Constructing the Word's First JPEG XL MD5 Hash Quine

throw0101d

One of my saved HN comments from @Retr0id:

---

    Beware of having too-small fingerprint hashes though, or not checking enough of the digits.

    $ echo -n retr0id_662d970782071aa7a038dce6 | sha256sum
    307e0e71a409d2bf67e76c676d81bd0ff87ee228cd8f991714589d0564e6ea9a  -
    
    $ echo -n retr0id_430d19a6c51814d895666635 | sha256sum
    307e0e71a4098e7fb7d72c86cd041a006181c6d8e29882b581d69d0564e6ea9a  -

---

* https://news.ycombinator.com/item?id=38668893

Retr0id

I later wrote an article explaining how I computed that: https://www.da.vidbuchanan.co.uk/blog/colliding-secure-hashe...

(Doing it the "obvious" way would involve infeasible amounts of storage space)

wizzwizz4

I'd be interested in seeing even your messy non-working code for that.

zygentoma

This is soo cool! Especially the prediction machine stuff. I had no clue this was possible.

And also at the same time a good reminder for everyone to find a browser that supports JPEG XL. I wonder if that was part of the reason to do this. :)

embedding-shape

> And also at the same time a good reminder for everyone to find a browser that supports JPEG XL

That's probably furthest down on my list of features I look for in browser, where the top two are "Not run by a for-profit company living on extracting data from users" and "Can have tabs vertically in sidebar in a tree-based structured format".

MrAlex94

Waterfox might be what you’re after?

- Supports JXL out of the box (including support for alpha transparency and animations)

- Vertical tabs with optional tree tabs (hired the original tree style tab developer to implement the feature)

- For profit, but I don’t want your data, collect it or use it to earn a living (telemetry/analytics/experiments disabled at build time and alongside a fair few patches on top to make sure external connections are limited to what’s necessary)

Sidebar, I’m the developer of Waterfox

embedding-shape

Firefox (with minor changes + addons) is what I use today, works well for what I care about. Thanks for the recommendation though!

While you're here, last time I came across your website (and it seems like it looks the same currently), I noticed that your browser comparison is not including Firefox, which is what you've forked from (as far as I can tell at least, it isn't made clear by the landing page actually, but the UI and name makes it obvious), which feels like it's a bit misleading almost intentionally.

progbits

Also one which doesn't add new image decoder with built-in VM that is rawdogged in C, like Safari.

Chrome and Firefox are making a very reasonable decision to wait for a memory safe decoder.

F3nd0

Chrome's involvement in the past few years has (until very recently) been anything but reasonable.

That said, have any of them subjected WebP or AVIF to the same strict requirements, or should we reserve those only for less complex codecs actually designed with images in mind?

wild_pointer

In the era of LLM-generated content, such a high-quality writeup is a breath of fresh air. Well done!

bigbuppo

Nobody was this excited about WEBP.

QuaternionsBhop

Webp was not as exciting. JpegXL has cool features like 20% improved lossless jpeg recompression and progressive decoding. Not to mention all the cool stuff used in the writeup like implementing a font in the prediction engine.

bigbuppo

And best of all, it's not a google product, so it can be trusted.

smokel

Shouldn't that be "World"?

b16m

[flagged]

blenderob

This is an automated bot incorrectly claiming posts are dupe: https://news.ycombinator.com/threads?id=b16m

Looks like it was banned: https://news.ycombinator.com/item?id=46188453

How is it still posting?

embedding-shape

> How is it still posting?

Banned accounts show up as [dead] as soon as they post, so most users don't see those comments, only those with showdead enabled. Commonly referred to as "shadowbanned" but usually operators also don't tell shadowbanned users they've been shadowbanned.

blenderob

This one didn't show up as "[dead]" for some reason.

poly2it

Looking at the posts, there are many formatting differences added across the comments. Either the account owner are posting the comments themselves, or they are manually modifying the bot script to add mistakes over time.

jsnell

It's fine to have a few duplicate submissions for articles that did not get any attention originally.