<bslsk05>
hackaday.com: Review: DevTerm Linux Handheld Has Retro-Future Vibe | Hackaday
<mrvn>
The SOC is the little green board on the last picture, right? The rest is just wirering?
<kazinsal>
seems to be
<kazinsal>
display is 8:3, 1280x480
<moon-child>
I have an electric typewriter
<moon-child>
it's neat but ... terrible ergonomics
<moon-child>
what I really wanna do is convert a typewriter into a teletype
<gog>
irc->teletype bridge
<gog>
you'll always know when the channel is active
<moon-child>
I mean, telnet
<GeDaMo>
Nah, what you want is a loom which weaves the output into cloth :P
<moon-child>
is an adequate irc client (aside from pingpongs). And I don't think you can do much better given a line-oriented interface
<gorgonical_>
mrvn: yeah that's right. It's just a nice wiring up so you can hack on it
<gorgonical_>
I'm getting into risc-v now and am always more interested when I've got hardware to play with and not qemu
<gorgonical_>
Since working on arm64 stuff my trust in qemu/virtualization in general has been shaken
<Bitweasil>
gorgonical_, I had to make a will save against one of those... or the ARM64 stuff...
<gog>
GeDaMo: i love it
<Bitweasil>
Oh man. Now I want a dot matrix IRC client.
<GeDaMo>
You'd be able to wear your chat log! :P
<Bitweasil>
I used to think about "logging to greenbar." Can't hack the logs if they're external and printed! :D
<gorgonical_>
As a grad student I have to give myself some joy. Beer only can be used so much for that
<Bitweasil>
Gotta graduate, get a good job in the low level weeds, they you can afford whiskey and vodka! :D
<Bitweasil>
I'm only half kidding. :/
<Bitweasil>
<gorgonical_> Since working on arm64 stuff my trust in qemu/virtualization in general has been shaken <-- Mind expanding on that?
<Bitweasil>
It's a hot mess all the way down, we all know it, just curious since I'm in the AArch64 weeds these days too.
<gorgonical_>
I got bit pretty bad recently because of qemus departure from hardware behavior. I spent days before I considered that qemu would be wrong
<mrvn>
Getting a loom to scroll fast enough for just this channel would be horribly loud.
<j`ey>
gorgonical_: on what bit?
<gorgonical_>
And then after poring over it realized qemu would only ever be about an 80% correct solution
<gorgonical_>
Tbf I had an older qemu version, but the board emu literally did not have nested virt support. The registers, etc were there. But actually nested virt did nothing at all
<Bitweasil>
Oh... heh.
<j`ey>
oh, this was qemu 2.x right
<Bitweasil>
Yeah, stuff like that should warn badly if you touch something that's present but not implemented.
<gorgonical_>
Whatever came with Debian 9 I think
<Bitweasil>
<meme>What year is it???</>
<gorgonical_>
The changelogs even said arm64 support was "nearly complete" for my version
meisaka has quit [Ping timeout: 256 seconds]
<Bitweasil>
[1] For basic version of non-complex ARM64 functionality that comes nowhere near edge cases. ;)
meisaka has joined #osdev
pretty_dumm_guy has joined #osdev
Ermine has quit [Quit: Cya!]
kingoffrance has quit [Ping timeout: 250 seconds]
<Bitweasil>
I remain generally concerned about (1) the ever-increasing complexity of the tech stacks we rely on and (2) the lack of younger people doing a lot of deep work in it. You're an oddity, really.
<Bitweasil>
There are a lot of people, 40+, who make up the core of the field.
<gorgonical_>
It helps that I was exposed to a lot of old PCs when I was young. Used a crappy windows 95 pc until I was like 13.
<gorgonical_>
In 07 or so
Ermine has joined #osdev
<Bitweasil>
Yeah, "Grew up on 486s or early Pentiums" is about the last wave of the low level folks, with a few rare exceptions.
<mjg>
heh
<mjg>
now that you mention it, someone did a quick check who is developing freebsd age-wise
<mjg>
i don't remember the distribution exactly, but basically late 30s is the youngest you can expect excluding some minor exceptions
<mjg>
i happen to qualify, with 486 dx 2 being my first pc cpu
[itchyjunk] has joined #osdev
srjek has joined #osdev
<mrvn>
C16
<Bitweasil>
I'm ~40, and I know very few people more than a couple years younger than me in the low level spaces.
<Bitweasil>
We had an 8086, but a 486 DX2/66 was the first PC I was the sole user of (everyone else wanted to use the hot new Pentium 200, so, uh... fine, I'll do my stuff over here).
<Bitweasil>
Eventually it ended life as a 586/150 on a 50MHz FSB.
k8yun has quit [Quit: Leaving]
<Bitweasil>
And either my experiments with installing Linux or having the FSB 50% overclocked eventually led to it being less-than-reliable, so I got a Pentium III before heading off to college.
<GeDaMo>
ZX81
JanC has quit [Remote host closed the connection]
pretty_d1 has joined #osdev
pretty_dumm_guy has quit [Ping timeout: 240 seconds]
_xor has quit [Read error: Connection reset by peer]
_xor has joined #osdev
<mrvn>
Can I make a "tee" in C++ coroutines?
wootehfoot has quit [Read error: Connection reset by peer]
JanC has joined #osdev
gorgonical_ has quit [Ping timeout: 240 seconds]
gorgonical_ has joined #osdev
pretty_dumm_guy has joined #osdev
GeDaMo has quit [Remote host closed the connection]
pretty_d1 has quit [Ping timeout: 240 seconds]
Vercas5 has joined #osdev
Vercas has quit [Remote host closed the connection]
gorgonical_ has quit [Read error: Connection reset by peer]
bauen1 has joined #osdev
mrvn has quit [Ping timeout: 250 seconds]
srjek has quit [Ping timeout: 240 seconds]
mrvn has joined #osdev
_eryjus has quit [Remote host closed the connection]
Burgundy has quit [Ping timeout: 250 seconds]
eryjus has joined #osdev
<Ali_A>
in the ELF format,
<Ali_A>
I know that `.init` section is responsible for process initialisation and I know that .fint section is responsible for process termination
<Ali_A>
I am just asking is this the place where the OS frees memory from the process? I Know no matter how complicated your memory allocations, once u terminate a process, a good OS (linux/windows/mac) will automatically regain the memory that you were using (or your application in this case)
<Ali_A>
I am just wondering, where does `freeing` memory happens ? is it happening at the `.fint` or it happens in other places?
<Ali_A>
(if it depends on the OS, then any OS example would be useful for me)
<mrvn>
Ali_A: that's kernel work. Nothing to do with the process.
<Ali_A>
mrvn so I assume the kernel will keep track of which physical pages or something equivalent and then reclaim them after the process terminates
<Mutabah>
yep
<Mutabah>
when it tears down the address space, it deallocates the used pages
<mrvn>
LostFrog: and 2 years later the standard adds that and BOOM
<mrvn>
ups
* LostFrog
wakes up.
<mrvn>
LostFrog: sorry, ewin
<LostFrog>
No problem.
Ali_A has quit [Quit: Connection closed]
[itchyjunk] has quit [Remote host closed the connection]