nabijaczleweli

наб

@nabijaczleweli

Dazed and confused, but trying to continue 🇵🇱/🏴󠁧󠁢󠁥󠁮󠁧󠁿/🇷🇺 ⚧ they

Maintains homie/hoe stasis. Store horizontally when not in use. Contains sulfites.

log in
nabijaczleweli
@nabijaczleweli

still really funny to me that for all the marketing and semantic poison what came out of the raspberry pi foundation the "using" UX sucks πß

there arent useful schematics or even vague board layouts or test points layouts. except sometimes, but not for your device, fuck you

the "documentation" sexion is somehow 30% marketing, 40% "wow look at our python bull shit!!!!!!!" (this is on the "hardware" page), 20% data laid out in the worst possible way, and 100% firefox ^F somehow doesn't fucking work

there's a whole "remote access" sexion, which....... ? im not opening to avoid permanent brain damage but it has subheadings upon subheadings on (translated) "systemctl enable ssh" and "ssh your-host" (for two platforms, neither of which target the raspberry pi hardware family, which????), "ssh-keygen", "mount -t nfs", and "ip a". bro. you're a notionally-educational hardware vendor. spit out cop cock and put documentation for your fucked up proprietary bootloader on your fucking documentation page; "Software PWM available on all pins" shut the fuck up. how is it possible that the UX of a vendor with 100`000`000 quid in yearly revenue is worse than some random lithuanians

see all
You must log in to comment.

in reply to @nabijaczleweli's post:

NireBryce
@NireBryce

I still see it as a deeply dishonest marketing scheme even if they didn't plan for that -- they expect a high baseline for computer knowledge, and tell people a raspi can work for them... and then SSH is sluggish because of bad PAM credentials on the recommended OS, or because your adapter does not provide quite the right number of watts... for a USB-C device.

login to reply
nabijaczleweli
@nabijaczleweli

idk about USB-C models, but mine is happy to violate the USB standard for power sucking, and indeed they "recommend" USB hosts with 2A current sourcing capability which fundamentally makes no sense; they should've make this like HDMI imo

as for PAM, idk. yescrypt takes like 430ms on this pi zero w which is Fine I Guess, a reference computer takes 30ms, and I don't think anyone expects better and how often do you run yescrypt (once a session? if you're logging in with a passphrase? but then you're logging in with a passphrase, so.); SSH with a RSA4096/SHA256 key takes 1.16s enter-to-prompt which isn't stellar but also Fine I Guess (I don't have other Wi-Fi hosts to log into so the 26ms to my router isn't really meaningful comparison-wise since i see like 25μs RTT there vs 5.6ms to the rpi); the figure jumps to 3.06s when sucket-activating sshd, but meh

the actual interactive usability ime i'd classify as "surprisingly good", to be honest, and I've been interactively debugging the deployment without wanting to off myself (but then i know better than to try to use a 500MHz uniprocessor system for anything processor-intensive, so); lord save you if you want to do heavy I/O though, but I don't, so idk. the default is still shockingly for /tmp on /, and i would probably be singing a different story if i didn't change that every time

but then all the points in the last paragraph are moot if this is advertised as a site and not an embedded/dial-out system (and, really, it's not suitable for the latter either because lol wifi, but a Normal User could potentially use it for that on a default-config bullseye (actual bullseye, not the Foundation's system; that sounds like hell on earth) image and not die probably; or this is a feldspar take), and it's fucking annoying to be used as the former because they're writing author/user fanfic in their documentation. ceterum, the raspberry pi foundation delenda est

really effective as a way of selling adapters and cheaply-rebranded electronic garbage though, you have to admit

login to reply
NireBryce
@NireBryce

nah, the PAM thing is that before recently it would cause slow ssh remote sessions because it was causing something else bundled to break, iirc

as for power, with the 3b it needs a constant 5.1v@2.5a, but nothing in the low voltage errors make it clear that even their branded ones cannot hold that constant.

the 4 needs 5/5.1v@3a, which chargers CAN provide

but both depend on cable length and thickness to actually maintain that -- my USB C 2.0 charging cables cannot.

and their solution is bad tutorials and not hardware revisions, if they haven't changed it in the last year.

the low voltage events used to only be shown on the physically connected screen and didn't show up in logs until a few years ago, too.

BUT.

they're cheap so I'm not sure i expect more

login to reply
nabijaczleweli
@nabijaczleweli

lmfao "we sell devices with psychotic power input specs and our own special chargers with the same output phy; they're unrelated tho" really is amazing. i wouldnt request a better voltage either, sounds like an incredible profit incentive for the foundation and its business partners, smart choice :v

but yeah idk, if something's cheap i expect it to have more documentation (or, indeed, more working) than marketing, but maybe thats unreasonably skewed

login to reply
Pinned Tags