<Jari-->
Windows task manager taking 40% of the CPU quad cores.
<\Test_User>
"don't use ms's stuff"
<Jari-->
Interrupts, too, etc. overhead. Slowing my AMD to literally unusable.
<Jari-->
\Test_User: this is hardware related
<gog>
interrupts tend to have a little more overhead on windows than linux yeah
<\Test_User>
Jari--: are you running on a 400mhz quadcore cpu then?
<gog>
but the question is whether task manager is constantly taking 40% or if you're just seeing it spike to 40% when it refreshes the task list
<gog>
by default it'll update once per second and that 40% might be an artifact of the accounting
<gog>
take top in linux for example
<gog>
you can make it eat cpu time by refreshing it multiple times a second
<gog>
and it'll show you that it's taking a lot of CPU time because of the overhead of walking /prof
<gog>
proc
<gog>
i find it really unlikely that task manager is the cause to any performance problems you're having
<zid>
taskmgr uses 1% cpu on my machine, I think it rounds up
<\Test_User>
I've seen ms's task manager eat up much more when there's more other load on the system
<gog>
that's possible too
<zid>
yea higher iowait or whatever
<gog>
but it's not likely the underlying cause
<\Test_User>
ofc, underlying cause is that underlying system designed by ms :P
<gog>
he ynow
<gog>
there's a good operating system inside the layers of bloat
<zid>
yea the kernel is fairly solid
<gog>
remember windows xp
<gog>
windows 2000
<zid>
it does a lot of stuff, doesn't crash much, etc
<zid>
the vfs is just a little slow
<zid>
I just wish they wouldn't ship it with 20GB of new themeing each year
<gog>
they're starting to put ads in the start screen
<gog>
idk if this is going to become a permanent feature
<zid>
not in w7 yet they haven't, phew, I'm safe
<gog>
yeh in 11
<gog>
windows 7 is decent
<gog>
windows 8.1 was fine if you used a start menu replacement
<zid>
if w7 becomes obsolete I'm either switching to xp64 or w10, haven't decided yet
<gog>
windows 10 was decent until the creator update
<gog>
which was just a bunch of crap
<zid>
maybe w2k?
<gog>
good luck finding drivers
<gog>
unless wdf was backported to windows 2000?
<gog>
oh it originated on w2k
<gog>
hm
<gog>
might be fine then idk
<gog>
but
<gog>
what about 64-bit support
<zid>
I'll run 32 copies of it in a VM
<zid>
that'll be about the same
<gog>
works for me
Belxjander has joined #osdev
gog has quit [Ping timeout: 260 seconds]
sonny has joined #osdev
sonny has quit [Quit: Client closed]
sonny has joined #osdev
sonny has left #osdev [#osdev]
Jari-- has quit [Remote host closed the connection]
Jari-- has joined #osdev
isaacwoods has quit [Quit: WeeChat 3.7.1]
nyah has quit [Quit: leaving]
Raito_Bezarius has quit [Quit: free()]
vdamewood has joined #osdev
sebonirc has quit [Remote host closed the connection]
Raito_Bezarius has joined #osdev
sebonirc has joined #osdev
Raito_Bezarius has quit [Max SendQ exceeded]
heat has quit [Ping timeout: 256 seconds]
Raito_Bezarius has joined #osdev
epony has quit [Remote host closed the connection]
epony has joined #osdev
<sbalmos>
ahhh, Win2k3. Those were the days.
