@tnt@chaos.social avatar

tnt

@tnt@chaos.social

RF / Electronics / Random stuff ...

This profile is from a federated server and may be incomplete. Browse more on the original instance.

whitequark, (edited ) to random
@whitequark@mastodon.social avatar

Can you answer this simple question about the C programming language?

$ cat test.c
<string.h>
<stdio.h>
int main() { printf("%d\n", strlen("\x01c")); }
$ cc test.c && ./a.out

What number does this program print?

tnt,
@tnt@chaos.social avatar

@whitequark If you have syntax highlight it makes it obvious

whitequark, to random
@whitequark@mastodon.social avatar

did you know: maximum-sized USB HS BULK packets where the payload is all-ones are about 16% longer to transmit than the ones where the payload is all-zeroes or random?

this is easily measurable using the with a slightly modified benchmark applet:

I: g.applet.internal.benchmark: mode sink-zeroes: 42.84 MiB/s (342.70 Mb/s)
I: g.applet.internal.benchmark: mode sink-ones: 38.91 MiB/s (311.25 Mb/s)

(the measured overhead is about 10% due to other factors, e.g. scheduling)

tnt,
@tnt@chaos.social avatar

@whitequark So will you implement scrambling in glasgow transport layer ? 😁

DrewNaylor, to videoediting
@DrewNaylor@mastodon.online avatar

PSA: Resolve Studio requires the USB drive to remain plugged in and will check every few seconds for its presence.

tnt,
@tnt@chaos.social avatar

@stooovie @DrewNaylor
Well there are 3 kind of licenses :

  • The USB dongle one
  • The software activation code one
  • The apple app store one (only on Mac obviously).

If you're using the USB dongle one , yes, you need the dongle to keep being plugged in while using it.

gsuberland, to random
@gsuberland@chaos.social avatar

to help you with the condition that causes you memory problems, you should take these tablets every day of the week. we put the tablets in sheets of ten. do you like this.

tnt,
@tnt@chaos.social avatar

@gsuberland They should do it like for contraceptive pills : you have a sticker you put on the sheet and they give you 7 different sticker in the box depending on which day of the week you started that sheet, so each pill ends up marked with the day of the week to take it.

gsuberland, to random
@gsuberland@chaos.social avatar

captain's log, star date 20240409.1

ow. computer, summon the EMH so he can get me some painkillers.

tnt,
@tnt@chaos.social avatar

@gsuberland Please state the nature of the medical emergency

tnt, to random
@tnt@chaos.social avatar

Me wondering why my signal level sucks...
And then :

ploum, (edited ) to random French
@ploum@mamot.fr avatar

Quelles sont vos alternatives à Google Drive pour des étudiants utilisateurs de LibreOffice qui doivent collaborer pour rédiger un rapport ? Les étudiants ne sont pas des geeks, pas de self-hosting.

UPDATE : Merci, plus besoin de réponse. J’ai Framapad, Cryptpad et OnlyOffice (du plus léger au plus complet)

tnt,
@tnt@chaos.social avatar

@ploum One does all the work and the others stay out of their way.

gsuberland, to random
@gsuberland@chaos.social avatar

What do the Goa'uld say when they want to measure the light quality of a white LED?

Jaffa, CRI!

tnt,
@tnt@chaos.social avatar

@gsuberland Speaking of which is there a (cheap) way to measure that without a full blown spectrometer ?

tnt,
@tnt@chaos.social avatar

@felix Interesting part, I had no idea this even existed. I wonder if anyone makes something like an OSHW screen calibrator using that.

tnt,
@tnt@chaos.social avatar

@gsuberland Yup, I've seen those before 🤩

gsuberland, to random
@gsuberland@chaos.social avatar

can anyone explain this section on preferred length equalisation patterns, from STM AN5122, where they say that switchback patterns (aka trombone routing) are preferred over serpentine patterns (aka accordion routing) when routing DRAM traces, due to serpentine "provoking orthogonal propagation"?

my understanding was that trombone routing causes more crosstalk and is only really used on low speed parallel buses. not sure what "provoking orthogonal propagation" means here in terms of SI impact.

tnt,
@tnt@chaos.social avatar

@gsuberland I don't read it as forcing a 4 layer stackup, just that if you use 4 layer board you should use SGPS. And that's what I would also recommend in the case, so what's wrong with it ?

timonsku, to random
@timonsku@mastodon.social avatar

Me: Why do I feel so groggy and tired?

oh..

tnt,
@tnt@chaos.social avatar

@timonsku Cold, wet and poisonous ... great environment :)

tnt,
@tnt@chaos.social avatar

@timonsku I am in the sdr room. Probably will exit in 30 min.

tnt,
@tnt@chaos.social avatar

@timonsku Nope, I'm home now 😅

ploum, to random
@ploum@mamot.fr avatar

When I graduated as an engineer, a friend told me "How do you feel entering a plane when you know that people like us build it".

I’ve worked several years in the automotive industry and saw how software was done there.

When the 737-MAX crashed, I entered into the rabbit-hole and built a really good understanding of the software issue.

The result are:

  • I don’t board a boeing plane anymore.
  • I avoid cars as much as I can.
  • I be sure to recommend cars with the less possible software.
tnt,
@tnt@chaos.social avatar

@ploum Watch a bunch of the Mentour Pilot videos .... then you won't want to board airbus either 😅

azonenberg, to random
@azonenberg@ioc.exchange avatar

Validating my new SDR receive setup with some known over-the-air test signals.

