search

w3c / wpub

W3C Web Publications
https://w3c.github.io/wpub/
Other
79 stars 19 forks source link

Project Scope #400

Closed geoffjukes closed 5 years ago

geoffjukes commented 5 years ago

I would be grateful if someone could, as plainly speaking as possible, explain the difference in scope between this project, and the pwpub project?

TzviyaSiegman commented 5 years ago

@geoffjukes Excellent question. Sorry it hasn't been clear. Here is the very short answer:

wpub is publications on a user agent, no packaging. The current version of the wpub spec is simply metadata and manifest. pwpub defines a packaging mechanism for wpub.

geoffjukes commented 5 years ago

Thanks @TzviyaSiegman. As a primarily infrastructure and operations engineer, I am not fully clear on your use of the term "User Agent".

If I use the words "Online" and "Offline" as being use-cases for Web publications. Would it be a stretch to say that this project is focused on the standards for providing "Online" access to a publication, and the pwpub project is focused on collecting those "online" resources into a single file, for access "offline"?

wareid commented 5 years ago

"User Agent" for our purposes is any platform that would display/use Web Publications, this is pretty broad, but can include web browsers, reading apps, devices, etc etc. Most of the time we're usually referring to web browsers/agents that use web views for rendering.

Your use cases for online and offline are correct! There's a little more nuance (what happens to an "online" WP if you just happen to leave an area with Wifi?), but when we talk about "offlining" something, we are referring to packaging it or downloading it.

deborahgu commented 5 years ago

"user agent" means, in this case, "browser or reading system" (where reading system will mean something like a software or hardware book reader -- in epub, this would be something like a kindle, readium, ibooks, sumatrapdf, a nook, etc). Basically, something that can read the publication.

The "online" versus "off-line" question is related to packaging, but is not the same thing. If I send you a packaged publication, that might be a single packaged file that you can receive as an attachment in an email message. You might decide to read that on a kindle on the train, offline, which would be very handy. But you might also decide to read it while connected to the Internet, which would give you access to some remote resources which might not be part of the package. This might include links to external resources, for example.

In fact, one of the questions of packaging is that the creator needs to determine what are essential resources which need to be included in the package, and what are non-essential resources which might not be part of it. So, for example, a creator might decide that the CSS is not an essential part of reading the publication, but the embedded videos are -- or vice versa. If something isn't included in the package, that resource can only be accessed when the user is reading online.

So "packaging" is about thinking of a publication as a single, easily distributable file, rather than a collection of separate web resources.

geoffjukes commented 5 years ago

Perfect! Thanks all, that does make things much clearer.

@deborahgu if you drop the word 'web' from your closing statement, you would be describing the asset deliveries that we receive from audio publishers (separate resources sent via FTP) and I am assuming that the purpose of the Manifest is to describe those resources.

My primary interest is in the Manifest specification, especially as it has several uses for Blackstone. I am secondarily interested in packaging, only in so far as it may become a download option for our customers.