user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<derzahl>
ill take that as a 'no'
<neggles>
isn’t it still lacking support for 16k pages?
<derzahl>
i was able to get the 16k page size patch applied to the latest version of ungoogled-chromium, 100.0.4896.127...
<derzahl>
but I am stuck on the protoc failure now
<derzahl>
im gonna go back to ug-chromium 98.0.4758.80, as Ive successfully compiled it for arm64 before
<derzahl>
but not apple-m1
<derzahl>
i recall v99 being problematic for regular arm64 as well. so if i can get v98 built on apple for now, i will still be very happy
<bluetail[m]>
remember me claiming that on macOS the display looks different? It still applies. On the m1 mac Mini I can see the raster of the screen, although harder to see on a new monitor!
<bluetail[m]>
I think it has todo with the way the gpu works on m1
darkapex has quit [Ping timeout: 480 seconds]
<tpw_rules>
marcan: would you accept a pr to m1n1 to vendor a binary of the boot logo? having to have imagemagick available during build is inconvenient to me and you already keep pregenerated binaries of the fonts
user982492 has joined #asahi
PhilippvK has joined #asahi
phiologe has quit [Ping timeout: 480 seconds]
conradev has joined #asahi
Moprius has quit [Quit: Konversation terminated!]
ciggi has quit [Ping timeout: 480 seconds]
ciggi has joined #asahi
marvin24_ has joined #asahi
marvin24 has quit [Ping timeout: 480 seconds]
<marcan>
bluetail[m]: the "raster" of the screen? that's not a thing
<marcan>
macOS just does whole display scaling which results in a softer image, compared to Linux which usually does not. You'll be able to do the same once DCP is in (whether it actually makes sense is another story, since subsets of Linux GUI frameworks actually have better fractional DPI support which macOS does not)
user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
user982492 has joined #asahi
<bluetail[m]>
<marcan> "macOS just does whole display..." <- Its not a OS thing. I did make the test. On Linux its the same with the m1 chip. Basically, on TN screens, I am capable of seeing quadratic rasters / lines on the screen which others said is 'Moire'.
<bluetail[m]>
It is so far ONLY seen on Apple Devices. macbook pro 2016, imac 2019 (intel), mac mini (m1)
<bluetail[m]>
I don't know what the heck is different - but something certainly is.
<bluetail[m]>
Heres the catch:
<bluetail[m]>
If you get to use a screen with more resolution than I do (1920x1080) and with bigger color range, it does not happen
<bluetail[m]>
And it absolutely doesnt matter if I'm on Linux or Windows
<bluetail[m]>
It is more difficult to see on the higher end TN Panel of mine
<bluetail[m]>
but I can still notice it
<bluetail[m]>
Just to be sure
<bluetail[m]>
* Just to be clear:
<bluetail[m]>
This behavior does not happen on any non-apple System I am aware of, and I have built a lot of rigs since the 2000s
<bluetail[m]>
I could proof it with a proper camera
<bluetail[m]>
* proper camera - that the overall image is different
<marcan>
Again, that's not a thing. The OS doesn't know nor care what display you have. The display just gets bits in via HDMI. If your display is acting weird with some OSes and not others then it might have buggy firmware or some other kind of behavior which is triggered by some subtle difference in the resolution/display mode.
<bluetail[m]>
Yea
<bluetail[m]>
thats what I say
<bluetail[m]>
It's hardware
<bluetail[m]>
I could reproduce this behavior on quite a range of apple devices
<bluetail[m]>
and it annoys me
<marcan>
sounds like a monitor issue then
<bluetail[m]>
nope
<bluetail[m]>
this is a firmware / GPU issue on apple devices
<marcan>
again, the GPU/firmware just ends up giving the monitor bits
<marcan>
the monitor decides what to do with those bits
<marcan>
the OS decides what those bits are and how they are processed
<bluetail[m]>
right. I can see it even on my CRT (through adapters)
<bluetail[m]>
It's just less noticeable on the highest end TN panel
<bluetail[m]>
But its certainly there
<marcan>
well, it certainly isn't what you think it is because what you're saying doesn't make any sense :-)
<bluetail[m]>
I cannot spot it on a raspberry pi
<bluetail[m]>
I cannot spot it on any x86 system I ever had
<bluetail[m]>
I can only spot it on the apple devices
<marcan>
and you say it happens with Linux on x86 macs?
<bluetail[m]>
it does
<marcan>
and not with Linux on other x86 boxes?
<bluetail[m]>
exactly
<marcan>
sorry, that's impossible. it's the same hardware.
<marcan>
you're either imagining things or you have something weird going on.
<bluetail[m]>
wait
<bluetail[m]>
I compare lenovo t490
<bluetail[m]>
against a mac mini m1 that runs asahi
<bluetail[m]>
and
<bluetail[m]>
imac 2019 that runs linux
<bluetail[m]>
and macbook pro 2016 that runs linux
<bluetail[m]>
and I can see it on every of those
<bluetail[m]>
I cant see it on my lenovo t490
<bluetail[m]>
i cant see it on my gtx 1660 super
<bluetail[m]>
I cant see it on a 980 ti
<bluetail[m]>
I cant see it on a gtx 460
<marcan>
using the same display connected to them?
<bluetail[m]>
not even on my spare 9800gtx
<bluetail[m]>
Yes
<marcan>
yeah, sorry, that *really* isn't a thing.
<bluetail[m]>
I have actually two pairs of the same new monitor
<bluetail[m]>
plus a third one
<marcan>
until proven otherwise the most likely explanation is you're imagining things
<bluetail[m]>
have two of XL2546K
<bluetail[m]>
1 of LG24GM79G-B
<bluetail[m]>
marcan: again. If I had a proper camera I could show the differences in the image
<bluetail[m]>
They are subtle
<bluetail[m]>
but they are there
<marcan>
x86 Mac hardware is the same as x86 non-Mac hardware when it comes to external display ouput.
<marcan>
there is no "Apple firmware"
<marcan>
there are no "Apple GPUs"
<marcan>
it's the same thing
<bluetail[m]>
Yes it makes no sense
<bluetail[m]>
I couldnt see it on my hackintosh
<bluetail[m]>
its absolutely bonkers
millenialhacker has joined #asahi
<bluetail[m]>
Perhaps somebody should do a blind test with me :D
<bluetail[m]>
* do a double-blind test
<marcan>
perhaps :p
<bluetail[m]>
But really
<bluetail[m]>
I might be the exception
<bluetail[m]>
Why? I have particularly sensitive vision
<marcan>
yeah, that's what all the audiophiles say...
<marcan>
"my hearing is special"
<bluetail[m]>
well I cant attest audiophile hardware to give me more joy than a dt 990 pro
<bluetail[m]>
except of one model
<bluetail[m]>
and that one model is uncomfortable
<bluetail[m]>
What I hear might not be what you hear.
