AdrianTheFrog

Check how large your photos library is on your computer. Now wouldn't it be nice if it was 40% smaller?

jpeg xl lossless is around 50% smaller than pngs on average, which is a huge difference

https://siipo.la/blog/whats-the-best-lossless-image-format-comparing-png-webp-avif-and-jpeg-xl

JPEG XL in lossless mode actually gives around 50% smaller file sizes than PNG

its royalty free and has an open source implementation, what more could you want?

I have a reverb g2 and the only linux projects that support it don't work with the controllers

Microsoft is going to kill WMR in the next couple years so I'll probably switch then, luckily it seems like people are working on controller support already so hopefully that's stable by november 2026

Mozilla could definitely be putting their development time into the areas that the browser is actually behind in

I use firefox, I mostly like it, but it still doesn't support chromium style tab groups (no, that one extension is not similar), and its webgpu implementation also doesn't work on most websites more than a year after Google made their version available by default

Read the response from the thing that read the Arch wiki

I got some used chromebooks on Ebay for $40 each (3855u, 4gb ram, 32gb ssd), I would recommend them if you don't have any money to spend on a laptop. It's not going to be running anything super demanding, but its shocking how much it can do.

Some things I have run on it and had a decent experience: Blender, FreeCAD, Portal, TMNF through proton, Celeste, Minecraft Java, MuseScore

yea, thinkpads aren't the only laptops that can be bought used

I've been looking at the paper, some things about it:

• the paper and article are from 2021
• the model needs to be able to use optional data from age, family history, etc, but not be reliant on it
• it needs to combine information from multiple views
• it predicts risk for each year in the next 5 years
• it has to produce consistent results with different sensors and diverse patients
• its not the first model to do this, and it is more accurate than previous methods

I actually thought that subplot was one of the more interesting ones