omoikane

> humans have a high ability to ignore negatives in things they're a fan of, including (and often especially) their own work

I thought this is due to the "backfire effect", because humans perceive criticisms of things they like as threats.

https://theoatmeal.com/comics/believe

This is a wonderful paper that includes many optimization tricks that are generally applicable, including various optimizations that they retrofitted into lld. But I thought there was one meta optimization that was especially interesting (from page 2):

> A new linker's early adopters knowingly accept such differences, which is one reason why drastic improvements tend to come from new linkers rather than existing ones

This is true of many things, not just linkers. I think this is encouraging for people to (re)try things that have already been done, because newcomers are allowed to take risks that established things can't take.

Bluesky didn't require phone numbers back when it was invite only, and then there was a period of time after they allowed everyone to create accounts where users need a phone number to verify. That requirement might have been replaced by a captcha according to:

https://github.com/bluesky-social/social-app/issues/2762

But I didn't try to create an account recently, maybe they went back to requiring phone numbers.

Alan Paeth's three-shear rotation is one of my favorite algorithms, especially because it's perfectly suited for rotating pixel art. Because shearing preserves all the original pixels, it even works with dithered images, unlike sampling-based rotation schemes.

It's also fairly simple to implement. I have implemented a version that rotates ASCII art, and that code made it to Ruby's list of sample programs:

https://github.com/ruby/ruby/tree/master/sample/trick2025/01...

I like Existential Comic's explanation of the simulacrum concept, using football/soccer injuries as an example:

https://existentialcomics.com/comic/476

I like how a website for tracking balloons escalated into critical infrastructure. Reminds me of 99 Red Balloons.

https://en.wikipedia.org/wiki/99_Luftballons

It's iOS only, relying on iOS to exclude isSecureTextEntry things from screenshots:

https://github.com/mozzius/expo-privacy-sensitive/blob/main/...

The 1% Fighter

> while (has_watermark(text)) slightly_rewrite...

I understand what you are trying to say but I am not sure any watermark detection API would definitively return a true/false answer, I would have expected something more like a numeric confidence value. I am also not sure if the API would be deterministic.

Maybe Bluesky is not quite the Twitter rival it set out to be, but it's got a better shot than Mastodon, according to the user count graphs here:

https://bsky.jazco.dev/stats

https://mastodon-analytics.com/

Bluesky has a higher total user count. I think the number of active users might be higher too, but I am not sure what's mastodon's definition of an active user for the second link above. Anecdotally, Bluesky feels more lively for me because that seem to be the default destination for Twitter refugees.

Related, Oscar Toledo also made a tiny Klondike Solitaire program for IOCCC, I wonder if it was inspired by this one:

https://nanochess.org/klondike_in_c.html

https://news.ycombinator.com/item?id=48450024 - Klondike Solitaire game for curses in 5k of C (2026-06-08, 17 comments)

> Sanjay (who just joined Twitter)

Clarification: this comment is saying Sanjay Ghemawat joined Twitter as a user recently (new account @Sanjay_Ghemawat as of July 2026), as opposed to Sanjay working for Twitter the company.

It's probably not that different from rangefinder style cameras[1]. I don't think the problem with this one is with its transparent viewfinder, it's more that its output quality just aren't that great according to someone who tried it[2]. Seems like a neat toy at $45 though.

[1] https://en.wikipedia.org/wiki/Rangefinder_camera

[2] https://www.reddit.com/r/Godox/comments/1urlp8u/first_test_s...

I am glad they also cloned the snake game.

https://github.com/8bitbubsy/ft2-clone/blob/master/src/ft2_n...

I also liked how the "Be" in "to Be continued" resembles BeOS.

It sounds like you are looking for Happy Vertical People Transporters with Genuine People Personalities.

> Destination Dispatch [...] are in general worse

I wonder if this is an artifact of how the author used random destinations. I worked in a building that used Destination Dispatch, and the common travel pattern seemed to be:

- Everyone who is not on the ground floor generally want to go to the ground floor.

