Re: Continuing discussion on Polyfills

Benjamin,

That is one way to look at it, but I would not suggest or expect the use of markdown or anything like that.  The input would be HTML with the option of CSS and/or Javascript.

The best way I can describe it is to build it (which I still hope to) but the best way to think of it is as injecting the reading system into the book.  It becomes a slightly different concept than having a reading system LOAD a book (such as Readium JS) because the reading system is loaded, then the content, whereas this loads the content and reading system at the same time – and they are linked, so if a browser were to cache all the files, it would technically have all the reading system as well.

Ultimately it is not TOO far from a web-based reading system that loads content except that on a conceptual level, it would be the “this is what we’d love the browser to handle for us, but it doesn’t yet.”

-Nick

From: Benjamin Young <byoung@bigbluehat.com>
Date: Tuesday, February 6, 2018 at 10:50 AM
To: "Ruffilo, Nick" <Nick.Ruffilo@ingramcontent.com>, Hadrien Gardeur <hadrien.gardeur@feedbooks.com>, Jeff Buehler <jeff.buehler@knowbly.com>
Cc: "deborah.kaplan@suberic.net" <deborah.kaplan@suberic.net>, "Siegman, Tzviya - Hoboken" <tsiegman@wiley.com>, W3C Publishing Working Group <public-publ-wg@w3.org>
Subject: Re: Continuing discussion on Polyfills


Nick,



What you've described so far sounds a lot like a static site generator, such as Jekyll: https://jekyllrb.com/




Authors write simple (stupid) content-centric document (typically Markdown) and organize them into a certain structure and then the tooling does the rest (outputs HTML, adds scripts and polyfills, etc).



One such static site generator that does all kinds of book-ish things is https://www.gitbook.com/




Their tooling is open source https://github.com/GitbookIO/gitbook (Apache License 2.0) and can be used to generate PDFs, EPUBs, sites, etc.


Is what you're proposing essentially this same idea?



Cheers,

Benjamin

--

http://bigbluehat.com/


http://linkedin.com/in/benjaminyoung


________________________________
From: Ruffilo, Nick <Nick.Ruffilo@ingramcontent.com>
Sent: Tuesday, February 6, 2018 9:56:02 AM
To: Hadrien Gardeur; Jeff Buehler
Cc: deborah.kaplan@suberic.net; Siegman, Tzviya - Hoboken; W3C Publishing Working Group
Subject: Re: Continuing discussion on Polyfills


I like the way this is presented and said.



Responding to a previous note from Ivan – When I was talking about Polyfills, I was not referring to anything on the back-end.  What I was recommending is that a back-end script is what injects any polyfills into the documents.  Once I have the example, it will be more clear.



And while everyone might not have access to a customizable server, I promise you that if there is value – services will pop up to make it easier.  When the web first started, servers were expensive, but services like Geocities gave you a small chunk of space on their servers so you could get started.  Publishing is not easy, and that’s OK.  I believe we need to focus on making the overall process as Logical, accessible, and future-proof as possible.



As for all the Accessibility feedback – thank you.  I hear it loud and clear.



-Nick



From: Hadrien Gardeur <hadrien.gardeur@feedbooks.com>
Date: Tuesday, February 6, 2018 at 4:33 AM
To: Jeff Buehler <jeff.buehler@knowbly.com>
Cc: "deborah.kaplan@suberic.net" <deborah.kaplan@suberic.net>, "Siegman, Tzviya - Hoboken" <tsiegman@wiley.com>, "Ruffilo, Nick" <Nick.Ruffilo@ingramcontent.com>, W3C Publishing Working Group <public-publ-wg@w3.org>
Subject: Re: Continuing discussion on Polyfills



IMO we should separate things on the authoring side from the user agent side.

Authoring

A Web Publication MUST be entirely readable without any kind of polyfill, extension or specific browser support.

This means that all resources should be reachable through navigation, ideally in a sequence that follows the reading order.

For every other affordance (I'm working on those in the lifecycle branch<https://github.com/w3c/wpub/pull/130>), the author MAY attempt to support them from the entry page (HTML document returned by the WP address). This can be achieved by pointing to a Web App from that page.



User agent



A WP aware user agent SHOULD enhance the experience of reading a Web Publication by providing additional affordances that won't clash with the author's intent (easier said than done, user settings for example are very difficult to handle).



IMO, presentation and navigation affordances MUST only enhance resources that are within the scope of the WP which means:

  *   all resources listed in the default reading order and list of resources
  *   but this excludes the entry page or any other resource not listed in our collections

For resources that are not within the scope of a WP, a user agent MAY provide two other affordances instead:

  *   switch to publication mode
  *   add to the list of publications