Jon Sneyers
@sneyers.info
JPEG XL and image compression in general, socialism/marxism/trotskyism, miscellaneous geekery.
I'm not working on it myself so I'm not the best person to ask, but I would expect something like the end of the year, maybe?
You can follow the development here: github.com/libjxl/jxl-rs
You can follow the development here: github.com/libjxl/jxl-rs
GitHub - libjxl/jxl-rs
Contribute to libjxl/jxl-rs development by creating an account on GitHub.
github.com
August 7, 2025 at 6:38 PM
I'm not working on it myself so I'm not the best person to ask, but I would expect something like the end of the year, maybe?
You can follow the development here: github.com/libjxl/jxl-rs
You can follow the development here: github.com/libjxl/jxl-rs
What exactly do you need?
The container is optional in jxl. If you want to create some examples with container, you can use cjxl --container=1. Or just use an input image with Exif or XMP metadata, which will force using the container.
The container is optional in jxl. If you want to create some examples with container, you can use cjxl --container=1. Or just use an input image with Exif or XMP metadata, which will force using the container.
August 3, 2025 at 8:00 AM
What exactly do you need?
The container is optional in jxl. If you want to create some examples with container, you can use cjxl --container=1. Or just use an input image with Exif or XMP metadata, which will force using the container.
The container is optional in jxl. If you want to create some examples with container, you can use cjxl --container=1. Or just use an input image with Exif or XMP metadata, which will force using the container.
See also: docs.google.com/document/d/1...
Why ISO should make all standards Publicly Available
Why ISO should make all standards Publicly Available Background ISO and its national affiliates derive income from the sale of ISO publications ISO sold 6.6 million CHF of publications in 2019 and re...
docs.google.com
July 15, 2025 at 7:57 PM
See also: docs.google.com/document/d/1...
Lossless float is not something libjxl has been particularly tuned for. More accurately: besides checking that it works, not much effort has been spent on getting it fast or well-compressed. I don't know how much room for improvement there is.
July 15, 2025 at 7:54 PM
Lossless float is not something libjxl has been particularly tuned for. More accurately: besides checking that it works, not much effort has been spent on getting it fast or well-compressed. I don't know how much room for improvement there is.
Ook weer zo'n clickbait titel. Wie is gewonnen???
July 15, 2025 at 8:20 AM
Ook weer zo'n clickbait titel. Wie is gewonnen???
Bij elke "optimalisatie" is de vraag of hetgeen geoptimaliseerd wordt eigenlijk wel datgene is waar de wereld beter van wordt. Vaak is dat niet het geval. Zie bijvoorbeeld ook "fiscale optimalisatie" of in de academische wereld: epsilon-incrementele publicaties om publicatiecijfers te optimaliseren.
July 14, 2025 at 8:08 PM
Bij elke "optimalisatie" is de vraag of hetgeen geoptimaliseerd wordt eigenlijk wel datgene is waar de wereld beter van wordt. Vaak is dat niet het geval. Zie bijvoorbeeld ook "fiscale optimalisatie" of in de academische wereld: epsilon-incrementele publicaties om publicatiecijfers te optimaliseren.
In this case, I think the main question is why academics would be tempted to cheat (either as author or as reviewer), and how we could reduce/remove those incentives.
July 6, 2025 at 5:07 PM
In this case, I think the main question is why academics would be tempted to cheat (either as author or as reviewer), and how we could reduce/remove those incentives.
That was not my point, but I don't agree it's always a bad argument. Sometimes things are broken beyond repair and it's better to replace them with something better than to keep applying band-aids and pretending things will be fine.
July 6, 2025 at 4:49 PM
That was not my point, but I don't agree it's always a bad argument. Sometimes things are broken beyond repair and it's better to replace them with something better than to keep applying band-aids and pretending things will be fine.
IMHO, unreasonable publication pressure — mostly caused by research funding guided by neoliberal principles and the widespread conflation of "objective and fair criteria" with "quantitative metrics" — has already seriously compromised peer review and academic publishing in general.
July 6, 2025 at 11:57 AM
IMHO, unreasonable publication pressure — mostly caused by research funding guided by neoliberal principles and the widespread conflation of "objective and fair criteria" with "quantitative metrics" — has already seriously compromised peer review and academic publishing in general.
In Belgium we do it this way: en.m.wikipedia.org/wiki/Consume...
Consumer price index (Belgium) - Wikipedia
en.m.wikipedia.org
July 6, 2025 at 5:57 AM
In Belgium we do it this way: en.m.wikipedia.org/wiki/Consume...
TBH, I prefer returning a bool (where true is OK and false is failure), and using some other way to signal which error condition it is, if there's a need for that. Usually just OK or not OK suffices, and calling the function in an if-else or if(!foo), without a return code variable, is convenient.
July 5, 2025 at 11:38 AM
TBH, I prefer returning a bool (where true is OK and false is failure), and using some other way to signal which error condition it is, if there's a need for that. Usually just OK or not OK suffices, and calling the function in an if-else or if(!foo), without a return code variable, is convenient.