<bluetail[m]>
But for that we measure those headphones
<marcan>
I'm talking about the people claiming they can hear above 20kHz and stuff like that
<marcan>
or that WAV sounds better than FLAC
<bluetail[m]>
And we could also measure if the video output is completely identical or not and if the emitted output of the monitor is identical
<marcan>
yup
<bluetail[m]>
marcan: That depends if a measly mp3 was converted to either of them.
<bluetail[m]>
But FLAC can contain a wav - its the same
<bluetail[m]>
well
<bluetail[m]>
FLAC takes a little cpu perf but its neglibible.. cant spell that word
<bluetail[m]>
marcan: On that note - in the Safari browser some colors are rendered differently. But thats a software bug. No sources, only measurements taken internally which I cant provide unfortunately.
<bluetail[m]>
You can actually compare it in screenshots
<marcan>
feel free to do a side by side comparison if you get a decent camera. it's entirely possible that you're seeing subtle differences in a given monitor with different settings in different OSes, but a blanket "Apple devices behave differently" makes no sense.
<bluetail[m]>
It doesnt make sense
<marcan>
color management is another story
<bluetail[m]>
I blame the m1 gpu in that case ... for the other devices I dont know
<bluetail[m]>
It's not explainable
<marcan>
that can be pretty different between browsers, especially if your display is calibrated
<marcan>
again, the gpu has nothing to do with any of this
<marcan>
the gpu just does what you tell it to
<marcan>
there's nothing magical about it
<marcan>
there's nothing special about it
<marcan>
safari doing whatever color management has nothing to do with the GPU
<bluetail[m]>
Okay - but then I dont really know what causes this
<marcan>
it's just whatever the code does
<bluetail[m]>
It's very subtle and the upgrade of the monitor did improve the problem
<marcan>
I think you're confounding different factors and confusing yourself here
<bluetail[m]>
I think I'm unable to explain what I see
<marcan>
software doing color management differently is normal (well, it shouldn't be, but it is because color management is hard)
millenialhacker has quit [Ping timeout: 480 seconds]
<bluetail[m]>
its just over the whole screen using the m1 mac, just a lot less on the newer monitor
<bluetail[m]>
on the old one its very visible
<bluetail[m]>
its not in that intensity though
<bluetail[m]>
its much finer
<marcan>
TN displays do that, especially if the COM voltage is miscalibrated
<marcan>
and it will potentially look different at different refresh rates depending on processing
<marcan>
it has nothing to do with the host
<bluetail[m]>
I get that. but what I dont get is the different appearance on two different hosts
<marcan>
there is also dithering involved with TN because TN displays are almost never 8 bits
<marcan>
two different hosts running the same OS with the same display environment and the same theme and the same modelines and the same gamma settings and the same HDMI color range and the same... etc?
<marcan>
because that kind of stuff shows up more with some colors than others, for example
<marcan>
but also, the effect is subtle enough it's entirely possible you're convincing yourself that it's there when you think it should be, and that it isn't when you think it shouldn't
<richardburleigh>
with respect, the best way forward would be if you can quantify what you're referring to. There are a heap of algos in computer vision to assess visual fidelity. I understand you are talking about specific monitor classes, but as another option you could always capture the raw HDMI output of a reference image and compare.
<marcan>
I have an NeTV2 so if I need to pull raw bits off of an HDMI link I *can* actually do that... but I'm pretty sure that won't be necessary to solve this mystery
<marcan>
color management differences between monitors and OSes are expected
<bluetail[m]>
Yes sure but if I boot the Arch KDE environment it should be comparable to my Debian (KDE) environment... right?
<bluetail[m]>
When both use the same color profile...
<bluetail[m]>
* to my x86 Debian (KDE)
<bluetail[m]>
If I boot into Windows on my Lenovo T490 I also cant see it on the TN's
skipwich has quit [Quit: DISCONNECT]
<richardburleigh>
Sorry marcan, my message was for bluetail. You can pull the raw HDMI output of a reference image on debian M1, vs debian (whatever) and compare
<bluetail[m]>
richardburleigh: That would be great, right? What device do I require for that?
<bluetail[m]>
Just to note: Visual evidence failed using a ipad 2019 12.9" pro camera
<bluetail[m]>
Before I forget it
<bluetail[m]>
marcan: if I go from the x86 Lenovo via VNC to my mac mini the problem is not existent
<bluetail[m]>
* mac mini m1 the problem
<bluetail[m]>
It has a mode where it allows for assumedly 'accurate' way of capturing a screenshot
<bluetail[m]>
I do use TurboVNC
<bluetail[m]>
* use TurboVNC with ... the settings that take the most bandwidth for visual fidelity
<bluetail[m]>
> <@_oftc_richardburleigh:matrix.org> Sorry marcan, my message was for bluetail. You can pull the raw HDMI output of a reference image on debian M1, vs debian (whatever) and compare
<bluetail[m]>
* That would be great, right? netV2 seems to be capable of that...
<richardburleigh>
since you're talking about visual artefacts, you shouldn't need anything fancy
<bluetail[m]>
its just mind boggling
<bluetail[m]>
First when I tried Asahi Linux (ARCH) it was in the hope of getting rid of this behavior
<bluetail[m]>
richardburleigh: Arent there like macro lenses you can temporarily attach on the ipad pro lens?
<richardburleigh>
honestly I didn't even know external displays were working on Asahi lol
<bluetail[m]>
well they are limited to 60hz
<bluetail[m]>
you cant use xrandr ...
<bluetail[m]>
But well
<richardburleigh>
regarding camera lenses, it depends what you are assessing.. there is so much digital processing done on the image, including automatic focus fumbles, that it is hard to compare 1:1
<bluetail[m]>
I would be just assessing the lower left corner of the matrix window
<bluetail[m]>
There is like a relative dark gray against a grayish blue tone
<bluetail[m]>
They appear as a surface (clean) on the x86 device
<bluetail[m]>
They appear with a grid that subdivides them on the m1 device
<bluetail[m]>
But it is MUCH less visible on the newer Monitor
<bluetail[m]>
I can spot it when I follow from for example left to right, as when I'm reading
<bluetail[m]>
* spot it on the m1 when I
<marcan>
probably just a different color
<bluetail[m]>
When I do that on the M1, I can see relatively large 'sqares' approximately the size 1 third of my pinky
<bluetail[m]>
I cant see them through VNC ...
<bluetail[m]>
* I cant see them through VNC ...
<bluetail[m]>
Yea
<bluetail[m]>
that would make the most sense
<marcan>
the M1 *is* going to do different output processing to a random x86 device, especially since the display controller is driven by DCP and is a different design anyway, so if you wanted to do a controlled test where we output the same exact HDMI bit patterns that would be a project on its own
<bluetail[m]>
But I just dont get how even though I use the same color profile on both devices they differ
<bluetail[m]>
They should be same right?