- People who are on the ground floor generally travel in large groups to the same destination.

This happens because people who worked on the same floor often leave for lunch at the same time, and return at the same time to the same floor. Destination Dispatch helps in this case because it's batching large groups of people with the same destination.

I thought this article was going to be about graphical imagery as opposed to textual elements. I was specifically thinking about certain levels in Pragmata where they had this somewhat broken and inconsistent look that was meant to resemble AI-generated imagery, but was actually all done manually:

https://www.ign.com/articles/pragmata-contains-a-stage-thats...

The contest ended a few days ago, some teams have already published their writeups. Usage of LLM-based tools appear to have mixed results. For example, I think #23 place claimed to have used nothing but AI[1], while the #24 place seems to be mostly manual code[2]. These are very good results considering they solved all problems, only 40 out of ~268 teams managed to do that[3].

[1] https://blog.greg.technology/2026/07/26/doing-okayish-with-a...

[2] http://www.sawicki.us/icfp/2026/

[3] https://icfpcontest2026.com/standings

I went through the 3 days without using any AI, working mostly in the language designed by the judges. I placed somewhere in the middle.

In that sense, this game is similar to Mosaic[1] where the clues are all initially visible, except in Mosaic the count includes all 9 cells since black squares can be placed on the numbers themselves.

[1] https://www.chiark.greenend.org.uk/~sgtatham/puzzles/js/mosa...

I think the number actually runs out at 69999:

https://celestrak.org/NORAD/documentation/gp-data-formats.ph...

In the linked bluesky thread, they have been estimating when they will need 6 digit numbers, and seems like it was a few days earlier than their previous prediction.

https://bsky.app/profile/tskelso.bsky.social/post/3mlg2dpuqd...

If you have control over the HTTP server, a different technique is to set "refresh" in the response header[1], which makes the client fetch new animation frames periodically. This is the technique used in a 2013 IOCCC entry to show a continuously updated clock that's rendered as PNG[2].

[1] https://developer.mozilla.org/en-US/docs/Web/HTTP/Reference/...

[2] https://www.ioccc.org/2013/mills/index.html

The "Refresh" header is encoded backwards on this line:

https://github.com/ioccc-src/winner/blob/619f554bbdb19e5003a...

Related, a collection of tiny chess programs by Oscar Toledo:

https://nanochess.org/chess.html

This is an interesting tool and maybe it works for the songs you need, but for classical guitar songs where printed scores are generally available, I would rather just buy those from some reputable publisher (e.g. Zen-On or Doremi).

It has happened quite a few times where I compare tabs in youtube videos with printed scores, and the youtube version turned out to be unique arrangements. It's the same with PDFs on certain sites, especially those that accept user contributed content.

I would rather pay money to get printed scores than to find out later that I have been practicing the wrong thing.

I hope this eventually ends up in https://coveryourtracks.eff.org/ so I can see how unique my math functions are relative to a larger population.

Not sure about <blink>, but I sampled a few random sites from their gallery[1] and all of them have <marquee>. <marquee> is deprecated[2] and no longer scrolls in Firefox, but still works in Chrome.

[1] https://neocities.org/browse

[2] https://developer.mozilla.org/en-US/docs/Web/HTML/Reference/...

Looks like they illustrate the dual nature of how line intersections in the grid map to equilateral polygons in the tiles. It wasn't obvious to me until I read the "how it works" section.

https://github.com/aatishb/patterncollider#how-it-works

I don't think most browsers enable transliteration out of the box.

Lynx does[1], but it does that using iconv, which I think doesn't use UTS #35 transliteration rules.

[1] https://lynx.invisible-island.net/lynx2.9.3/CHANGES.html#:~:...

I like how retirement progress is tracked in a "freedom bar". I have heard that amount of money being described with a different f-word.

I can't find any restriction on the submission language, maybe that should be explicitly spelled out. 1000 words in German would carry a different amount of information from 1000 words in English or 1000 characters in Japanese.