I used some random antennas I had lying around, didn't bother to VNA them so they might not be well matched to 2.4 GHz. But good enough for a quick test and I wanted to use an ISM band to keep the FCC happy even though I was doing a short, low power test (0 dBm / 1 mW input power to the antenna, sending 250ms of unmodulated carrier at each frequency within the sweep).

I'm seeing exactly what I expect, so everything seems good. Is ~80 dB of path loss typical in this kind of a setup (TX and RX maybe 8m apart using what appear to be simple dipole antennas) or is that a sign one or both of the antennas are badly matched to 2.4 GHz?

TX is Siglent SSG5060X-V vector signal generator with no modulation or external PA, just 0 dBm carrier right into the antenna. RX is my AntSDR running the UHD-compatible firmware.

Screenshot of gqrx listening to my AntSDR with the sweep pattern displayed on the waterfall view

tnt,
@tnt@chaos.social avatar

@azonenberg Did you calibrate the SDR somehow ? I'm not familiar with the AntSDR but most UHD device output is uncalibrated, the "dB" values you'd see in GQRX are just for relative measurement and have no meaning in absolute terms.

tnt, to random
@tnt@chaos.social avatar

Does anyone have resources / guides / examples on how to build OSHW zigbee end devices on linux ? Like without needing a windows VM to install some vendor dev environment ...

tnt,
@tnt@chaos.social avatar

@LaF0rge I don't have a target MCU/transceiver ... I'll take whatever can run whatever stack I find :)

I couldn't really find any OSS stack though ... there were some efforts but they look old/abandonned.

I'm fine if I need to link in a vendor blob as long as I can do the dev on linux with preferably a standard compiler. When doing wifi on ESP I also got to link the espressif blob for wifi ...

ploum, to random
@ploum@mamot.fr avatar

My blog is unavailable due to the hosting provider, , being heavily DDOSed. Like @Codeberg and also

The only way to protect yourself against such attacks seems, those day, use .

Which raises an interesting question: with his distributed infrastructure, what would prevent Cloudfare itself from DDOSing every non-customers ?

What makes Cloudfare business different from a criminal racket?

tnt,
@tnt@chaos.social avatar

@ploum

One common thing for a DDOS being very effective is to come from a lot of sources with nothing in common (like lots of compromised devices across the world). If all requests were coming from IP ranges assigned to cloudflare, it'd be pretty easy to block ...

But if you have any evidence that they initiate the DDOS, it would certainly be racket. But if they are not the source, that's a pretty big/obvious difference ...

tnt, to random
@tnt@chaos.social avatar

Anybody that can offer remote access to a system with an intel Arc GPU ? (to some linux install I can mess with ...).

I implemented some CL/GL interop between mesa and intel-compute-runtime but it's not working on A750 (and presumably other Arc) but I don't quite feel like buying another GPU ...

(But heh, if someone wants to send one my way that's also an option 😅 )

timonsku, to random
@timonsku@mastodon.social avatar

why-dsi-is-annoying.png

for some reason DSI displays specifically love to have under documented init sequences.

tnt,
@tnt@chaos.social avatar

@timonsku

Because some of those values would be meaningless without tons of deep technical info and you would never need to change them anyway.

Thing like bias current tweaks, wait times, .... stuff that was made configurable in the chip mostly just to be safe and not have to respin but that realistically is constant and shouldn't be changed.

At least for some of those ... others ... who knows ...

tnt,
@tnt@chaos.social avatar

@timonsku

Yeah, at least the "good" ones have the init sequence in the datasheet (with no explanation whatsoever just as a "write this to the display after reset").

The "bad" ones don't have it in the datasheet and the FAE will tell you if you're buying enough of them.

And if you're not buying enough ? Well you don't matter ...

gsuberland, to random
@gsuberland@chaos.social avatar

LTT: "we're going to slow down our production pipeline to make clearer, more accurate content"

LTT, a couple months later: "here are three videos about the same topic that we shot out of order and which, through the magic of editing, paradoxically reference each other with jumbled chronology"

tnt,
@tnt@chaos.social avatar

@gsuberland I'm still not sure WTF happened ....

Like did they change MB twice ? Because in today's video they say there was no apparent damage. And melted tube would have been very apparent.

So I guess the PSU and MB died, they change MB then a couple days later, disconnected the pump, caused melting tubes and waterleak killing the new MB they just put in ?

timonsku, (edited ) to random
@timonsku@mastodon.social avatar

Any favourite workflow tips/blog posts for working on Kernel driver (modules)?
Looking for nicer workflows for rapid iteration on single modules.

Executed on metal, emulation not an option here.

tnt,
@tnt@chaos.social avatar

@timonsku Oh, I always test stuff using an "Out-of-tree" style makefile first so I only build the module I'm working on.

tnt,
@tnt@chaos.social avatar

@timonsku I mostly write new stuff. For in tree patching I usually just do make modules or make drivers/xxx/yyy.ko and that usually goes pretty quick if you have already built the kernel ones.

But yeah, on occasion, I've ripped things out (when I didn't want to even build the kernel fully one and wanted to replace a module from stock kernel).

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • cubers
  • DreamBathrooms
  • thenastyranch
  • ngwrru68w68
  • Durango
  • magazineikmin
  • Youngstown
  • InstantRegret
  • rosin
  • slotface
  • tester
  • kavyap
  • ethstaker
  • megavids
  • osvaldo12
  • khanakhh
  • cisconetworking
  • tacticalgear
  • everett
  • mdbf
  • Leos
  • anitta
  • GTA5RPClips
  • normalnudes
  • modclub
  • provamag3
  • lostlight
  • All magazines