been doing it for years, not just after the VSOC thing.

Intel has intentionally been looking away for 10+ years, and not giving clear guidance.

[safe trench]

at this point why even have 2 brands if ones a dumpster fire for some time

which cpu do people even pick if one often times just blow up with problems

do i get 3 different gpus just to be safe if 2 or 1 blows out with problems? and 2 different mobos with 2 different cpus???

[twin dew]

But to completely different thing:
LibVirt documentation is bad and Virt-Install documentation is just horrible...

[safe trench]

bro

i hate intel intergrated graphics

it has a major problem with running minecraft 1.7

and if you open up a support ticket writing a paragraph, guess what

[twin dew]

Which IGP generation and what driver version?

[safe trench]

they dont lay a finger to give a damm

[safe trench]

whats and igp generation? and it was somewhere mid august i think

[twin dew]

If you don't even know what are Intel IGP generations, then what CPU?

And you also don't seem to know how to check the installed driver version?

[safe trench]

it wasnt any time recent, acutaly a long time ago!

[twin dew]

Too lazy to actually write the model?

[safe trench]

yes.

https://community.intel.com/t5/Graphics/Minecraft-showing-as-one-solid-color/td-p/1450228

1.7 doesnt work in general

[twin dew]

And later game version fixed the issue?
If game isn't following spec, then it isn't always the GPU makers job to make their drivers accomodate the games faults?

[safe trench]

yeah

[twin dew]

The point is that drivers expect spec compliant input, games are supposed to output spec compliant.

[safe trench]

but still i think as the consumer we shouldnt be the ones to just straight up downgrade our gpus with an older type of the game just to play some 1.7 game

i wouldnt care but guess what! all the good mods stop there

after they told dean that it was fixed on an older version they just dropped him

[twin dew]

But Gen 12 (first gen Arc/Xe) in that CPU.
And the drivers were a mess at start, been getting better over time.

[safe trench]

iris xe

[twin dew]

That is just branding.

[safe trench]

oh?

well then

[twin dew]

https://www.intel.com/content/www/us/en/download/785597/intel-arc-iris-xe-graphics-windows.html

But again, the "Iris Xe Graphics" is just branding.

[safe trench]

is that latest or the one before minecraft just got destroyed

[twin dew]

32.0.101.5768, 7/29/2024

[safe trench]

oh latest

[twin dew]

But the whole Xe arch can only do DX12 natively, everything older is emulated.
And that emulation has had major issues.

Xe-LP version in the Gen 12 IGPs.

[safe trench]

DAMMIT

but nope intel just cannot fix it

[twin dew]

Been getting much better with time, but as baseline.

Because putting lot of work for just one old game version isn't worth the time?
And it might get fixed or might already be fixed now from improvements for other stuff?

But there is reason newer driver versions have given in cases 400% FPS improvements to some games, when the game had been completely broken before.

And why GN still says that don't get the dGPU Archs if you don't want to roll a dice with older titles.

And IIRC Java Minecraft uses DX9.

[safe trench]

if im gonna be honest you might be right and intel is honestly the only gpu producer im rooting for currently only because their first gen chips were insane for a first gen

[safe trench]

oh really?

[twin dew]

And the point why mods stop at 1.7.10 might be that the MS devs made major refactoring and upped things like that, so the mod makers would need to almost completely rewrite their mods to support the newer game versions.

[safe trench]

i mean there is an 1.20 verison of one of the mods that i really wanted to use but didnt work

my compliant is that its not even forge causing the problem if i try it regular it wont work

[twin dew]

But it is also known issue that the Intel (and AMD etc.) support forums are place where stuff goes to die.
Getting actual support is almost impossible on any of them.

The lowly peons who are in there from company side aren't allowed to push things upwards.

[safe trench]

of course it is theyll dodge support claims harder than lebron can toss balls

[twin dew]

And in many cases the "official" responders are not even paid but are communicy volunteers.

[safe trench]

WHAT

[twin dew]

And not employees.

