Claude says: JPEG XL beats JPEG 2000 on essentially every practical axis:

**Compression efficiency** — JXL achieves better quality at smaller file sizes than JP2, both for lossy and lossless. At equivalent visual quality, JXL files are typically 20–60% smaller.

**Lossless JPEG recompression** — JXL can losslessly re-encode existing JPEG files ~20% smaller, then decode back to the *identical* original JPEG bitstream. JP2 has no equivalent capability.

**Speed** — JP2 (especially with its wavelet codec) is notoriously slow to encode and decode. JXL encodes faster and decodes fast enough for practical real-time use, with a well-optimized reference decoder (`libjxl`).

**Progressive decoding** — Both support it, but JXL's progressive rendering is more sophisticated, allowing a useful low-resolution preview very early in the stream, which matters for web delivery.

**Feature set** — JXL supports HDR, wide gamut, alpha, animation, layers, and up to 32-bit/channel depth natively. JP2 has some of this but with far less ecosystem support.

**Royalties and patents** — JP2 has a murky patent history that scared off adoption. JXL is royalty-free with a clean IP situation.

**Browser/ecosystem support** — JP2 was never adopted by Chrome or Firefox (only Safari). JXL has broad support in Safari, Firefox, and Chrome (behind a flag, then natively). The web ecosystem simply rejected JP2.

**Encoder/decoder ecosystem** — `libjxl` is actively maintained. JP2 tooling is fragmented (OpenJPEG, Kakadu, etc.) and often requires proprietary libs for best quality.

The short version: JP2 was designed for print/archival workflows (DICOM, digital cinema via JPEG 2000's cinema profile) and never translated well to the web. JXL was designed from the start to replace JPEG on the web while also serving archival use cases — and it does both better.

Once FireFox is outlawed, only criminals will use FireFox.

A competitor to Stripe and Adyen worth considering is Mollie. Adyen is focused on large scale enterprises while Mollie wants to work with any size entity.

Fun fact: both Adyen and Mollie are based in Amsterdam, walking distance from each other, and there's probably a joke or two for the slashdots in there somewhere.

Another Drupal developer here, with some experience working with the Feds. Most government websites as I am aware of are Drupal websites.

Writing as a developer, I can tell you we are not allowed any access to live systems, which is good. That allows us to work in our sandboxes and break things before we commit to the GIT repository branch we're developing to eventually be merged into the main branch and released one day. In other words, the only connectivity we're allowed is uploading to the git server.

In a perfect world, we'd have resources including time to scan everything for everything prior to our GIT commit to the repo. I hear ai (and mythos) are a thing.

I'm just sayin'.

That said the NPM vendor directory is generally excluded by GIT, because none of that stuff belongs in the repo because it can easily be rebuilt on the staging server that gets tested prior to going into production. And the admins upstream aren't supposed to trust anything, period. In a perfect world.

1984

This is what Putin, Xi, Kim Jong Un want him to do. Trump got the memo and is doing as is expected of him.

Apparently Mythos is so advanced Anthropic is afraid to release it, (and there's no government regulations on AI). Anthropic *has* made it available for security researchers so they can use Mythos to look for bugs in widely available open-source code so that code can be fixed before Mythos is relased. Mythos has found thousands of zero-days.

https://www.techradar.com/pro/...

I told the person that replied to me they were a God, if that keyword helps. Muchos gracias to anyone in advance that even considers making an effort to assist me. Nerd.

That was so long ago as evidenced by the timestamp! No way! Am I that old already?! I think its awesome you managed that citation! 17 years? No way!

BTW: Serious question because someone seriously wise and learned long ago replied to a comment I once made with EXCELLENT advise about LED's for home lighting, but I can longer find the comments, and I'm certain it was in the last 6 years or so, but slashdot seems to have comment retention limits -- and I trusted slashdot, and I was stupid and lazy for doing so. Does anyone know how I can find an old comment/reply of mine, (via google, etc?) with regards to LED home lighting?

Solid point, but wouldn't that somehow make the entire exercise moot? Because that wouldn't result in a truly random and acceptable number? I suppose you can add a computer to translate, somehow, between Base 6 and Base 10, but I don't think so, (and I'm way out of my league here. A man's got to know his limitations).

I'll opine even more so. Its typical.

I can recollect a video of a machine built for a casino as I recall, but I might be wrong. Its purpose was to constantly generate truly random numbers.

Essentially a very large number of quality dice was automatically rolled and finally fell into a row, a horizontal tray actually, to define the result of the dice. The tray itself was part of a conveyor belt and the the dice traveled vertically upward, and each row of dice was captured with an OCR camera to convert into a random number. But I can't find the video on YouTube.

Typical Slashdot first post comment with regards to an article about generating reliable randomness.

Goed gedaan hoor!

Steve Jobs was totally irrelevant without Woz. Steve was a fake-it-till-you-make-it kind of guy. A lot of Trump qualities actually. Woz made Steve.