(In context of web gardening)

1. Many hypertexts do not require elaborate navigational apparatus.
2. Rigid hypertext structure is costly.
3. The shortest path is not always the best.
4. Large hypertexts and Web sites must often contain both parks and gardens.
5. Use punctuation sparingly; unwanted interruptions are tiresome and intrusive.
6. The boundaries of parks should be especially clear, lest readers see them as mere wilderness.
7. Rigid structure makes a large hypertext seem smaller. Complex and intricate structure makes a small hypertext seem larger, inviting deeper and more thoughtful exploration.

Highways are judged by efficiency: distance, cost, safety, and time. Garden paths play a different role; they lead us through the best routes, not the shortest. They may bend to pace our journey, curving here to reveal a view, twisting there to lead us through a shady grove or a sunny clearing.

Rigid hypertext is streetscape and corporate office: simple, orderly, unsurprising. We may find the scale impressive, we admire the richness of materials, but we soon tire of the repetitive view. We enter to get something we need: once our task is done we are unlikely to linger. We know what to expect, and we rarely receive anything more.

Make an OSS version of it for the web: why not building on the web for the web?

At times, wilderness is exactly what readers want: a rich collection of resources and links. At times, rigid formality suits readers perfectly, providing precisely the information they want, no more and no less. Indeed, individual hypertexts and Web sites may contain sections that tend toward each extreme.

Often, however, designers should strive for the comfort, interest, and habitability of parks and gardens: places that invite visitors to remain, and that are designed to engage and delight them, to invite them to linger, to explore, and to reflect.

A website about web, links and hypertext... well it is a garden when someone can lost itself a bit.

Baseline is a new term that defines a feature broad adoption. There are two stages:

1. The new feature is available in all last versions of the browsers
2. The new feature is widely available after 30 months after the first stage

We can then speak of baseline 2023 that groups all features usable in the considered browsers.

Limitations:

• the browsers considered are Safari, Firefox, Chrome and Edge. The rest is unknown even if they are listed on caniuse.com
• it does not take screen readers and accessibility tools in consideration as part of "availability"

I like the model: Subject - Predicate - Object

Such as: Xerxes - Parent - Brook.

It can be modeled as such:

parent(xerces, brooke).
parent(brooke, damocles).

Note of Tim Berbers-Lee.

Put a frame around "engineering" and " ownership"

The overwhelming motivation behind it seemed to be “I made something, here it is”. Sharing things for the sake of showing them to the world. Somebody had created something, then put it online so you could see it. Visit their website (wait for the dial-up to finish), and it’s yours.

Large companies find HTML & CSS frustrating “at scale” because the web is a fundamentally anti-capitalist mashup art experiment, designed to give consumers all the power. — Mia, with valuable secrets 🤫 (@TerribleMia) November 24, 2019

You can stand out of the crowd by simply treating the web platform as what it is: a way to deliver content to people.

The best growth hack is still to build something people enjoy, and then attach no strings to it. You’d be surprised how far that can get you.

Metaphosr of the web as Dark Forest, cozy web, dark web, corporate web, etc...

And there are The Towns, the IndiWeb or the small web. There is the fun !

Let's create a holiday card generator by learning how to get access to a user's webcam and compose a screenshot.

Because the federation is disturbing for casual users, it seems normal to

remind ourselves of what social media used to be: a way to connect around shared interests, talk to friends, and discover new content. No grifts, no viral fame, no drama.

The barrier to be free on the web is highly correlated to the level of expertise.

You could loosely map some of them by how easy it is to get started if you have no technical knowledge. [...] The more independence a technology gives you, the higher its barrier for adoption.

Owning, control and independance on the web should be just as easy as signing up for a cellphone plan.

An acronym standing for "Publish (on your) Own Site, Syndicate Elsewhere".

It seems great :D

Great explanation of the rel attributes that provide:

• information to the resource linked
• behaviors while navigating

How to log performance information client-side.

Using an object model helps while designing some web pages or content.

An object in a web project is something that has structure, instances, and purpose. Instances are all the specific occurrences of an object.
An instance of the object link can be "When to choose a progressive web app".

So, as we make our user experiences more object-oriented, we also make them more intuitive.” - Sophia Prater, ooux.com.

With this practice, 4 unintuitive objects are identified. The author use the example of a zoo website and they are:

• isolated objects: they are disconnected from related objects. The penguin animals are not related to the demos and habitat from the list of all the animals. The habitat is also included on the map view of the zoo, but there are no links to the Animal or the demo on this map view either.
• broken objects: ones that are not directly manipulatable. Demos are useful to show events to the user. As someone who would like to attend the penguin demo, I want to click on an individual demo to see more details, but it is not possible.
• masked objects: styled to look the same, but are actually different. On the DisneyPlus web view, there is no way to differentiate between a series or a movie. As you want to see a movie (one shot) instead of a series (multiple shots), you have to check on the detailed view in a panel.
• shapeshifter objects: opposite of masked ones, they are styled to look different, but are actually the same. Date formatting, links, and so on should look and behave the same across the website or the experience platform. They have to be consistent.

The internet works because

• there is a stack of protocols built to make things work. Each protocol solves one thing.
• all miraculously work together because these standards are open.

Tim takes the example of the network stack: Ethernet Packet, Internet Packet, TCP Packet, the port, and the email protocols.

Protocols and standards are everywhere. He takes more examples. When you publish a web page for example: it can be both human and machine-readable. It can be accessed through a URI and when someone follows a link to your web page, their browser opens a TCP/IP connection to TCP port 80 on the machine which is registered as serving the (www.whatever.com, etc) in question. All of that is because the URI specification says that what you can tell about a URI depends on the first bit, in this case HTTP. Tim explains in depth why these relationships exist.

An XML document is a less specified version of an HTML document. The namespace declaration gives a URI indicating the namespace used to interpret this XML though. And more...

An RDF document is based on XML and a triple: a value of some property of some object, or some relationship between some object and some other object. How to figure out what a triple means? A URI defines it, and its standard can be read while dereferencing it. The color example is great! So the stack for this document piles up from the Ethernet to the RDF MS 1.0 and RDF MS 1.0 definition of rdf:type.

A pattern is that each technology evolves into three stages: using numbers or strings, then using a URI, and then a dereferenceable URI. As we move on to later protocols, the protocols themselves become more diverse, so URIs were created instead of simple versions with numbers or strings. "The third stage of civilization is the one at which the identifiers can be looked up on the web.".

This stack prevents one from sending a nasty email to someone and then protesting that the message didn't mean anything. So if the stack is so strict, how does one send a nasty email message when one doesn't mean it? Many protocols have ways of breaking the chain, of including information that is not part of the meaning of the message.

For the email: an attachment. "So being able to refer to something without asserting it, whether you call it attachment, packaging, or quoting, is an important feature of a language. The fact that you can do this removes the last excuse for anyone claiming not to have meant whatever they did say in the main message!"