<sbalmos>
xp64 in classic mode might be a welcome change back
<zid>
xp64 is 2k3 with xp's theme pack
<zid>
rather than being a rebuild of xp's kernel for 64bit
<zid>
guess that branch didn't get any love
<geist>
yah i dont thin it did. think of it as for some sort of specialized workstation stuff at the time
<geist>
that didn't really pan out until later
<geist>
also iirc there wsas some builds of XP that had PAE enabled until they found out that there were so many buggy drivers that couldn't handle it they basically turned it back off again in SP2 or whatnot
<geist>
the bugginess being that there was lots of 32bit PCI devices on a 32bit sytem so drivers had to handle going through bounce buffers and whatnot to deal with >4GB
<geist>
but most didn't, or didn't do it properly, etc
<geist>
for servers you could get away with it because you could generally be more strict about which drivers were certified for server use
srjek has quit [Ping timeout: 268 seconds]
Vercas6 has joined #osdev
Vercas6 has quit [Client Quit]
Vercas6 has joined #osdev
Vercas6 has quit [Client Quit]
Vercas6 has joined #osdev
Vercas6 has quit [Client Quit]
Vercas6 has joined #osdev
Vercas6 has quit [Client Quit]
Vercas6 has joined #osdev
bradd has joined #osdev
sebonirc has quit [Remote host closed the connection]
sebonirc has joined #osdev
zaquest has quit [Remote host closed the connection]
zaquest has joined #osdev
sebonirc has quit [Remote host closed the connection]
sebonirc has joined #osdev
amwps290 has joined #osdev
troseman has quit [Ping timeout: 248 seconds]
bgs has joined #osdev
<geist>
watching space launch fun!
Vercas6 has quit [Ping timeout: 255 seconds]
<Ermine>
Who's launching?
<geist>
NASA, Artemis. just launched
<Ermine>
Yay
<sham1>
Finally
Vercas6 has joined #osdev
<Ermine>
Are there people on board?
<sham1>
Apparently not
<geist>
no it's the first launch, so it's a test launch with nobody on board, but it'll in an hour burn for moon
<vdamewood>
Can we burn the moon?
Belxjander has quit [Quit: Exit()]
<kazinsal>
god considering how nervewracking the first minute or so of that launch was *without* people on board, the first crewed mission is going to be a serious butt-clencher
<kazinsal>
all the way up I was just thinking "please don't blow up please don't blow up"
<mrvn>
Looking at the nasa websites it's damn hard to find the launch. Even the news site only has "Core stage main engine cutoff, core stage separation complete, 19 minutes ago" in an easily overlooked window. Why isn't the launch the headline with live and recent video?
<geist>
yeah i have to admit nasa is not as good as providing real time infos about their launch as the spacex things i've been watching a lot of
<geist>
now the official feeds are mostly just a cycling of various onboard cameras
<mrvn>
the blog seems to have a running commentary of what's happening
<vdamewood>
Rocket go 'whooosh'
bgs has quit [Remote host closed the connection]
<kazinsal>
hundreds of ULA and Rocketdyne engineers are having a coordinated adrenaline-wearing-off-vomit right now
<mrvn>
Do they not have any native american astronaut in the program? Only found a woman and a black man to send to the moon?
<mrvn>
Or is the woman black to get a double score?
<mrvn>
"Artemis missions enable a growing lunar economy ..." It would be hard to shrink it from 0
<mrvn>
hehe
<kazinsal>
iirc from the third Artemis mission onward they're going to be combined lunar surface missions and lunar orbital space station construction missions
<mrvn>
Why do they have to launch when I have to work? I wanna watch.
<kazinsal>
launch windows for translunar injection are pretty tight
<kazinsal>
the solution to that issue is like, handwavium antimatter drives and shit
<Mutabah>
Ooh, is A1 launching?
<Mutabah>
It survived the storms?
<kazinsal>
yep, it's up there and burning away happily
<Mutabah>
Nice!
<kazinsal>
core stage is separated and the Orion solar arrays have deployed. next burn is in 10 minutes
<kazinsal>
TLI burn should be in 40
<geist>
yah in this case it happened to be at night, but basically you gotta launch whenever the disc of the lunar orbit crosses the launch site
<geist>
i thin it's like a 2 hour window
<kazinsal>
yeah
<pitust>
why are the lunar transfer windows so hard to hit? i can kinda eyeball it most of the time in ksp
<bslsk05>
gist.github.com: Bootable 16bit C program with gcc-ia16 · GitHub
<Nuclear>
I was looking for a 16bit C compiler, and remembered gcc-ia16. It's fun ... look at the cute way I used to place the boot signature with the link script :)
<zid>
nod I wrote nearly the same script for someone else 2 days ago
<Nuclear>
stage1 in asm, stage2 in C loaded high in conventional memory, to then go ahead and load the kernel low
<Nuclear>
(I'm just starting playing around with the idea for a new project, an OS for 8088/8086
<Nuclear>
)
<sbalmos>
I'm still just writing down my OS features/goals outline
<zid>
sbalmos: then colouring it in with crayons after
<sbalmos>
zid: I'm classier than that. I use dry erase marker.
