3 ms·

I've always loved MDN and wish them well, but - how is it acceptable to run off github for a project that wants to encourage authoring sites on a supposedly fe

I've always loved MDN and wish them well, but - how is it acceptable to run off github for a project that wants to encourage authoring sites on a supposedly federated medium (the web)? - do we really need an enormous "web documentation project" for something entirely man-made, and which once set out to facilitate easy self-publishing? Even thinking about these and similar questions means that, rather than attempting to document the craptastic overcomplicated web, we're probably better off to leave the web behind us for good, and start to concentrate on defining HTML+CSS subsets (such as for purely static docs, for light content apps, and so on), to distribute simple static text via alternate p2p protocols. W3C and WHATWG have failed miserably to do so.

breck 6y ago
> - how is it acceptable to run off github for a project that wants to encourage authoring sites on a supposedly federated medium (the web)? You may have a point here in reality, but "technically" it should be very easy to push to multiple backends (GitLab, for example). So I could see lock-in to GitHub not ever being more than a hypothetical problem.
bastawhiz 6y ago
> do we really need an enormous "web documentation project" for something entirely man-made, and which once set out to facilitate easy self-publishing? What point are you trying to make? Should Oracle throw out its docs for Java because it's a large language that's man-made and designed for easily writing software—but lots of folks here think it's not good? > start to concentrate on defining HTML+CSS subsets (such as for purely static docs, for light content apps, and so on), to distribute simple static text via alternate p2p protocols And this will still require docs, and tutorials, and information about the P2P protocol, and how it's kept secure and ~anonymous, and how to facilitate discovery, and how to avoid a network partition, and how to build non-static tools like search engines, and how to securely interact with those tools, and how to do archival work, and a million other things that make that system an ecosystem...it's not easy to build a resilient distributed network that actually works. The fact of the matter is that people use the web whether you think it's good or not, and having decent docs for it is important. Nothing is stopping you or anyone else from building an alternate web, but so far there hasn't been anything that people actually care to use.
- tannhaeuser 6y ago
  > What point are you trying to make? That maybe our time is better spent by taking the chance to simplify the languages of the web along use cases rather than document the whole of it in an encyclopedia. The Java example is a good one, since IME tools such as javadoc, and just writing down the purpose of something and why and when it was introduced, immensely helps cleaning up your API surface. The web as it is is way too complicated when it really doesn't have to be, as witnessed by browsers vanishing.

breck 6y ago

> - how is it acceptable to run off github for a project that wants to encourage authoring sites on a supposedly federated medium (the web)? You may have a point here in reality, but "technically" it should be very easy to push to multiple backends (GitLab, for example). So I could see lock-in to GitHub not ever being more than a hypothetical problem.

bastawhiz 6y ago

> do we really need an enormous "web documentation project" for something entirely man-made, and which once set out to facilitate easy self-publishing? What point are you trying to make? Should Oracle throw out its docs for Java because it's a large language that's man-made and designed for easily writing software—but lots of folks here think it's not good? > start to concentrate on defining HTML+CSS subsets (such as for purely static docs, for light content apps, and so on), to distribute simple static text via alternate p2p protocols And this will still require docs, and tutorials, and information about the P2P protocol, and how it's kept secure and ~anonymous, and how to facilitate discovery, and how to avoid a network partition, and how to build non-static tools like search engines, and how to securely interact with those tools, and how to do archival work, and a million other things that make that system an ecosystem...it's not easy to build a resilient distributed network that actually works. The fact of the matter is that people use the web whether you think it's good or not, and having decent docs for it is important. Nothing is stopping you or anyone else from building an alternate web, but so far there hasn't been anything that people actually care to use.

tannhaeuser 6y ago
> What point are you trying to make? That maybe our time is better spent by taking the chance to simplify the languages of the web along use cases rather than document the whole of it in an encyclopedia. The Java example is a good one, since IME tools such as javadoc, and just writing down the purpose of something and why and when it was introduced, immensely helps cleaning up your API surface. The web as it is is way too complicated when it really doesn't have to be, as witnessed by browsers vanishing.