Les formations ont accès à toutes ces notes, les pondèrent et aboutissent à une forme d’hyper-précision où chaque élève se voit attribuer une moyenne à trois chiffres après la virgule. Cela signifie qu’on aura un élève avec une note de 12,256, un autre avec une moyenne de 12,257, et le but, c’est d’avoir le moins de doublon, d’avoir une note la plus discriminante, la plus individualisée possible, pour ensuite classer les personnes les unes derrière les autres.

De la même manière qu’on a un Règlement européen sur la protection des données, certes imparfait, il nous faudrait un règlement sur la protection des calculs.

L’autre enjeu très fort est celui de la participation des citoyens. [...] Dans les banques, l’assurance, le service public, il est anormal que les usagers ne soient pas représentés dans les discussions sur les effets du calcul.

Le numérique nous éloigne de ces enjeux, parce que la technique facilite le fait de décider pour d’autres

About Hayao Miyazaki to OpenAI allowing to generate image with the style of the studio Gibli.

Tante draws the following conclusion:

For the longest time OpenAI’s systems would try to block people from generating images in the style of certain artists. This was obviously for copyright reasons, the didn’t want to get sued (even more than they already are). Which is something they just changed very explicitly. You can now easily generate stuff in the style of Studio Ghibli and Sam Altman made his avatar on X-The Nazi Network a ghiblified version of himself.

This means OpenAI has to make more and more announcements so they look to the investors like they’re still cool and interesting. Any old garbage will do, like a literary fiction writer bot or something.

• https://pivot-to-ai.com/2025/03/17/openai-brings-you-statistically-average-literary-fiction/

We can reach the limit with AI though:
https://pivot-to-ai.com/2025/03/27/sam-altmans-studio-ghibli-memes-are-another-distraction-from-openais-money-troubles/

Bad #UI
experience: I was trying to send a friend some money via #CashApp
, but could not verify my identity. Here’s what their tech support found out:

“...from the documentation you have provided, we see that you have a legal one letter name. While we understand that is the name chosen and legally granted, we regret to inform you that we cannot proceed with the request at this time.”

I sent a response thanking them for figuring out the problem, and I wouldn’t try to get verified -- or use CashApp -- again.

I also sent them the URL for “Falsehoods Programmers Believe about Names”

Next time you're washing your hands next to somebody, cup your hands until the water overflows, then look at them dramatically and say "This water is getting out of hand".

Similar to previous shaares I stored

1. . If you look at 99% of all websites in the wild, everybody who worked on them seems to be allergic to semantics and shit. Headings are random levels, loosely based on font-size, Form Fields have no labels, links and buttons are divs.
2. The whole industry doesn't understand semantics and shit.

So when I teach about HTML I always start with the elements that are obviously interactive. I show them the multitude of UX layers of a link.

s and layers of UX that are added to a well considered form. I show them what happens on a phone when you use an input with a default text type instead of the proper type of email.

See the example between a span with an onclick-event and a proper link: the proper link opens a specialised context menu on right-click.

Why "Semantic HTML" fails? There’s no clear UX feature to point at.

First we need to get people exited about HTML by showing all the free yet complex layers of UX you get when you use the interactive elements properly.

You need a good idea of what UX is before you can understand things like the option to nagivate through the headings on the page with a screen reader.

I didn't know the radio button has an "indeterminate state" :o
I am curious about HTML form validation with HTML and clever CSS.

In other words, talks about what HTML does, and much less about what it means in theory. Let’s talk about user experience, and let’s stop talking about semantics and shit.

Brandolini's Law in picutre

Making the claim

I read that vaccine rewrite your DNA.
5 seconds. 0 source.

Debunking the claim

Okay... let's talk about transcription, translation, cell nuci....
6 hours. 15 citations.

Debunking take hours. Misinformation take seconds. Think before your share.

What is the opposite of "vibe coding"?

It could be someone who does everything like a professional who is never needed. "Overkill coding"?

"The Jedi use the cascade for knowledge and defense, never for attack."

...

"But only the Sith deal in absolutes!"

It was so simple, anyone who wanted to could create a free account [on Geocities, Tripod, FortuneCities, or Freeservers and build a website to share their hobbies and ideas.

The web was more about browsing and exploration.

It is worth remembering a website [...] can also be art. The web is also a creative and cultural space that [can be Free from convention defined by commercial product design and marketing].

If the commercial web is "industrial", you could say that the small web is "artisanal". One is not better than the other. They serve different needs and both can co-exist in an open web.

There is a lot of old good website, internet archive links and examples

To quote https://rhizome.org/editorial/2015/nov/30/oldweb-today/

Today's web browsers want to be invisible, merging with the visual environment of the desktop in an effort to convince users to treat "the cloud" as just an extension of their hard drive. In the 1990s, browser design took nearly the opposite approach, using iconography associated with travel to convey the feeling of going on a journey.