<zid>
ngl crayons are too tasty to waste on colouring in
<sbalmos>
the yellow and orange crayons make a delicious grilled cheese sandwich
PapaFrog has quit [Ping timeout: 260 seconds]
Raito_Bezarius has quit [Ping timeout: 260 seconds]
PapaFrog has joined #osdev
<sbalmos>
creating an EFI System Partition puts a shim MBR stage1 bootloader in the first sector for BIOS compatibility. But what code does this ESP shim bootloader load next? Everything else on the ESP is a UEFI app, which would expect being loaded from Boot Services and expect an EFI-init'd environment.
<FireFly>
sure, but he heard some tech person say it before and thought it sounded fancy
<bslsk05>
twitter: <fixedpointfae> running low on disk space on macOS or linux? remember to empty out your recycle bins (located at /usr/bin and /bin)
<gog>
just shove the whole bunch in his throat
<FireFly>
and can repeat it to try to out-tech the interviewee
<heat>
"do you think you know better than me? you're fired"
gxt has joined #osdev
<FireFly>
I mean that does seem to be the vibe I'm getting from his companies :P
<heat>
"but this is just an interview"
<gog>
yeh my advice to any engineer that works for elon that wants to keep their job is to just quietly know more than him
<FireFly>
honestly I'm just sad for the employees stuck there for visa reasons
<gog>
unless you work for twitter which is probably irretrievably headed to insolvency
<gog>
in which case go nuts
<FireFly>
or don't answer "yes" to the recent email thing and get 3 months' severance apparently
<zid>
Make twitter comment saying you're going to buy twitter, twitter stock spikes, you own 8% of twitter. FEC says that's manipulation and now you have to actually buy it. Twitter is in a bunch of debt, ???, bankruptcy.
<gog>
afaik if they don't take the reinstatement offer then it counts as quitting
<zid>
We're seeing the ??? phase.
<gog>
and they lose the severance
<gog>
but if they take their jobs back they might not see 3 more months of pay anyway
<gog>
so it probably doesn't make any difference
<FireFly>
gog: idk the email explicitly said people who don't answer yes will get 3 mo severance from what I could tell, but it does sound crazy to me especially with what I see on the internet from US companies...
<FireFly>
but yeah I've no idea
<gog>
and if the company goes bankrupt then fired employees become one of the lowest class of creditors
<FireFly>
true
<gog>
which is fucking bonkers to me, but that's US labor law in a nutshell
<gog>
trump has fucked over thousands of people with this tactic
<bslsk05>
twitter: <donie> Here’s the text of the email Musk sent to Twitter staff overnight.   Those who don’t commit to being “extremely hardcore” by 5pm ET today must leave the company. ‼️   Story: <cnn.com/2022/11/16/tec… https://t.co/expt0d63dH> https://pbs.twimg.com/media/Fhr5owsXEAEcB76.jpg
<FireFly>
with the "anyone who has not done so by 5pm will receive three months of severance"
<gog>
the catch is that's not a binding statement
<gog>
without a formalized agreement he can say whatever and do another
<FireFly>
makes sense I guess
<gog>
tech workers organize pls
<FireFly>
yes
<heat>
WE NEED TO BE EXTREMELY HARDCORE
<heat>
mjg, STAY HARD
<FireFly>
(I say, not being in a union either, but shhh)
<heat>
HARD-CORE
<gog>
i am a union member
<gog>
it's pretty dope
<FireFly>
union { cat gog; }
<gog>
yes
<heat>
unions are not dope
<heat>
careful with the UB
<gog>
union {cat_t cat: girl_t girl;} gog
<zid>
unions are dope
<gog>
;
<zid>