The actual support forms go to employees, but the forums are often just smoke-screen.

[safe trench]

i feel like the govenor should give regulations on support forums for the consumer not to just have to rely on the full hearted community instead of the bastards the people who moderate and support these fourms

[twin dew]

But back to the LibVirt documentation issue:
The LibVirt documentation documents all the options, but doesn't tell what they do or mean.
Virt-Install documentation just mostly tells to refer to help-commands of the tool (which just give the options, not what they mean or do or how you use them), and to refer to LibVirt documentation (where the options don't match 1:1 to the Virt-Install).

[safe trench]

dang

[twin dew]

Oh, and for some hypervisor specific options the LibVirt documentation doesn't even mention them.

But Virt-Install can add them etc.

And the names the LibVirt wants for them don't of course match to the hypervisor documentation, as those are for their own config file formats or command line switches, not for the XML LibVirt uses.

So if you don't use the GUI Virt-Manager, you are fucked.
Don't know how fucked you are if you use the Virt-Manager.

[twin dew]

For most things LibVirt documentation about the options is pretty complete, but for some others it is just weirdly incomplete.

Mainly for concepts that the writes think self-evident but are highly linux nitty-gritty specific.
Like what are the effects of using file vs anonymous vs memfd for virtual machine memorybacking, and what does the whole memorybacking concept means etc.
"only" problems because I don't have enough previous knowledge, but would be nice for the documentation to provide the information or point to somewhere it can be learned.

access
Using the mode attribute, specify if the memory is to be "shared" or "private". This can be overridden per numa node by memAccess.
But question is, when should you set the memory as shared ?)
And what is the default ?) (almost certainly private)

[languid gulch]

i can't wait for the intel bullshit to get locked up in court for like 5 years

[twin dew]

And all the exec level people to bail with golden parachutes in meantime.

[willow pike]

fascinating

if i remember correctly, part of the problem was also board makers not using the extremely expensive voltage regulators properly and instead pegging it to max, which won't help if the max is also wrong

[twin dew]

Just norm that MB makers have been overvolting the various subparts of the CPU to the max to give best chance that whatever XMP RAM the consumer bought would work.

[willow pike]

yeap

[twin dew]

Several were using 1.35V setpoint as default once you set RAM frequency that was above the 5200MT/s.
Some boards might have been using 1.37V setpoint.

For VSOC, other voltages related to the IMC are also pumped up.

To level that Asus shows the "default on XMP" voltages for some of them as red or purple (extreme danger).

VDDP is default 0.7V or 0.8V, most MB makers still set it to 1.15V by default once you go above 5200MT/s.

And when manually tuning, Asus displays 1.0V and up as red, and 1.10V and up as purple.

For that.

When the colors have normally been blue? for near stock, yellow for normal OC, red for "are you sure?" and purple for "only for extreme cooling, will cause degradation".

Or were there white and then blue at low end, or just white? Don't remember, not that important in this context.

[wanton orchid]

you didn't mentioned straight-up broken documentation where advertised options are not recognized by actual libvirt or requires other undocumented attributes or even have different names

[twin dew]

They only publish documentation for the latest version, and don't have way to look it for older ones...

And most distros don't use the latest versions...

[wanton orchid]

I burned myself to it
I'm tired of self sucking lazy "community" projects (only work for commercial users under bizarre closed specific config fixes)

[wanton orchid]

you imply arch linux does not use latest ?

I mean
some option never actually been like written in the official doc

a lot of the doc stuff details is not actually supported

[twin dew]

Ok.

Some do state they are only for specific hypervisors.

[wanton orchid]

typical bullshit documentation used as marketing more than to document anything actually

[wanton orchid]

and yet does not work with it

or at least does not specify other requirements

[twin dew]

Cannot comment on that, but yes, in general the documentation is bad.

Expecting the reader to be expert in linux kernel insides etc.

And in the insides of the chosen hypervisor.