<marcan>
no, as I said
<marcan>
it's different hardware
<bluetail[m]>
ic
<marcan>
it should be possible to get it to output the same mode and colors, but you can't expect it to do by default
<marcan>
which is why I was focusing on the x86 comparison, since *that* one doesn't make any sense
<bluetail[m]>
You should take the x86 comparision with a grain of salt as this information is older than 1 month
<bluetail[m]>
I have neurological problems which falsify information that are older
<bluetail[m]>
But this - right now is something I can replicate right now
<bluetail[m]>
* right now on a M1
<marcan>
different OSes, different hardware, different drivers will end up doing different color processing and HDMI encoding and coupled with cheap TN monitors (it's kind of hard to take color experiments seriously on a TN monitor...), yes, that can have different results
<bluetail[m]>
Perhaps it is only on the M1 and my brain has mixed it up with being the same on x86 macs
<bluetail[m]>
The reason I use expensive TN's in the first place is because I tried a couple of IPS monitors whose backlight I was uncomfortable with
<marcan>
yeah, but TN monitors can't even maintain a consistent gamma across the whole display due to viewing angles, so... yeah
<bluetail[m]>
It's basically like this:
<bluetail[m]>
Lower brightness so that its comfortable -> Now gray appears almost as black.
<bluetail[m]>
Add more brightness so that grays appear as gray -> Eyes are watering
<bluetail[m]>
I havent tried VA (yet)
<bluetail[m]>
All I knew is that I required to use a monitor that would do its job for me
<marcan>
that sounds like a monitor issue again, bad gamma/wrong brightness control?
<marcan>
decent IPS monitors properly calibrated should have excellent color and proper gamma
<marcan>
TN... all bets are off there really
<bluetail[m]>
marcan: Well, given that I had a couple of models that were calibrated from the get-go and had calibration hardware on-board and it was the same, I doubt that. (Expensive EIZO) precision Monitors... and a couple of 4k 144hz ones..
<bluetail[m]>
The EIZO's were better than the gaming ones though
<bluetail[m]>
in terms of how easy they were on my eye
<marcan>
monitors advertised as high frame rate are usually meant for gamers and good/accurate color is way down their priority list
<bluetail[m]>
You said earlier that there is a technology to avoid color banding, right?
<bluetail[m]>
What I did see is that the IPS screens provided me with no such banding effect
<bluetail[m]>
It is visible on the TN's I have
<bluetail[m]>
When I was setting up the IPS screens to fight against my 'problem'
<bluetail[m]>
I ended up with visible banding, it was slightly better
<bluetail[m]>
But colors appeared as 'washed out' as in TN
<bluetail[m]>
Perhaps thats a 'optical-cognitive' problem ontop
<marcan>
TNs have lower bit depth so they will have color banding, especially if the dithering is not properly calibrated, and depending on the angle
<marcan>
that's why IPS is better
<bluetail[m]>
I see
<bluetail[m]>
Perhaps thats the thing
<bluetail[m]>
I might be unable to process 8 bit screens with dithering on...
<bluetail[m]>
But I'm uncertain
<bluetail[m]>
I havent tried the remaining technologies
<bluetail[m]>
I can use CRT, TN ... But I have not succeeded with 'IPS' yet. People say its the backlighting thats different somehow
<bluetail[m]>
I havent tried a VA screen yet
<bluetail[m]>
But when i say I did change the settings to get the colour banding back, I did that because I could look at the screen with more ease
<marcan>
IPS does have the "ips glow" effect which some people dislike, but if the backlight is set at an appropriate level it's rarely an issue
<bluetail[m]>
Well...
<bluetail[m]>
10% backlight on all IPS except the EIZO I had
<bluetail[m]>
grays were almost black
<bluetail[m]>
anything more of brightness and my eyes started to water
<bluetail[m]>
I use full brightness on the xl2546k
<marcan>
that sounds like you were changing a brightness control that isn't actually changing the backlight
<marcan>
like additive/subtractive brightness/offset control
<bluetail[m]>
I see
<marcan>
you should use test patterns to figure out what your monitor controls are actually doing
<marcan>
a lot of "brightness" controls don't actually control the backlight properly
<bluetail[m]>
Now for the next 6 years I will probably use TN but afterwards, I might try VA or whatelse then is as matte as possible and as easy on the eyes as possible
<marcan>
I have a 2015 5K iMac and the brightness control does proper backlight control on linux and the lowest setting is very dark (but still readable on medium lighting)
bisko has joined #asahi
<bluetail[m]>
On the LG27GP950-B this was more like a linear shift from dark blue to blue and so on
<deagle50[m]>
Installed asahi yesterday on my MBA
<deagle50[m]>
Really impressive, thank you all
<deagle50[m]>
When brightness control lands it might be a daily driver
<bluetail[m]>
marcan: I heard the older models were easier on the eye for many people
<bluetail[m]>
The apple forum is actually well populated with people who cannot look in current 'RETINA' screens for different reasons
<bluetail[m]>
Whenever people complain they are sent to the eye-doctor
<bluetail[m]>
I was at the eye-doctor (optometrist) and everything was good
<bluetail[m]>
* heard the some of the older models
<bluetail[m]>
* different reasons. But I just added a reddit one.
greguu has quit [Quit: WeeChat 3.0]
<bluetail[m]>
Back in the days I was talking to some companies who were able todo anti glare to (some) mac screens
<bluetail[m]>
They said that in the past they could get the imac glass off the screen and then somehow make it matte. But that would be no longer the case for newer ones.
<bluetail[m]>
Perhaps with those high-end precision screens, could it be possible that it is simply an sort of sensory overload?
<bluetail[m]>
(similar as some people cannot stand some STAX headphones and complaining they are 'too detailed and thus exhausting')
<marcan>
I'm certainly not going to tell you what display does or doesn't give you a headache, brains are weird
<marcan>
that person says using the output into a known good display also does it, but output processing will also be different on a Retina Mac and a non-Retina Mac because the way the OS does scaling is different when the main display is Retina and enabled (not sure about disabling it), in macOS
<bluetail[m]>
Yea right, cause of the sub-pixel thing and all that 'funny' scaling stuff.
<bluetail[m]>
Actually I have a colleague who recommended a 5k screen to somebody visually impaired because of the smoother font - so that the person can zoom onto it using the magnifier
<bluetail[m]>
In the past that was done with 'reading cameras' that cost like 20 grand
Ry_Darcy has joined #asahi
<bluetail[m]>
Also - thanks for the nice discussion, I appreciate it. I kind of didnt sleep this night. Its already 7 am, I should nap a bit :)
darkapex has joined #asahi
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
bisko has joined #asahi
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
bisko has joined #asahi
user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
user982492 has joined #asahi
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
ptudor_ has joined #asahi
ciggi_ has joined #asahi
ciggi has quit [Ping timeout: 480 seconds]
ptudor has quit [Ping timeout: 480 seconds]
<Glanzmann>
derzahl: You can apply the 4k page patch tg.st/u/0001-4k-iommu-patch-2022-03-11.patch or use the prebuild chromium: tg.st/u/chromium-532e10195c-dyn-2.tar.gz
user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
giskard has joined #asahi
<derzahl>
Glanzmann: oh cool thanks. Doesnt it need the 16k page patch? I'm actually compiling my own right now with ungoogled patches now, i think I worked out the kinks this time
<derzahl>
is there a PKGBUILD for yours?