I use unions all the time
<zid>
I wrote one yesterday
<heat>
gog, so you're either cat or girl per the C standard
<bslsk05>
'Udienza Generale 05 gennaio 2022 Papa Francesco' by Vatican News - Italiano (01:12:34)
<heat>
hahahhahahahahahaa
<heat>
i mean, i think the pope loved it so
<zid>
pope is cultured
<FireFly>
I mean it is a good bop
vdamewood has joined #osdev
xenos1984 has quit [Ping timeout: 256 seconds]
<heat>
"new papal bull replaces ave maria with megalovania"
vinleod has joined #osdev
vdamewood has quit [Ping timeout: 260 seconds]
xenos1984 has joined #osdev
xenos1984 has quit [Ping timeout: 256 seconds]
vinleod is now known as vdamewood
xenos1984 has joined #osdev
bauen1 has quit [Ping timeout: 268 seconds]
bauen1 has joined #osdev
srjek has quit [Ping timeout: 268 seconds]
kaichiuchi has quit [Ping timeout: 260 seconds]
spikeheron has quit [Quit: WeeChat 3.7]
kaichiuchi has joined #osdev
kaichiuchi has quit [Max SendQ exceeded]
kaichiuchi has joined #osdev
kaichiuchi has quit [Max SendQ exceeded]
carbonfiber has quit [Ping timeout: 260 seconds]
carbonfiber has joined #osdev
friedy has joined #osdev
carbonfiber has quit [Max SendQ exceeded]
spikeheron has joined #osdev
geist has quit [Ping timeout: 252 seconds]
Vercas6 has quit [Remote host closed the connection]
Vercas63 has joined #osdev
gareppa has joined #osdev
geist has joined #osdev
kaichiuchi has joined #osdev
kaichiuchi has quit [Max SendQ exceeded]
kaichiuchi has joined #osdev
<gog>
apparently elon is stepping down as twitter ceo
dude12312414 has joined #osdev
<FireFly>
huh!
<FireFly>
see that's what I was saying would make sense from day 1, but I didn't think it'd happen
kaichiuchi has quit [Max SendQ exceeded]
<FireFly>
maybe it's so the downfall can be blamed on someone else :p
<gog>
maybe
<gog>
or it's a genuine attempt to rescue the investment
<gog>
perhaps the people he owes money to convinced him that it was his own idea
carbonfiber has joined #osdev
<FireFly>
perhaps
<FireFly>
it's the only good decision I've heard about twitter since he took over
carbonfiber has quit [Max SendQ exceeded]
<GeDaMo>
Maybe he thinks he already did enough to save it :P
<FireFly>
not to be all conspiratorial, but part of me said earlier (like.. idk a week ago? time is so weird currently) that it seems he's making all the decisions to ruin the value as quickly as possible, wondering if there's some way he could possibly stand to gain from it
<FireFly>
like idk.. shorting the company, insider trading, insurance fraud or something
carbonfiber has joined #osdev
<gog>
that would be in character for him
<gog>
he has used twitter to manipulate the stock market before
carbonfiber has quit [Max SendQ exceeded]
<gog>
SEC should have slapped him harder for the taking tesla private at $420/share stunt
<zid>
I already posted why
<zid>
Make twitter comment saying you're going to buy twitter, twitter stock spikes, you own 8% of twitter. FEC says that's manipulation and now you have to actually buy it. Twitter is in a bunch of debt, ???, bankruptcy.
<zid>
SEC*
<zid>
He's just going to try write it off, rather than paying its debt for forever
<FireFly>
gog: apparently he's trying to argue that SEC thing is "not valid because done under duress", in court, now from what I hear
<bslsk05>
twitter: <themaxburns> The attorney questioning Musk seems...unconvinced by that argument.   "Are you trained as a lawyer?"   MUSK: "I have some familiarity with the legal system. If you're in enough law suits, you pick up a few things."