[wanton orchid]

and in the end in the inside of libvirt itself for option to interface with said hypervisors

[twin dew]

Yup.

[wanton orchid]

So a documentation that expects you do not need the documentation

very symptomatic

I call that a bait

[twin dew]

The "documentation" for Virt-Manager just didn't exist.
FAQ that didn't basically include anything and some screenshots.

At least on the project web site

[night girder]

We shit on AMD too no? When they drop the ball.

Now it's intel's turn 🤡

[willow pike]

we dunk on all corporations

except FICSIT

they do not waste

[twin dew]

And I had forgotten to unmount the VM btrfs from the host before starting the VM, and now the filesystem is toast.
Time to redo.

Because I had to bootstrap the stuff instead of normal install, as I don't know how to succesfully install with the Debian Installer while also modifying the target file system in command line.

To change the hashing algorighm used for the BTRFS from crc32.

Might be salvageable, not worth the effort.

Needed to do that mount and umount too many times, as I didn't know that all disks that needed to be enumerated by UEFI needed <boot order='#'/> entries, not just the one containing EFI partition.
GRUB2 didn't see the root disks because they didn't have those originally.

Oh, right, debootstrap didn't work with mounted subvolumes, it couldn't handle var directory already existing even when it was empty...

and the installer uses the same utility.

And Falcon 9 is flying again, with SpaceX doing 3 launches in 30 hours.

Sensor to oxygen line connection failed and started leaking, with possible assembly failure in attaching a support for the sensor.
And SpaceX deemed the specific sensor not to be even needed anymore and is removing it from new production.

• probably other stuff to make sure similar thing doesn't happen again.

[twin dew]

Stupid Linux... Being only able to hibernate with at least 50% of RAM free...

Even when there is fuckton of swap available, not intelligent enough to put some stuff into swap, then do the hibernation image creation and writing to another swap partition...

[safe trench]

just dunk on everyone

someone else quickly explodes into a dumpster fire after someone else's case is solved

[twin dew]

But what the fuck is going on... the according to top and like the programs are using 40MB of RAM, but at same time 775MB is being used according to free...

[dire igloo]

exactly this. if someone fucks up, call them out.
Only way to reduce corpo bullshit is to hold them accountable

[twin dew]

But seems I will need to give that one VM that needs to be able to hibernate 2GB of RAM.

[safe trench]

then we need govenor regulations

we cant just afford 3 different cpus

and 2 different mobos

just incase 1 blows up with problems

[night girder]

wtf!

you don't just shit on Linux you know ...

[willow pike]

only true linux users shit on linux

he clearly is

[night girder]

and they shit on non linux users for shitting on linux.

while shitting on linux themselves.

[willow pike]

an ouroborous of shitting

[night girder]

what a shit show.

[safe trench]

try not to say shit difficulty: impossible

[night girder]

💩

👆 I didn't say it.

[visual tree]

I'm surprised Google recorded streets 2 weeks ago and already uploaded them on Google Maps

Usually it takes them 1 year to update street view after recording streets

I should check if they recorded me walking to the office

[twin dew]

But seems I need to make that specific VM migratable (for libvirt save to disk) by not using passthrough CPU and VF NIC, but emulated CPU and emulated NIC...

[safe trench]

why do people want 500 fps 😭

[twin dew]

Bigger Number More Better!

[safe trench]

60 is just good if not better and if your that diehard 120, seriously why cap your fps to unlimited your just asking your gpu to draw extra power for no reason and 60 compare to 120 is nominal

[pure karma]

because they buy stupid refresh rate monitor they cant use

[safe trench]

THEY DO KNOW THEY NEED A 500HZ MONITOR TO PLAY AT 500 FPS?

[pure karma]

games can barely run 60-120fps yet alone 165,240,360 and so on

4K is only worth it for a tv

[safe trench]

its so silly watching fortnite kids set their fps to unlimited thinking they will win every game while they wonder why its so laggy not knowing they are sending their parts at full throthle right?

