Eḿbed SQL examples in a web page. It uses a browser-compatible compiled sqlite under the hood .
- Users tend to use a list of all the links or headings on the page instead of the tab key! The user could then jump directly to the heading or link that sounded like what they were looking for.
- Navigating to the homepage from an internal page: all of our interviewees searched for a link being announced as "Home"
- To navigate to a certain page, users opened a list of all the links on a page. it was common to type the letter on the keyboard to search for the available links starting with the first letter of what they would expect the page name to start with. A "Get in touch!" link is likely to not be found
- idea here: we need a standard about common pages! (Contact, Pricing, etc...)
- None of the interviewees enlarged the window to take the full screen. This means some users will experience the mobile layout and behavior of the menu navigation, rather than the desktop version we intended. So a computer does not mean a desktop layout
- None used skip links because they are less efficient than their other methods. Moreover, the skip link is sometime broken as the keyboard focus does not shift!
Use a web UI to ensure compatibility.
Operations are made through system calls of Node, Bun or deno or the git cli command as the FileSystem API is not widespread.
A lot of so useful projects for web development!
Check them here: https://unjs.io/
Mdash is a new kind of UI library. It's 100% standards-based and it's tiny.
Mdash components are comprised of standard HTML, custom HTML, and Custom Elements. As such, Mdash works with any framework (or no framework) and works with all types of web projects: SSR, SPA, PWA, static site, and even popular email clients.
Only 6.8kB. It uses a minimal appearance but adds components that are used everywhere such as switch buttons-
It is possible to display high, medium and low screen sizes by popularity
Useful for flat design
La solution, la seule, permettant de traiter le mal plutôt que ses symptômes est pourtant connue. Elle consiste, je vous en ai déjà parlé, à créer un index indépendant du web. C'est à dire à réinstaller durablement dans l'espace public une "ressource", un "bien commun" dont l'exploitation par des acteurs privés à fini par installer des usages essentiellement privatifs, là où pourtant l'ensemble des caractéristiques techniques (son architecture) et des usages (ces 'contenus' qui 'font' le web) de ce "bien" relèvent, par nature, de la sphère publique commune.
Aussi, développé dans l'article:
Il n'y a pas de barrière technique ou juridique au déploiement de cet index indépendant.
Avec de bons débuts d'aggrégation de données: Hathi Trust pour les livres, Internet Archive pour les sites webs.
Les technologies open-source sont aussi présentes, de même que la recherche au niveau européen. Les services quotidiens ont tous ou presque un équivalent open-source.
One project to unify them all
You simply substract the side distance [y-axis] from the larger corner radius.
That way you end up with the right one for the smaller shape.
Seen on Twitter but I could not find the author again
Postuler que la publicité est un business « honnête » revient à dire que ces milliards d’euros ne seraient pas dépensés sans elle. Et donc revient à démontrer que la publicité appauvrit la population d’une manière démentielle en entrainant une surconsommation qui détruit littéralement la planète.
Lorsque vous achetez des publicités chez un de ces géants, tout ce qu’ils vous offrent en échange sont… des chiffres qu’ils inventent.
Parce que ce sont des boîtes noires et que personne ne peut vérifier le nombre d'affichage, etc...
Le géant Procter&Gamble a d’ailleurs annoncé avoir coupé complètement tous ses budgets publicitaires pour les réseaux sociaux et n’avoir perçu absolument aucune différence dans les ventes.
Avec mes 2500 followers sur Facebook, j’avais découvert à l’époque que chaque message touchait en moyenne 1%, mais que, en payant, je pouvais monter à 5% voire 10% de « ma communauté ». Facebook me faisait croire que j’atteignais un public alors qu’en réalité, je devais payer pour contacter moins de gens que si je leur avais conseillé de s’abonner par mail ou RSS.
En fait, de manière générale, tout fournisseur qui prétend vous vendre l’observation de métriques décidées par lui, mesurées par lui et optimisées par lui est par essence un escroc.
Briser l’incroyable puissance de ces monopoles morbides ne passe pas par l’utilisation d’alternatives ou de succédanés, mais par la réalisation profonde que nous n’avons tout simplement pas besoin d’eux.
Clever and efficient
There is an alternative to the SPA model: a new model with HTML only. It uses modern DOM-swapping interactivity libraries like Hotwire, HTMX, and Unpoly and... something that returns HTML as backend :)
Since version 7.0 has Ruby on Rails Hotwire backed-in!
-
SSR and Jamstack
-
Active Memory Caching
In summary, if you want to increase the performance of your application, you can use server caches to speed up your APIs, but if you want to persist your app state, you should use the local storage cache.
- Data Event Sourcing
Useful for real-time applications. Connections are made with Websockets.
4.a Prefetching
Pros: Prefetching waits until the browser’s network is idle and is no longer in use and will stop when you trigger usage by clicking a link or triggering a lazy loading function.
Pros: Prefetching caches data within the browser, making page transitions faster when redirecting to a link.
4.b Lazy Loading
Lazy loading can only help you delay downloading resources and doesn’t make your resources smaller and more cost-efficient.
- Resumability
Essentially, Resumability uses the server to do the heavy lifting and then gives the client a minimal amount of JavaScript to execute via serialization.
A captcha with proof of work (~2s). This can definitely be better for every user as they do nothing. It's free software, privacy-friendly, and it doesn't use IP addresses (so users won't suffer what reCaptcha or CloudFlare impose).
They have a GitHub organization: https://github.com/mCaptcha
A way to avoid to pull with a request at regular intervals:
- the user subscribe to events similar to JS Events
- the server send them at its pace
I find it hard too :/
Rule of thumbs:
- Show errors immediately if issues are severe
- Late validation is almost always better
- Validate empty fields only on submit
- Reward early, punish late: if a user edits an erroneous field, we should be validating immediately, removing the error message and confirming that the mistake has been fixed as soon as possible (reward early). However, if the input was valid already and it is being edited, it’s probably better to wait until the user moves to the next input field and flag the errors then (punish late).
- Validate after a copy-paste
- Allow users to override inline validation
- Just in time validation
- For short forms, consider validation on submit only. For complex forms, use the task list pattern
Examples of original icon hover effects 👍
Examples of original button effects 👍