<nicolas17>
derzahl: the patch is for the linux kernel to make it use 4k pages
<nicolas17>
then normal chromium works
millenialhacker has joined #asahi
millenialhacker has quit [Ping timeout: 480 seconds]
<Glanzmann>
derzahl: So the chromium I sent you is from Chainfire, IIRC.
<Glanzmann>
I did not build it myself. chainfire did and he reported the issue upstream, I just tested it on 4k pages, but it also works on 16k pages (that is what chainfire tested).
<Glanzmann>
The issue is reportedly fixed in upstream chromium and will land in version v102 IIRC.
the_lanetly_052 has joined #asahi
user982492 has joined #asahi
nicolas17 has quit [Quit: Konversation terminated!]
<derzahl>
ohhh gotcha
<derzahl>
right, im attempting to backport the 16k page size chromium patch
<marcan>
that's corporate speak for "it's in your head, guys", which as I said is the mostly likely explanation for these issues, at least the ones that don't have any sensible software explanation
<marcan>
occam's razor says that if a small minority of people are experiencing health effects from something that defies all logical and scientific explanation, chances are it's in their head. same story with people saying 5G gives them headaches, etc.
<marcan>
as I said though I'm not saying all reports are such; there *are* differences in visual output between different OS and hardware configurations, e.g. the Retina business
<bluetail[m]>
marcan: I kind of agree and I kind of disagree
<marcan>
but all that is easy to quantify, you just need a capture card
<bluetail[m]>
I thank you for taking the time
<bluetail[m]>
I'm certainly not one of those 5G guys
<marcan>
so it either boils down to something like "this OS config makes the text blurrier/whatever due to the scaling and that gives me headaches" (which is entirely possible), or the pixels are the same and then (modulo HDMI InfoFrame related business, you *can* go deeper in the investigation but it should be fairly easy to test for any obvious changes with a camera) you're left with it being a purely ...
<mps>
not sure about all this but I can't use 3G phone too much because I clearly know that I have some pain on the head side on which I use phone
<marcan>
mps: I was talking about people saying 5G in their environment (and, like, not at <3m from an antenna) causes health effects
<marcan>
a phone held to your head has many subtle ways of causing minor physiological effects which some people might be sensitive to
<mps>
marcan: yes, I wanted to say we all are different
<marcan>
from the physical pressure, to the heat produced by the phone, to the subtle but nonzero dielectric heating effect of the radio waves themselves
<bluetail[m]>
It's easy to dismiss... Did we not agree on the hardware causes difference marcan ?
<marcan>
right, but there's "I'm sensitive to little things most people aren't" (that's different), and "I'm sensitive to something that defies scientific explanation" (usually not a real effect, and easily shown with a double blind test)
<bluetail[m]>
Well
<bluetail[m]>
I am certain I can catch it on photo because I can see it very clearly
<mps>
also, looking too much on not good quality display causes some strange 'light' in my eyes
<bluetail[m]>
I posted a pic in libera hardware earlier and people said that was just moire/aliasing effect
<marcan>
looking at computer screens in general *is* stressful for various reasons and a lot of people are sensitive to that
<marcan>
same with backlight flicker, etc
<bluetail[m]>
but my brain is wired in a way that I can see them and my brain isnt filtering it out
<marcan>
I'm certainly not going to dismiss people who say certain displays/software/whatever cause them eyestrain or headaches; indeed we are all different
<bluetail[m]>
That might be a opti-cognitive thing, but certainly not psychological
<marcan>
I'm just saying that once we get down to the level of A/B testing things which at a digital level should produce identical output, we run out of sensible explanations other than it being a psychological effect
<bluetail[m]>
marcan: I am certain I wouldnt fail that
<bluetail[m]>
At least not with my old lg 24gm79g-b
<bluetail[m]>
it is so clear and distinct that I can tell immediately
<marcan>
I'm talking about the x86 story which you yourself said could've been a bad memory :)
<bluetail[m]>
marcan: And I'm talking about m1
<mps>
long ago I read somewhere that australian aborigines see every image separately on movies which uses old tapes with 24 images in second
<marcan>
as I said M1 vs not has enough scope for different output that I am not at all dismissing that one
<bluetail[m]>
I just say I cant see it on my laptop for whatever reason
<bluetail[m]>
marcan: then we both agree on that one. :)
<bluetail[m]>
Yea the x86 one must been faulty memories
<bluetail[m]>
I tell you one thing
<bluetail[m]>
I 'lost' 10 years of my life due to daily seizures during REM phase (rapid eye movement) -> learn phase
<bluetail[m]>
Most things were just wiped but some did stick nonetheless
<bluetail[m]>
Its still a problem with old information getting fuzzy and need to be repeated a ton till they are ingrained
<marcan>
the problem with all these stories is most people experiencing these problems don't know how to "debug" these issues and arrive at controlled tests
<marcan>
so you end up with a mix of people who *are* sensitive to a difference, but who lead themselves into red herrings and wildly incorrect theories (this is the same for laypeople trying to diagnose the cause of weird bugs, but it's especially bad when health effects are involved)
<marcan>
and people who outright have broken hardware/bad screens/etc
<marcan>
and people for whom it's really all in their head
<marcan>
and then weird theories end up getting collectively reinforced
<marcan>
because the third group will easily lock onto such a theory and provide "evidence" for it
<marcan>
the first group will just confuse themselves with bad tests/experiments
<marcan>
and the second group, well, who knows, when their hardware is bad
<marcan>
and it's really frustrating because you just can't pinpoint any actual problem (even when there might be one) but *everyone* is convinced they have the answer or highly plausible theories
<marcan>
and you can't just say it's not a real effect because it might be, for some
<marcan>
but you also can't just test one thing and figure out if it's a problem or not because everyone has arrived at the story from a different angle and through faulty conclusions
<marcan>
all you can do is pick one person at random and try to figure out what *they* are experiencing, and they need to let you throw away all their preconceptions first because chances are the backstory is a mess
<marcan>
it's a lot easier when there are easy explanations, e.g. "this screen gives me headaches" (could be anything from color settings to panel flicker to backlight flicker, etc)
<marcan>
but when a group of disparate people somehow end up arriving at the conclusion that something that doesn't make sense is the cause of their collective woes, it's really hard
<marcan>
because they're probably wrong, even when there *is* a cause of their woes, at least a subset
<marcan>
but good luck convincing them of that
<marcan>
that's why intel had a bunch of things tested and concluded there was no problem and closed the thread
<marcan>
maybe some of the people reporting there *did* have some other problem caused by something else, but what are they going to do?