<gog>
"shitty management" is not duress
carbonfiber has joined #osdev
<FireFly>
yeah no doubt
kaichiuchi has joined #osdev
bauen1 has quit [Ping timeout: 256 seconds]
<heat>
gog, doesn't he literally have other partners in the investment too
Vercas63 has quit [Quit: Ping timeout (120 seconds)]
Vercas6 has joined #osdev
<gog>
yes
Burgundy has joined #osdev
sebonirc has quit [Remote host closed the connection]
sebonirc has joined #osdev
antranigv has quit [Ping timeout: 260 seconds]
antranigv has joined #osdev
gareppa has quit [Quit: Leaving]
bauen1 has joined #osdev
<Bitweasil>
We'll see what comes of Twitter, but... it certainly seems like "Hotshot runs his mouth, annoyed he actually got held to what he signed" is a useful description of the events so far.
<Bitweasil>
"You suck, I should just buy you!" "Yeah, yeah, whatever... *signs paper in front of him*" "You mean I *what*???"
bauen1 has quit [Ping timeout: 256 seconds]
bauen1 has joined #osdev
<gog>
hubris comes sometime in the autumn
<gog>
or something
epony has quit [K-Lined]
<zid>
I copy pasted some code from a different project, it now doesn't work
<zid>
stupid floating point errors.
<zid>
Turns out I never handle the cases where things are exactly aligned properly, makes a nice /0
<zid>
the previous code used it for drawing spinny things and the rotation amount caused fractions, I randomly picked a nice even spinny speed, crash
<gog>
spinny
* gog
goes spinny
<zid>
oh it works WAY better though if you remember to call sort
<Bitweasil>
lol.
<Bitweasil>
Nice.
antranigv has quit [Ping timeout: 256 seconds]
antranigv has joined #osdev
GeDaMo has quit [Quit: I'm going to show these people something you don't want them to see. I'm going to show them a world without you.]
<Bitweasil>
\o] Hack the planet, man! Hack the planet!
<Bitweasil>
Also, *Discord*? Ugh.
<zid>
It's easy af to use as an image host
<Bitweasil>
Fair enough.
<Bitweasil>
I just hate Discord as a concept.
<zid>
I just have a private channel open and I hit paste
<zid>
and I get a url back
<Bitweasil>
"What do we want? Decentralized private chat servers! How do we want it run? By a centralized company that can make sure we don't do anything they don't like!"
<Bitweasil>
It's like IRC, if Meta ran all the IRC servers.
<zid>
I need to backface cull now, so that should be testing the crossproduct of two of the verts vs the dot product with [0,0,1] I think idk
<Bitweasil>
I know that's a graphics optimization, what does it do exactly?
<zid>
it's not even an optimization
<Nuclear>
zid: it's better if you use vertex winding for backface culling
<zid>
it stops the faces that are at the back drawing on top of the ones at the front
<Bitweasil>
Oh, OK.
<zid>
Nuclear: that's what that uses
<zid>
the crossproduct gives the opposite answer if you do cross(v0, v1) to cross(v1, v0)
<zid>
rather than using backed in normals
<Bitweasil>
"If this triangle is entirely covered by others, don't render it"?
<Bitweasil>
... yeah, I'm happy to have a working framebuffer these days.
<bslsk05>
isocpp.org: P2723R0: Zero-initialize objects of automatic storage duration
<zid>
If you take two vectors (two of the lines on my triangle, in this case) and take the cross product, that gives you a vector pointing orthogonal to both
<zid>
meaning, pointing 'out' of the face
<Bitweasil>
My computers are progressively less functional. A 12 year old recently asked of me, "Why are your computers all so... *weird*?"
<Bitweasil>
*nods*
<zid>
if it's facing the camera, draw it, if not, don't
<zid>
now the back faces get culled, and don't draw potentially on top of the front faces
<Bitweasil>
(I'd been trying to get Plex to play back properly, and well wouldn't you know, the kernel updated without the DKMS nVidia drivers updating, and the sound module wasn't there, and... etc)
<Bitweasil>
Ok.
<Nuclear>
backface culling works for visible surface determination only for convex objects though. You need something more for the general case. Ideally a zbuffer