[twin dew]

Nah, just no vsync and get shit ton of tearing, but think they get less input lag with that.

[safe trench]

4k vs 2k is crazy tho

but 1440p gaming is 4k 1080p is regular

[pure karma]

honestly i feel like thats mostly just because everyone here is all about deals and amd lol

[safe trench]

1080p and 1440p isnt too big of a difference

[pure karma]

it is noticible on anything bigger than 24''

[safe trench]

about a whole different topic, why do people think oleds just suddenly stop burning in if they have it on a screensaver

[pure karma]

honestly i think oled burn in is something to just not bother worying about

my old monitor has burn in and its not even oled

[safe trench]

people do know burnins happen because of heat right?

[languid gulch]

yea, i noticed my current VA panel has a bit of burn in

[safe trench]

its not possible for them to just atach an whole water cooling system on it and let you still see

[languid gulch]

my mom got to experienced a pillowed S4

[dire igloo]

One factor, the brighter an LED has to shine, the more it wears out. The other factor is the type of content displayed.
If you reduce the amount of energy an LED uses by picking a dark screensaver, it will burn out slower.
If you then reduce static content (e.g. hide task bar) and pair it with constantly changing content, you'll also ensure that all pixels wear out somewhat evenly which also reduces the amount of apparent burn in.

Plus there's anti-burnin technology which uses stuff like pixel shift and whatnot.

Bottom line:
Pixels wear out because of heat.
Burn-in happens because of static content.

"burn in" in the sense of distinct patterns being visible on the screen which resulted from pixels wearing out

[twin dew]

OLEDs wear out based on usage too, not just heat generated by that.
All leds dim with use time, with uneven rates, OLEDs have more specific problems with that.
Current displays just track the individual OLED usage and try to compensate by changing the asked brightness.

Until they run out of headroom and the OLED cannot get bright enough anymore.

[twin dew]

With CRTs it was because the color phosphorus coating that the electron gun activated took damage from that. Basically solved problem near the end.
With LCDs somewhat similar, the color filter cells getting damaged. Same again, not really happening anymore.
With OLEDs it is still happening, just that the compensation methods have gotten very good and hide it until certain point.

[dire igloo]

what about Plasmas?

[twin dew]

Major issues, don't remember the exact mechanism.

[dire igloo]

I know they were super bad offenders for burnin

[twin dew]

Seems to be the same as with CRTs, but because of the tech used.

But because of the tech used (direct gas ionization) the better phosphorus created for CRT:s still weren't robust enough.

[twin dew]

Ah, seems I will need to switch from UEFI emulation to BIOS, as the UEFI is reserving 650MB of RAM for stuff...

[night girder]

500 FPS now means 60 FPS in the future (eventually).

[wanton orchid]

wdym ?

there is no point in using 60 since 2000s

the basic use sweetspot is 143
the pro/optimal use sweet low limit is ~157
ranging up to double that for fidelity (adapting frequency for content)
so ~320
everything above that can be fully emulated lossless

So absolutely no point in doing 500fps screening but to compensate for retarded graphic driver and game processing

but also no point in going back lower

the lowest point for old standard stuff fidelity is 120
but again can also go on 240 perfectly fine

140-320 should be the target maximum production range
with single to double ratio of adaptive single frame refresh rate

such monitor could screen absolutely anything transparently rate wise from an human standpoint

[languid gulch]

i remember the same concept from like 20 years ago from the "movies are 24fps & that's just fine" crowd 🤣 the "more is better & gooder for more money" win out

[twin dew]

When the various basic refresh rates like that were selected way back when as the lowest that was still somewhat tolerable.

[maiden coyote]

No, I'd like to understand how lcds burn in though.

[twin dew]

Part that color filter after the actual LCD element degrading, part that at least at some point the LCD crystals themselves could get "stuck" if the same value was kept on too long.

And with LED backlights, you can get uneven wear on those so the brightness differences between parts of the screen can get worse with time, as the LEDs lose brightness at different rates.