<marcan>
I see this kind of situation all the time with non health related effects too, e.g. some community is running into a problem and ends up developing their own alternate reality lore for what causes it and how to avoid it, then you finally look at the code/problem/etc and it's real but they were all wrong with their theories (but many of them totally convinced they had it)
<marcan>
conversely, dismissing collective reports of problems like this for "weird stuff" without some kind of investigation is also bad (*cough* OBS audio death bug *cough*)
<marcan>
and I mean, we've all run into WTF theories
<mps>
yes, making conclusion about anything by personal experience leads to weirdness
<marcan>
like the other day I ran into a weird "these devices connected in this way cause everything to fail" situation and it turns out one of the causes was an AC adapter connected to a little insignificant part of the setup, but it wasn't just that, there are still other sporadic issues, and it's very hard to debug
<marcan>
I've come up with invalid theories myself many, many times
<marcan>
and electronics was a mistake, etc :)
<bluetail[m]>
All I know - there is something different and it bothers me. I don't know how to fix it. Perhaps my colorimeter is just bad or something else in the chain is causing this. Perhaps I have to activate some setting in my monitor. I might not even know what causes this.
<marcan>
but you just need to either just keep digging until you find the actual problem, or just accept that sometimes hardware is cursed
<bluetail[m]>
marcan: I am about to buy a Arduino to get into Electrical Engineering because I'm such a noob at it... I know the inner workings are covered this way maybe, but I need kind of an introduction
<bluetail[m]>
marcan: I remember blaming my Nintendo Wii was defective when every night I game rips over WiFi somebody turned their microwave on and then interrupted the WiFi :D
<bluetail[m]>
That was funny
<bluetail[m]>
until I found out the actual cause...
<bluetail[m]>
It was just a defective microwave shielding instead
<bluetail[m]>
And the game dumper that time didnt support to restore the upload
<mps>
this reminds about story with so wise man as Socrates. he concluded that females are 'lower' beings than males because females have only 28 teeth while males have 32
<marcan>
I once had a wifi router that, with the antennas disconnected, when plugged into the *mains* in a corner of the room, killed the DSL connection in the house. with no ethernet/data connection whatsoever
<bluetail[m]>
WTF :D
<mps>
Socrates counted teeth only on his wife Xantipa and no other females :)
<marcan>
we didn't have SDRs back then like we do today and I didn't have any test equipment to diagnose it, so I just chucked it up to insane radio gremlins
<mps>
so yes, making conclusions with a lot of data, experience and deep thinking usually leads to wrong conclusions
<marcan>
yup
<mps>
s/with/without/
<marcan>
and even if things are actually "in someone's head" that doesn't make it not real, it's just a really unfortunate situation
<mps>
yes
<marcan>
so as I said I certainly won't dismiss reports of health effects but I may dismiss their bad conclusions as to the cause :-)
<mps>
as I read (and real experience) health effect could be caused by (well) strange thinking
<marcan>
(I will dismiss audiophiles outright though, that particular collective delusion has gone on for too long and if you refuse to do controlled experiments you're forgoing your credibility, which they ~always do)
<marcan>
(but that's one most of us are familiar with)
<mps>
there are so called psyhosomatic illnesses
<marcan>
absolutely
<marcan>
"it's all in your head" just means the trigger for the issue is mapped through relatively high level thought in your brain, not that the end result isn't a real physical effect
<marcan>
human beings are weird and brains are weirder even
<mps>
yes :)
<mps>
according to some reports healthier are those who are cool and calm and don't care much about health than those 'health fanatics'
<richardburleigh>
even base anxiety can cause things like headaches through subconscious association.. for example, people who have spent a lot of time working in a stressful environment may have experienced mild panic attacks when looking at emails. Even after they change jobs and the stress is gone, the trigger can remain and give all the same physiological symptoms.
<marcan>
mps: "don't care much about health" in the sense of not worrying about health, yes
<marcan>
(not in the sense of being unhealthy)
<marcan>
richardburleigh: certainly
<mps>
marcan: yes, I meant this
<marcan>
I guarantee there is some subset of people saying something like "Apple products give me headaches" where the root cause is some repressed Apple-associated trauma in their past
<marcan>
(substitute Apple products for literally anything else)
<mps>
my english is not so good to clearly express what I have in head
<marcan>
it's actually scary just how fallible our brains can be, and being aware of that is important
<marcan>
like I know bluetail[m] mentioned faulty memories but we *all* suffer from that, especially over time, to some extent
<marcan>
it's just how brains work
<mps>
marcan: fully agree
<mps>
brain could make delusions in us very easily
<marcan>
my brain throws away details all the time, like whether I told something to a specific person or not, or whether a video I watched came from someone else sending it to me or my feed
<marcan>
ends in funny things like me talking about the video again with the person who sent it to me :p
<marcan>
(ADHD doesn't help with that either)
<richardburleigh>
honestly, about 10 years ago I quit a terrible job expecting to find another one quickly.. I didn't.. and even now with a decent income, I still feel background panic when I receive the energy bill that I wouldn't even notice if it was deducted automatically
<marcan>
that's rough
<richardburleigh>
nah, best thing that ever happened.. taught me how to save and always have an emergency fund
<marcan>
heh :)
<marcan>
I used to think it was weird how so many people didn't have an emergency fund... but over the past few years I've come to realize just how unbelievably screwed up the labor market is in most places, especially for millenials+
<marcan>
I was lucky enough to come from a middle-class family and my parents were always good with money (but not penny-pinching) so I figured, can't everyone just do the same thing?
<marcan>
turns out, no...
<j`ey>
marcan: now your emergency fund is stored in the value of all your apple silicon machines :P
<marcan>
lol :)
<marcan>
you might not be wrong after my tax bill this year :p
<marcan>
turns out when most of your income is patreon without withholding...
Ry_Darcy has quit [Remote host closed the connection]
<richardburleigh>
I read a book a while ago about how many CEOs end up homeless because they never learnt the fundamentals of managing their own money. They matched any income increase with proportionate debt, and the moment the company fails or they lose their job, they are instantly broke and in credit default.
<marcan>
yeah, there's definitely a subset of people who just cannot manage money, and the super-rich are definitely often in that class
<marcan>
it's also a vicious cycle for some poor people because they're so frustrated with their normal that any excess income instantly goes into splurging to feel better about life
<marcan>
and you can't blame them for that either
<marcan>
also the financial sector in e.g. the US is utterly predatory
<marcan>
credit cards etc
<sven>
I never realized that people actually had debt on the credit cards for months before I spent time in the USA
<marcan>
yeah...