The web was "browsed", discovered by neighborhoods or [webrings], a circular collections of websites around a topic or a theme. To get orders and efficiency, one shared web directories such as the DMOZ open-directory project. One would simply go to a website and discover others on a page such as Retro Stuff

There are today gatekeepers keeping the attention only on some platforms: Facebook, LinkedIn. I would add TikTok now.

The commercial web emerged by the 2000s. Today's web is mostly commercial now. They invented words like "native advertising" and "sponsored content". They brought a completely different set of priorities: engage their audience, convert them and retain them for as long as possible.

Then came the product-oriented website and its sanitisation: it is polished, follows conventions and is optimised for efficiency. Modern websites are designed to direct user behavior towards certain goals: a purchase, a click, a share or a sign-up.

But the web remains and is also a creative space. the web is really a lot simpler than that. You really only need two things: a web host and HTML (and basic CSS for formatting).

If the commercial web is "industrial", you could say that the small web is "artisanal". One is not better than the other. They serve different needs and both can co-exist in an open web.

There are ways to discover it. Here some highlights from the past:

• Internet Archive, which, thanks to the 439 billion web pages saved since the mid 90s, lets you travel back in time and see how a website looked in the past.
• Restorativland, a "restored visual gallery of the archived Geocities sites, sorted by neighborhood".

and other highlights from the present:

• the search engine Wiby.me: for old-school, interesting and informative webpages, with a useful "surprise me" button that takes you to a random result.
• the Neocities.org is a modern web host that lets anyone create a basic website for free and be a part of a community where you can follow other webmasters.
• Curlie is "the largest human-edited directory of the Web. It is constructed and maintained by a passionate, global community of volunteer editors".

Some websites presented:

A fan-made site from 1998 about comedy duo Stan Laurel and Oliver Norville Hardy. Why? The siteinfo states "Why not!!!" - https://geocities.restorativland.org/Hollywood/Studio/5352/index.html

A student built the website Fractal Explorer to explore fractals. It includes very clear explanations, image galleries and step-by-step guided tours - https://web.archive.org/web/20020223163039fw_/http://www.geocities.com/fabioc/

The National Coca-Cola Bottle Clearing House lists all bootles since the 70s

NetHistory an informal history of BITNET (before internet) and the Internet - https://web.archive.org/web/20010516205238/http://nethistory.dumbentia.com/

Joan stark's ASCII Art Gallery - https://web.archive.org/web/20010420182629/http://www.geocities.com/spunk1111/indexjava.htm

European butterflies since 1998. Some pages such as Agrodiaetus ainsae have more informations than the dedicated wikipedia page.

The webtender since 1995, features over 6000 recipes, a handbook with information on bar glassware, tools, measurements and ingredients, a forum, a wiki and even lets you search for recipes based on what you have in your bar - https://www.webtender.com/info/

Disstant Skies is a fansite about the RPG game Crystalis - https://distantskies.neocities.org/

and you can build your own presence on the web!
Neocities.org allow to make a file and publish it with a free account. Spending hours browsing the websites made by passionates

There is also bear.blog dedicated for blogs.

The discussion follows on HackerNews https://news.ycombinator.com/item?id=23326329

Not in France.

3.5 millions fines after a ransomware attack because no security measure was set.

Even so, the problem won't disappear. There will be legal dispute about what is an ad and what is not.

If I share something on my blog, is it an ad?

The author clearly depicts traditional ads.

Eval scoped into a function has a default global scope: wtf.

and a lot of things

USA fuck yeah

Seit heute morgen sind ein Dutzend in Deutschland illegaler Casinos ohne Casino-Plattform und somit offline. Denn "the mill adventures" hat die illegale Instanz seiner Casino-Software abgeschaltet.

Wir können daraus lernen:

• Recherchen zu illegalen Casinos wirken.
• Wir können Casinos ihre Plattformen nehmen, was im Gegensatz zu den von der GGL geforderten Netzsperren, auch tatsächlich funktioniert.
• Ich bin den Betreibern der Casinos sehr nahe gekommen.

De même, la CJUE confirme mon interprétation du considérant 129 du RGPD. Une Autorité de Contrôle doit minimiser les efforts administratifs pesant sur les personnes concernées.

La CNIL et le Conseil d’État contreviennent donc aussi au RGPD en permettant à la CNIL de ne traiter qu’une partie du dossier et en reportant le reste et la vérification de la conformité à une éventuelle nouvelle plainte.

Screenshot: https://s3.imirhil.fr/firefish//fee5f857-a267-42c6-937a-5c78bf0d1bf9.webp

Pour les personnes qui veulent savoir combien de mémoire consomme un serveur XMPP avec 2 virtualhosts, 11 utilisateur·ices et 13 salons, la réponse est : ~ 165 MB de RAM

Encore moins gourmand, ce sont les serveurs IRC (cf: https://pleroma.lord.re/objects/4748377b-6551-4319-b2b3-95f714cb649e)