You have backlight, classically single CCL, now bunch of various types of LEDs depending on display quality, with white edge-lit LEDs as the cheap option.
Then you have diffusors etc, then the LCD panel that controls how much of the backlight is let through, then color filters that determine the colors of the pixels.

[dire igloo]

movies are a constant 24fps with heaps of natural motion blur

which is why they appear so smooth

[twin dew]

48 FPS movies are already a thing.

[dire igloo]

just providing context why "movies are 24fps and that's fine" doesn't apply to gaming

[maiden coyote]

Movies were probably 24fps because that's how it was during the film era

[twin dew]

And the point was that that is the lowest that was acceptable in 1920s.

When audio was added to movies in late 1920s.

[maiden coyote]

Higher fps means faster film speed and that was the lowest acceptable limit to create "smooth" motion without wasting film

[twin dew]

Yeah, but that the standard was set back then and is still used, except for few outliers that have gone with 48FPS.
Or direct-to-TV stuff that often uses the TV rates, which were determined by the AC frequency.

[maiden coyote]

Imax stuff had 60fps film, and it turned into something like 400lbs of film for an hour or something insane like that

[twin dew]

Also much bigger film space per frame

So even at same FPS it would have been much bulkier.

[maiden coyote]

That too, I think normal movies were 35mm film, and imax was shot on 75-120mm film

[twin dew]

Normal movie format is 35mm, IMAX film was 70mm, so little over 8 times the area per frame.

No, that was wrong? Seems imax spec is for one direction and the 35mm is for diagonal.

[maiden coyote]

Yeah. Normal film you'd buy for you camera, and then imax was "medium format" film commonly used by professionals

[twin dew]

Because for the normal 35mm to normal 70mm it is only 3.5 times the area.

But something weird with wikipedia on that, \o/

[maiden coyote]

Wiki says 65mm film?

And 70mm. I guess they had a few formats

[twin dew]

Which page?
https://en.wikipedia.org/wiki/IMAX says 70mm.. and then disagrees with the 70mm film page it links on the area difference...

[twin dew]

Ah, that part, you mistake the "conventional 65mm camera" part to Imax film size

When that points to the normal film camera using 35mm film

Ah, no... both... Wtf..

[maiden coyote]

Maybe not. The image they show for it is definitely not 35mm

[twin dew]

Edits at different time giving different values.

On same page.

[maiden coyote]

That ain't 35mm lol

[twin dew]

Ah, found the reason why the "Install Grub" part always fail.
I don't install base stuff thingie, so Grub hasn't been downloaded for installation XD

I mean "Standard utilities" part, as that just contains huge amounts of crap.

[maiden coyote]

https://en.m.wikipedia.org/wiki/70_mm_film

That's what they lugged up everest for the 1998 documentary

[languid gulch]

there's the crazy people who go on adventures, and then there's the batshit crazy people who go along with them to record it all

[tough owl]

how many CDs would I need to burn 1900 songs to cds?

[twin dew]

About 70 min per CD usually, originally 60min and 80 min CD-R blanks did exist.

If as actual CDs, and not just music files on data CD.

[tough owl]

its about 5.4Days worth of music

[willow pike]

just had war flashbacks to when i did that shit

[twin dew]

That 70 min is max you can do inside the actual spec.

[twin dew]

So little over 110 70 min disks, but in actuality more as you probably don't want partials.

[tough owl]

Apple music is weird. Itll let me playlists to CDs but not music albums

as if thats any different 🤣

[twin dew]

But with BIOS emulation instead of UEFI emulation, 1GB of RAM is nice and fine for that VM...

[twin dew]

70MB firmware reservation vs 650MB...