<richardburleigh>
yeah.. I usually never use the word 'evil', but payday lenders are absolutely deserving of the title
<marcan>
yup.
<marcan>
I only use my credit card as a "debit card but people expect you to have a credit card", and the limit is often annoying (*cough* Mac Studio *cough*)
<sven>
everyone I know in .de just uses them and pays off the entire balance every month while the banks in the USA wanted me to only pay off like 10% or so and just keep accumulating debt :/
<marcan>
come on Rakuten, why won't you let me throw money at you so I can keep using the damn thing this month? :p
<marcan>
apparently paying Apple via bank transfer is sometimes possible but you need to actually talk to someone? it's weird
<marcan>
actually I said "the limit is often annoying" but the limit was never a problem until, well, I started buying Macs.
<sven>
:D
<sven>
i don’t think I ever ran into the limits of mine, but trying to buy a few macs might just do that ;)
eroux has joined #asahi
<marcan>
my limit right now is just under €11k, and I'm sure you can see how, well, an M1 Pro + M1 Max + misc expenses that month could cause a problem...
<marcan>
the Mac Studio did go through since it was just one thing that product cycle, but barely... payment failed the first time for some reason, but apparently eventually succeeded? I was worried for a while.
<richardburleigh>
just wait until the Mac Pro replacement comes out :D
<marcan>
yeeeeeah...
<marcan>
I think I'm going to have to call a rep and beg for bank transfer payment for that one...
<j`ey>
the most expensive mac pro currently is 50k$.. I wonder what the AS one will top out at
<marcan>
hopefully less but... who knows...
<marcan>
I will say the machines aren't going to waste besides testing though; I have a couple close local friends in a less financially lucky situation that I share with as long as I get to test whatever Asahi stuff I want when I need to
<marcan>
feels wrong to just literally have a pile at home
<richardburleigh>
much respect
<mps>
I have few still very usable x86 machines keeping dust in garage, but have no idea whom to give them
<mps>
no one wants machines with linux
<marcan>
heh
<jannau>
half of that $50k is the 1.5TB ram upgrade Apple is selling with a markup over 2 though
<richardburleigh>
I recently moved to South Korea from Australia and haven't yet brought over my system with 2x 2080ti GPUs. No friends were interested in holding it, so it's just sitting there unused. I genuinely feel guilty about it with the current GPU shortages.
<j`ey>
I gave away my old desktop to a random person from the internet, that felt kinda nice and charitable
<marcan>
depends on the age, but honestly one thing I learned is that hoarding old hardware is, to a point, kind of useless. like, the machines might be "usable" for some definition but often making a concerted effort to get the machine to someone who would actually appreciate it is, in itself, more of a time/money cost than the value of the machine itself
<j`ey>
jannau: i assume theyre just normal slots you can put ram in then
<marcan>
giving them away is always a good idea but if nobody local will take them... you might be in that situation
<marcan>
unfortunately logistics is a nonzero cost :/
<mps>
marcan: yes, this is my experience
<richardburleigh>
to be fair, a lot of 'old' Macbooks are still extremely capable for what most people do on a computer
<marcan>
I mean I'm typing this on a 2015 iMac
<mps>
richardburleigh: yes, my wife use macbook from 2009 year and it works quite well
<marcan>
~7-10 years is the time horizon for hardware that was decently good/high end when it was new, for "normal" users, in my experience
<marcan>
older than that and you start limiting yourself to very nondemanding users only
<marcan>
stuff that was cheap new at that age is often unusable with modern software
<marcan>
(typically not enough RAM)
<marcan>
(but often you can't even upgrade it to usability at that point)
<mps>
my son upgraded this old macbook to 8GB ram and changed old rotational disk with SSD one, machine is still good for development
<marcan>
that's usually how it goes; if it was high end when new, it can have a pretty long life, but if it was cheap / barely scraping by from the get-go... nobody's going to want it 10 years down the road
<marcan>
yeah, these days 8GB is just barely good enough for decently broad usage these days as long as you don't open 100 browser tabs like I do :)
<marcan>
I regularly run out of RAM on this 32GB machine and I swear I don't know how it's this bad
<marcan>
I'm convinced there is something horribly wrong about xorg/amdgpu/the x86 ecosystem in general because even Firefox does this now and the M1s seemingly have no issue with 8GB
<everfree>
you can get very far on 2GB ram with a 32 bit os, but at that point it horseshoes back into being a power-user only machine with the little microtweaks you've got to do to get everything working well
<marcan>
yeah, I mean it'll boot and you can work on it, for sure
<j`ey>
my main personal machine (until I can switch to the m1) has 4GB lol
<marcan>
but you won't get the experience people "expect" of course
<richardburleigh>
yeah, I saw your thread marcan about the engineering reasons for not having upgradable RAM in the M1 Macs and of course it is part of why they are such amazing hardware. It is still kind of sad that they will be ewaste some day.
<marcan>
richardburleigh: the thing is laptop RAM upgrades aren't as big a deal as people make them to be, IMO
<mps>
well, old arm32 chromebook with 4GB ram works fine with about 12-30 tabs in firefox
<marcan>
sure, it extends the life of the machine a *bit*
<mps>
though not so fast at compiling
<marcan>
but not *that* much in the grand scheme of things
<marcan>
the good news is at least Asahi will let people have decently up to date software with lightweight DEs after Apple drops support :)
<mps>
I like keyboards quality on these old macbooks, better than on the new ones imo
<Glanzmann>
my main machine had 8 GB of ram until I switched to the m1 mini. The only thing I ran out of memory on was when I opened 100 PDFs at the same time using evince.
<everfree>
browsers also use far less ram on it than i experience them using on 64bit OSes fsr, my only thought is pointer size / alignment? but it doesn't make sense to me that 32 bit would literally cut the ram use in half the way i experience
<Glanzmann>
I use fvwm2, firefox, chrome, mupdf, gimp, xterm, tmux, mpv, sshfs and thats probably it.
<everfree>
there's gotta be more than pointers in a browser
<richardburleigh>
marcan: Fair point.. I imagine very few people bother upgrading components in a laptop vs buying a new one
<marcan>
everfree: a *lot* of browsers is pointers...
<richardburleigh>
even when it is possible
<marcan>
but also, any chance those 32bit machines have smaller screens?
