Show more

@ptrc @rail_ ...and force your IPv4 address to be behind CGNAT.

@grzeeesiek @rail_ eeej czapka czuwak trąci takim absurdem, że aż rozważam kupno xD

Been recording some profiles to try to explain _why_ my Rust translations of C programs tend to default to being faster.

Here's an unexpected culprit: strlen/strcmp.

The "natural" string representation in C is a char* pointing to a buffer that ends in a NUL (ASCII 0). To determine its length you have to crawl along it to find the NUL. To compare two strings, you need to do the same. In particular, because strcmp's return value isn't just a flag, but an integer giving the lexicographic relationship between the two strings ... it can't early exit if the lengths differ.

The "natural" string representation in Rust is a &str, which combines a pointer and a length. Finding the length is a constant-time near-no-op; comparing two strings for equality can trivially early-exit.

This might sound trivial, but profiles suggest it's a significant factor in wartcl being 4x faster than partcl across the board. In part, I suspect a lot of C programs do a lot of redundant strlen-computation unintentionally.

@rotopenguin @catsalad It was UDF, wasn't proprietary (though tools such as InCD that added support of relevant versions of it to Windows were) and it actually works on Linux: docs.kernel.org/cdrom/packet-w

@Blort @CalcProgrammer1 Yes. This needs to be fixed while processing (lens shading correction) and you need to profile your lens first to know what correction to apply - same with color matrix, chromatic aberration and barrel distortion. I've made a (not very high quality) lensfun profile for L5 some time ago: gitlab.com/dos1/glowup/-/blob/

See: wilson.bronger.org/lens_calibr

@rmader @Blort @CalcProgrammer1 I'm not worried about GLES3 arriving some day, @austriancoder's excellent work on etnaviv is progressing slowly but surely 😀

@pavel @usia @cas In case of , the front cam already does 10-bit, and you should be able to easily get it to work with lower res on the back cam. The problem is with higher bandwidth requirement of full res mode at 10-bit that AFAIK nobody really looked at and debugged so far. Help welcome.

while a few lines in bash already provided all the required functionality, the intercom control panel still needed some extra code for proper UX

@naugtur I think it's more a matter of fediverse's liberal boosting culture. It's easier for posts to break out of your followers' clique, and they don't have to get massively viral to do that.

I get it that my mobile Linux posts will have an easier time to find their audience here than anywhere else, but the difference is jarring. Even with posts about music or cats, here it actually feels like people read them. On Bluesky, they're being accidentally found by a single traveler who passes by 😄

@rail_ men will literally create several activitypub implementations instead of going to therapy

@rail_ @ptrc Also, those R1.1s are bidirectional and the city's current track building strategy is to cut costs and avoid building terminus loops, which requires such rolling stock.

Also, they're pretty similar to already operated Siemens Combino trams.

Also, they're likely to outlast not just 105Na, but also Tatra RT6N1 trams, the least reliable tram that rides on Poznań's tracks.

Also, Poznań is waiting for 30 new trams literally right now and prepares a tender to buy 30 more.

Also... 😭😭😭😭

@grumpygamer I'm using CommaFeed. While it can be self-hosted, you can just sign up at commafeed.com as well.

The recording of "postmarketOS - what is it and what's new?" has been released!

video.fosdem.org/2025/h2214/fo

Not in the video: three very nice people in the front immediately holding up their laptops as somebody in the audience asks about postmarketOS on other devices than phones 😀

Check out the other talks from the FOSS on Mobile Devices room too. I saw most of them and they were quite interesting!

Looking at my profiles, it's funny how dead Bluesky feels compared to Mastodon and the Fediverse. There's barely any interaction on my posts - and most of what's there comes from accounts bridged from Mastodon anyway. Hashtags don't seem to do anything.

It still does better than Twitter though 😂

@valpackett @postmarketOS I lucked out, as USD exchange rate was extremely convenient at the time and I had just finished a summer project and received money that almost exactly covered the Freerunner's price 😁

@camelCaseNick @postmarketOS I think it's just a matter of pmOS gaining some wider recognition these days, I don't think it's intentional. The only past project in this space that was similarly well-known was perhaps Openmoko, but its nature was a bit different.

SHR also targeted some Android phones, but it never really went far with them. Today's hardware ages much slower and is much more similar across brands than it used to 15 years ago, so it's easier for community efforts to progress.

@dos @postmarketOS yes I presume OpenEZX might have been the first attempt at open source software for the application processor of a smartphone. At least I didnt know any other such project back then. It sadly never went very far, as several key contributors including I got "distracted" with doing #openmoko - but that was also a good thing, of course

Show more
Librem Social

Librem Social is an opt-in public network. Messages are shared under Creative Commons BY-SA 4.0 license terms. Policy.

Stay safe. Please abide by our code of conduct.

(Source code)

image/svg+xml Librem Chat image/svg+xml