[safe trench]

unfair for the hate for intel

people are just haiting on intel way more than amd when they were basicly making motherboards a fire hazard

[stray badger]

AMD doesnt make motherboards. In addition, that issue was not as widespread and was very quickly fixed. Intels problem affects all their chips and has not been fixed yet

[safe trench]

august:

[twin dew]

That fire hazard part was Asus, not AMD.
It just needed broken CPU to happen, which was on AMD.
But AMD communicated clearly, fast, and made premature CPU replacement easy.

So AMD and Asus got flak for it, but AMD reacted the right way and took the hit.

Asus PR said they would do it right too, but at least in US that seems to not have happened completely.

[safe trench]

and for intel its only like the high end z series mobos

[twin dew]

No?
It is the CPU in all 65W or higher variants.

MB makers were undervolting the CPUs, which caused instability in another way.
Both being under the same label right now.

All 13th and 14th gen Intel CPUs have issues, both desktop and specific laptop CPU variants.
For desktop, officially from Intel right now it is "only" 65W and up that can get rapid permanent damage from the CPU requesting too high VCore and damaging stuff inside it (but some 700/900 T variants also affected).
For laptop, the highest end laptop chip series (HX?) also has issues, but it might be the same or different.

Intel has been trying to blame everyone else, but the actual cause is starting to come out finally.

And also tried to hide the issue for long time it seems.

And because Intel really-really-really doesn't want to respec all the chips in the 13th and 14th gen, replace all current ones and pay out compensation for lowered performance, they seem to be doing everything to downplay the issue, not take RMAs etc.

[gilded helm]

Equivocating the Zen 4 Asus issues with what's going on right now with Intel is ludicrous.

We can't even be sure we know the extent of what's going on with Intel right now.

[twin dew]

The base was similar, AMD thinking the safe VSOC was 1.4V when it wasn't, and "using"/giving that out, which then damaged and killed CPUs on all MB makers.
But AMD reacted quickly, took the blame, fixed the issue and implemented RMA program.

And that was in OC territory of the voltage range.
The Intel issue is in "normal operation" range.

For AMD, MB makers were just doing auto-overvolting to the danger zone if you set RAM frequency above the "stock max" of 5200MT/s.

[gilded helm]

GN's recent video collects all of Intel's Raptor Lake BS in one place. But holy crap, that video is poorly produced. Steve talks through the script way too quickly for most of the video, I know at least one other person who checked to see if they had increased the playback speed somehow.

[languid gulch]

it would have been more than an hour long if he hadn't talked fast 🤣

[dire igloo]

We don't even know when all of it will be over either

[twin dew]

And that GN video might have been sped up in Post for those segments, just for that reason.

[gilded helm]

Longform writing would have been a better medium for what GN had to say.

[languid gulch]

they probably also cut out a fair bit, he did say this was going to be a multi-part series

[twin dew]

With the native speakers not understanding that it might cause issues for some viewers.
I personally sometimes watch videos at 1.25x or 1.50x to make it go quicker.

[languid gulch]

he needs wendell to slow him down

[dire igloo]

Gamers Nexus are great journalists but awful presenters/educators and I wish it wasn't like that.
They could do so much good if they had LTT/HUB levels of digestibility

[languid gulch]

at least the fastest parts were the ranting, not quite as informational

[dire igloo]

But I doubt they'd ever change in that aspect (sadly)

[twin dew]

And they are doing written versions for most of their videos now, just with at least 1 week intentional delay.

gamersnexus.net/

[twin dew]

Because YouTube pays, their own site doesn't.

[night girder]

Tweakers tested out the new Intel Default Bios Settings to see how much performance loss there is:

[twin dew]

Which one(s), and on what MB?

Because depending on CPU model there were 2 or 3 different ones.

[night girder]

Gigabyte Z790 Aorus Master X

2023 = old bios settings, 2024 = new bios settings.

[twin dew]

Baseline, Performance, Extreme.
With Intel recommending the highest to be used if possible depending on MB VRM limits.

[night girder]

tweakers.net/reviews/12320/hoeveel-trager-worden-intel-processors-door-de-nieuwe-default-settings.html

[twin dew]

The site uses some scripting that blocks browser translates from working.