<marcan>
because there's texture sizes too
<mps>
everfree: I even run 32bit userspace on 64bit arm
<mps>
in some cases
<marcan>
not on an M1 :p
<everfree>
ah yeah screen res is different, im comparing a 1000x480 32bit machine to a 1366x768 64bit machine I used to use
<everfree>
32bit one also isn't doing hardware acceleration, while 64bit one was
<marcan>
honestly though, I'm looking forward to daily driving an M1 not only because they're great etc, but also because it will mean I'm more familiar with my own computer than I've ever been... and that means I can debug stupid problems/performance issues properly
dawgy[m] has joined #asahi
<marcan>
(and I will be motivated to do so more than I otherwise would be)
<ChaosPrincess>
marcan: so, when are we getting x32 equivalent for arm? :P
<marcan>
I've debugged/fixed many a deep bug, in random software, but some things and subsystems exceed my time investment:reward:motivation threshold
<marcan>
(like this iMac's lack of GPU power management right now... I debugged it as far as "SMU crashes" and the options there are either yell at AMD people or go on a full blown SMU reverse engineering escapade and I just don't have the time)
<marcan>
(or RT/latency issues in general on, like, every Intel box to some extent)
<marcan>
so hopefully I have a lot more room to debug/fix the deep stuff in M1 machines
<sven>
yeah, I’m very much looking forward to that as well
<richardburleigh>
ChaosPrincess: Are you sure there isn't already a project working on 32bit? The rpi community have achieved some pretty amazing things
<ChaosPrincess>
rips are natively capable of running 32bit no?
<sven>
there’s something exciting about fixing some obscure but that requires you to understand how most of your hardware works :-)
<marcan>
yup
<ChaosPrincess>
and x32 is not quite 32bit arm, its 64 bit, but with 32bit pointrs
<richardburleigh>
ChaosPrincess: Fair enough. My mistake :)
<richardburleigh>
Yeah, I imagine Alyssa has more knowledge of the M1 GPU than most individual developers at Apple
<marcan>
certainly of the shaders/3D engine side :)
<marcan>
that's not hard though, to be honest, because at most large companies very few if any people have good overall knowledge; people tend to specialize a lot
<marcan>
let's see if Lina can figure out the hardware/firmware side :p
<sven>
yeah, I’m sure there are a few engineers that understand most of the HW but there will be many more who just have a very good understanding of some individual component
<marcan>
yup
<marcan>
and a good chance nobody at Apple has as broad knowledge of the systems as we do, because you just don't need that person
<marcan>
broad knowledge of a subsystem or subset, sure, but literally touching ~everything?
<richardburleigh>
yep, I work for a tech company and half my day is spent trying to find the specific person who has knowledge on how some part of our product works
<sven>
could be, depends on how they structure their orgs. I could see them having a few principal engineers who have broad knowledge
<marcan>
yeah but still within a certain subsystem
<marcan>
someone's going to have broad knowledge of XNU, or of the GPU, or of the electrical design, or of ASCs/RTKit, or the boot flow...
<marcan>
but someone covering all of those? not so sure
<sven>
there certainly won't be more than few. maybe some veterans who have been there since the early iPhone days ;)
<marcan>
most principal engineer type people at companies like that will rather know what other person to send you to instead :)
<marcan>
I mean, like, I just don't see anyone having the time/business reason to spend time with everything
<marcan>
it's not much of a thing in big orgs
<sven>
oh, that's what I meant with "broad knowledge of the entire system"
<sven>
someone at least needs to have a big picture view to be able to tell "oh yeah, that's an issue inside xyz please go talk to $whoever"
<marcan>
oh yeah, for sure
<marcan>
what I meant is it's not so likely there's anyone at Apple that has actually, like, hacked at the bits of all the subsystems or reviewed most of the code that goes into them
<marcan>
but you're certainly going to get those types once you get down to subsystem level
<sven>
yeah
<marcan>
so there's going to be people laser focused on, say, the shader compiler
<sven>
i could totally see someone who has been with them since the early iphones who has hacked on everything over the years though ;)
<marcan>
and maybe one or two people who can hack on anything GPU, from firmware to Metal, probably
<marcan>
but go higher than that and it's not very likely
<sven>
yeah, above that it's probably more of a broad knowledge and how everything fits together + some special knowledge in some areas
<marcan>
yup
<sven>
they also get to cheat since they have access to documentation which makes everything easier :D
<marcan>
well... maybe.
<marcan>
if they got disclosed :p
<marcan>
you've interned there, you know how they work :D
derzahl has quit [Remote host closed the connection]
<sven>
i don't because i wasn't disclosed on anything back then :D
<marcan>
:D
<marcan>
(nb: this was a long time ago, don't panic, sven's not tainted, also that)
<sven>
yeah, this was what, 12 years ago now
<marcan>
I do remember your pokemon themed Apple TV cluster :-)
<sven>
right. i ended up writing a black box fuzzer because I just couldn't get access to anything that wasn't public :D
<marcan>
sounds like me and nintendo...
<sven>
and even getting access to _public_ ios builds that ran on their hardware that didn't even jtag enabled required me to beg people to log in on my machine for 10 minutes to restore them
<marcan>
lol
<marcan>
I was going to ask if you at least got JTAG-fused boxes...
<sven>
ofc not, I also didn't have symbols or anything. my entire team was sitting in an area I didn't have access to. my manager actually had to escort me in their for our 1:1s
<sven>
*in there
<marcan>
yup, me and nintendo alright
<marcan>
the funniest moment was when I asked for a Tegra devboard and they brought me a Jetson, I plugged it in... and it had a prototype/early version of the NES Mini interface software on it
<marcan>
"oops"
<sven>
:D
<marcan>
funny how we both ended up doing the same thing (black box fuzzer...)
<marcan>
I found a very PSJailbreak-style panic in their USB stack with that, hopefully they fixed it by release :)
<sven>
i vaguely remember some usb bugs too, but i can't remember if they were exploitable or just DoS
<sven>
I also never saw if they were fixed because I lost access to the bug reports the moment they had been assigned to the correct team :D
<marcan>
I never knew if it was exploitable, it just threw up a panic and I wrote it up, but it might've been
<marcan>
wait no, it was an infinite loop I think? there was another thing, a panic when a joy-con reported >100% battery
<marcan>
that one might've actually been an OOB array access
<marcan>
another funny thing was another random switch in the building happily pairing with my faux-joycon python script at one point
<marcan>
"oops"
<marcan>
was tempted to start pressing random buttons and seeing if they notice :)
<sven>
:D
<asahi-test[m]>
<bluetail[m]> "this is a firmware / GPU issue..." <- could this be somehow related to the issue?
<bluetail[m]>
at this point I think I might fail a A/B the test using the new monitor. But succeed using the old Monitor to tell 'something is off' .