[night girder]

i9 14900K losses 8%-9% ... in performance.

[twin dew]

But was that 14900K run on the Performance, or Extreme?

[night girder]

Well, kinda what was expected.

Extreme

[twin dew]

Ah, and that also has the Baseline, it was the KS:s that didn't have.

Ok, that was the question, as the differences are pretty big between the presets.

[night girder]

i7's don't have extreme.

[twin dew]

900K:s have all three, lower down only two, 900KS only two.

[twin dew]

And this is the best case, as the next microcode that fixes another overvoltage issue isn't out yet, and might cause even more loss.

[night girder]

De Core i9's moeten worden ingesteld op het Extreme-profiel, met een IccMax (maximale stroomsterkte) van 400A. De K-modellen mogen continu 253W verbruiken, bij de KS-modellen is dat 320W.

[twin dew]

But the point was that if they had done those tests on the Baseline, and not Performance/Extreme (depending on CPU), then the results would have been unusable.

[night girder]

Tweakers is pretty solid with benchmarks.

They drop the ball sometimes, but then comments will point it out and they fix it.

The community is pretty active and keeps them in check, whish is why I like tweakers.

[twin dew]

But couldn't know before it was confirmed to have been set.
Which I couldn't verify myself even after you linked the stuff because of that autotranslate block their scripts cause.

And I don't understand why everyone keeps posting that same low-resolution version of that image...
When there is much higher resolution clear version available.

[night girder]

What?

The sheet?

[twin dew]

This one.

[night girder]

Because that's what Intel shared.

[twin dew]

There is the original high-resolution one on Intel forums, and then someone copied that and attached to web article, which got automatically downscaled, and everyone has been sharing that one around.

[night girder]

Makes sense, we are lazy/efficient creatures.

These temps though:

[twin dew]

community.intel.com/t5/Processors/June-2024-Guidance-regarding-Intel-Core-13th-and-14th-Gen-K-KF/m-p/1607807

[night girder]

99 - 99 - 101 😒

[twin dew]

1514x776 to 980x502 with weirdness from the bad downscaling.

[twin dew]

IIRC there is even version going around in smaller numbers that has gone through two of such cycles to almost unreadable.

around 600xsomething?

Possibly someone pasting in Discord and people copying the click-once version and not the full pasted version that needs second click to access.

[willow pike]

a few months from now it's gonna be fully deep fried with an ifunny logo on it

[wanton orchid]

https://cdn.discordapp.com/attachments/457105940844380160/1269991273448865862/2500.png?ex=66b2130a&is=66b0c18a&hm=d8958e1a98ee4483b15a051313fc304bdd564a6ace42d55b67829835308bf36e

right click > copy link

unless you are on discord mobile and that's completly fucked

[twin dew]

That just gives link to the image, not the image itself.
Or is some installed client specific thing?
On browser on the site there is small "Open in browser" text below the image at bottom left corner that opens the full image and not autoconverted sample.

[visual tree]

Not sure if my math is wrong but only €2,44 per month sound too low

And this calculation assumes a dehumidifier will have maximum power draw which is not going to happen

[wanton orchid]

what

what do you want it to consume over maximum power draw ?

[visual tree]

Maximum power draw is just a theoretical scenario, it won't be working at full power afaik

I am just checking if my calculation is right because I always though a dehumidifier consumes a lot of electricity even if it's a small unit with 12 liters per day

Guess appliances have gone really energy efficient in the last 10 years

I am planning to buy a 12 liter dehumidifier and use it during winter around 8 hours per night. Just wanted to check how it contributes to electricity bill and got surprised when I calculated only €2,44

Small price to pay to prevent mold growth on the walls if you ask me

[night girder]

That 0.041 is pretty low.

[visual tree]

We have higher and lower tarriff when it comes to electricity prices. During the day, the price is almost double and at night it's 0.041 (including VAT)

But since you mentioned it's low, I guess I am privileged lol