<bluetail[m]>
`Monitors with FRC (Frame Rate Control) use temporal dithering to alternate between different colors in order to ‘fill-in’ the unsupported colors by the display. `
<bluetail[m]>
I suspect that is whats happening
<bluetail[m]>
the newer monitor supports more colors and doesnt behave like this
<bluetail[m]>
`However, some monitors do seem to be more prone to this issue. Pixel inversion artifacts are more obvious and common with high refresh rate TN panel displays.`
<bluetail[m]>
Perhaps I could find out on AsahiLinux (ARCH) if I could force 6 bit mode
<bluetail[m]>
* 6 bit color mode
<paciento[m]>
avoid displayninja if you can. iirc, they are know for plagiarism
<bluetail[m]>
I didnt know
<bluetail[m]>
thanks for warning me
<paciento[m]>
no worries
<bluetail[m]>
Anyways this is intriguing and its also hard to describe when you have it
<bluetail[m]>
marcan: Thats my new best guess. I guess I can only try once xrandr is setup since I think that you cant set a limited RGB range on macOS
<bluetail[m]>
s/setup/implemented/
<Cy8aer[m]>
<richardburleigh> "marcan: Fair point.. I imagine..." <- I ever did it: upgrading RAM changeing the disks (e.g. to SSD) or adding some (if there was an M.2 on the main board). With the mbp I am no horrified to e. g. age out my nvme so I soon prepared measures against it (https://wiki.debian.org/%20SSDOptimization)
<Cy8aer[m]>
because I swap my devices after 6, 7 years at earliest.
<Cy8aer[m]>
(the earliest)
<Cy8aer[m]>
And it is still a problem for services when there is not the expected os on the devices. I guess Apple will be very happy with a linux machine...
<Cy8aer[m]>
Even with care...
<Cy8aer[m]>
(I am horrified - sorry a no to much)
<Cy8aer[m]>
Another thing is the battery. I hope that the loading hardware is able to stop loading at e. g. 80%-90% - read somewhere that it is not possible with macOS) and someone builds up an interface for it sometimes, to have the battery live longer.
MajorBiscuit has joined #asahi
<Cy8aer[m]>
I learned: forget the service, you need to repair it by yourself or protect your hardware not dying. It's the same with my SailfishOS-Devices...
<Cy8aer[m]>
... and stock windows laptops - even these black ones with the red LED...
darkapex has quit [Ping timeout: 480 seconds]
sepkov[m] has joined #asahi
c10l has quit [Quit: Bye o/]
Moprius has joined #asahi
c10l has joined #asahi
<marcan>
the battery charging can be controlled from userspace, so you can set policy there
<marcan>
that is already supported
<marcan>
it's not a fixed threshold, you need a daemon to actually turn off charging when it reaches your desired level and such
<mps>
oh, didn't know. that is really good thing
<mps>
marcan: I thought you implemented this in kernel driver
<marcan>
not the threshold, no
<marcan>
the kernel driver just exposes the charge control that SMC does, with the standard sysfs API for this that already exists
<mps>
only writeable file is charge_behaviour in /sys/class/power_supply/macsmc-battery
<marcan>
yes, that's the one
<marcan>
you set it to either 'auto', 'inhibit-charge', or 'force-discharge'
<mps>
I see 'auto'
<marcan>
which is the default
<mps>
yes
<marcan>
so your policy daemon should set it to 'inhibit-charge' once you reach 80% or whatever level you want
<Cy8aer[m]>
marcan: good and bad news: some thinkpads have a threshold inside the loading chip.Of course - if the power management is working I probably will not switch the device off anymore. But when loading it switched off it will then load up to 100% ☹️
<mps>
marcan: what is 'force-discharge'?
<marcan>
what it says on the tin
<marcan>
run off of battery, even if plugged in
<mps>
I guessed so, thank for confirmation
<marcan>
I don't know how the macs handle being shut down as far as charge control goes; it's possible it sticks at the last setting?
<marcan>
I think it will reset at boot time until the OS sets it to the desired value, since at boot time the SMC resets
<marcan>
since it's not a threshold though, it might work if you shut down the machine already charged to the desired level, but it won't work if you leave it charging shut down since nothing will be there to stop it
<asahi-test[m]>
<marcan> "the battery charging can be..." <- nice, i can see the battery charge level on the console with acpi --battery --details, really useful!
<marcan>
but obviously at that point you probably wouldn't be shutting down very often anyway
<mps>
marcan: thanks. I will try to write simple daemon to test this
jluthra has quit [Remote host closed the connection]
<kprasadvnsi[m]>
M1 Mac mini as Gitlab runner running Asahi linux minimal.
Ll3macorn[m]1 has joined #asahi
<Ll3macorn[m]1>
hi
<Ll3macorn[m]1>
im having issues
<Ll3macorn[m]1>
i do the curl and sh thing and then it spits out tons of things that are like "Can't unlink already-existing object" or "Can't restore time"
<Ll3macorn[m]1>
also this "tar: Error exit delayed from previous errors."
<Ll3macorn[m]1>
from the end
<deagle50[m]>
Any ETA on display brightness control?
* eta
on display brightness control
<Ll3macorn[m]1>
yeah so can anyone help me
<asahi-test[m]>
Ll3macorn[m]1: what is the first error you get?
<asahi-test[m]>
cannot advice because i have no experience with that
<asahi-test[m]>
s/advice/advise/
<asahi-test[m]>
asahi-test[m]: however i have made several installs from an admin promt on macOS and that never happened
<Ll3macorn[m]1>
its a extraction problem
<asahi-test[m]>
and i have never turned that o
<asahi-test[m]>
s/o/on/
<Ll3macorn[m]1>
should i try the installer-v0.3.9s install.sh
<Ll3macorn[m]1>
if i do try i get the error "python.framework" is damaged and cant be opened
<jannau>
I'd say dcp will have brightness control well before it is finished. It currently doesn't have it. I've seen how it works on the macbook pro 14"
<jannau>
it's on my todo list for the initial release
<asahi-test[m]>
Ll3macorn[m]1: also i have always used the script from this page
<Ll3macorn[m]1>
yeah but the damaged error
<asahi-test[m]>
asahi-test[m]: meaning this one curl https://alx.sh | sh
<asahi-test[m]>
* this one "curl https://alx.sh, * | sh" ( without the quotes )
<Ll3macorn[m]1>
zsh: exec format error: /Users/baileyboylson/Downloads/installer-v0.3.9/Frameworks/Python.framework/Versions/3.9/Python
<asahi-test[m]>
Ll3macorn[m]1: i think you should address the first error first
<Ll3macorn[m]1>
asahi-test[m]: what do you mean?
<asahi-test[m]>
Ll3macorn[m]1: about the time machine. Just to be sure you are using the script on a M1 mac right?
<Ll3macorn[m]1>
asahi-test[m]: yeah ofc
<asahi-test[m]>
Ll3macorn[m]1: ok then, can you close the terminal, open another one and use this script?
<asahi-test[m]>
you do not have any linux partitions created yet
<Ll3macorn[m]1>
?
<asahi-test[m]>
meaning as you said the script only gave you the first error. So that is stopping the whole process. As i said, have no experience with time machine so i cannot advise further.
<marcan>
it's not time machine, why would it be time machine
<marcan>
Ll3macorn[m]1: try sudo rm -rf /tmp/asahi-install and run it again after that