OPDS – Easy to Use, Open and Decentralized Content Distribution | Hacker News Reader

OPDS – Easy to Use, Open and Decentralized Content Distribution

opds.io

71 pointsrobin_reala25 comments

The Website is really bad in communicating what OPDS actually is. "Distributing Content" can be anything, from ads, videos, images, websites, video games, ebooks, news, songs, etcpp.

But the drawings suggest that this is only about (e)Books and other text-form articles. They seem to actively avoid calling it what it is and instead using "Content". Is this some kind of special lingo?

My impression from looking at this a bit after seeing this submission is that yes, it's mostly eBooks in practice, but I think they use the general wording because it's not strictly limited to that - you could pack audio or video publications in the formats too.

This is a neat idea in the way it works with very well-supported open web standards. It feels like it could fit within the Dat/Hyperdrive ecosystem as a means for syndicating content held within decentralized content distribution networks.

As one of the architects of Ara [0] I would be curious to know if proof of content ownership and/or intellectual property protections are offered by these protocols? This has always proven to be one of the more challenging aspects to decentralized content distribution, in my experience.

[0] https://ara.one

The image of literal sheep inside tiny barbed wire fences is a bit extreme but I do feel the same problem.

I want to be able to purchase something then access it reliably without re-purchasing it.

I think the main problem is that digital copy/distribution is so cheap relative to content creation that the existing model of bundling those costs to emulate the inherent properties of physical distribution naturally leads to walled gardens.

OPDS seems to address only a small portion of the problem by describing a how publishers should host their content in a standard way that makes it easier to access/index.

Am I missing something in the specification?

Is there a list of software/services supporting this somewhere? (It's an easy mistake to make, but I find that difficult with many "standards" websites - how do I judge if this is actually in practical use, and if yes in what fields?)

I submitted this because I was investigating supporting the new 1.2 features for Standard Ebooks collections. Our main feed is at https://standardebooks.org/opds and you can find a feed for all of our books at https://standardebooks.org/opds/all

I currently use it across multiple devices:

KOreader supports it on Kindle,Kobo,Linux,Android

Foliate (great EPUB reader) supports it on Linux

On Android - MoonReader and a few other apps. Ubooquity app for eg.

On iOS- I use KyBook, and I’ve been planning to work on something that integrates this into the Files app.

The mobileread wiki has some more, but the page has some spam: https://wiki.mobileread.com/wiki/OPDS

It'd be interesting to see something like this built on podcasting standards, which has successfully filled these design goals for a couple decades now.

It'd require a handful of thoughtful RSS extensions, and possibly OPML extensions for "Shelf" interop, etc.

With OPDS, I don't understand what enforces the "buy" scenario. Or is it not enforced at all?

OPDS may work fine to distribute free digital catalogs, but I find it pretty hard to integrate it with purchased material. You definetely won't have the same experience as a proprietary platform.

So your end - user downloads a random opds-enabled client, and adds your catalog. Let's say you want to make sure the client owns the material. You can't expect the client to redirect to a sign - in page, that's not part of the protocol. Purchases also do not work well (try the O'Reilly opds for example) . It's just a mess.

It's not clear to me what exactly opds is trying to solve. I found it such a stretch when they claim it's an alternative to proprietary book platforms. These systems have tons of people thinking about the entire user experience. What opds offers is just a tiny fraction of what's necessary to build that.

It only supports Basic Auth, so the client is expected to support some out of band auth to get a token?

I think the new OPDS JSON spec had some notes on OAuth, not sure if that resulted in anything.

How is illegal content handled?

I think this is a centralized ecosystem "brainworm."

Every feed publisher would be responsible for providing a feed that does not include illegal content. This is just a protocol for cataloging and distributed information about content. So, it'd be like asking "how does TCP/IP" handle illegal content? It doesn't and that's not the purpose.

Now, publishers of feeds could indeed keep illegal content from being linked(in a naive way) by pairing this protocol with another protocol that can check hashes of file content against a database of copyrighted works signatures.

If copyright holders want to protect their works online they should be making it easy, by providing such a service.

The new spec (2.0, not implemented by any client) is made by a committee that is also building a standard DRM implementation to go alongside it. (Search for Readium).

That was when I lost all interest in the new spec.

Is this supposed to be a supplement to RSS, or in what ways is it unique or different from another XML-like feed standard?

It builds on Atom, and provides a collection-based vocabulary, with at least at the start was focused on ebook distribution.

According to the site, they're switching and the future work builds on JSON-LD (IMHO: ughs) and Schema.org (instead of Dublin Core)

What does any of this even mean?

If you are looking to run a lightweight OPDS server: check out elibsrv. The indexer is written in C, which writes to a SQLite database, which is then presented using a lightweight PHP script. And as a plus, it supports OPDS search, which most other implementations don’t (such as Ubooquity).

Ubooquity does support OPDS search.

hmm, doesn't seem to work on either of the clients that I Use (Foliate/Kybook2).

As far as I remember, I strictly followed the OPDS specifications when I implemented search, so it should work with any OPDS client. Just in case,I just tested it successfully with Moon+ Reader on Android.

I can't test the clients you use as I have neither a Linux or iOS device at hand, but if you have any kind of error when trying to use them, I could take a look.

Might just be an issue with my install. I'll do an upgrade and report an issue if I still can't get it working.

I was using the superb Marvin on iOS for my own content. Unfortunately the OPDS app for NextCloud doesn’t seem to work for a while now.

I'm the author of that app, it was made for a much earlier version of Owncloud and it would take more work to get it up to date with the current state of the platform than it would to start anew. It also was not made to support many of OPDS' more 'advanced' features like sharding and search. I've started to look into making a new version, possibly integrated with my other Own/Nectcloud app 'Reader', look for more info on that front in a few weeks.

Yeah, Marvin’s great. Pretty much the best ebook reader out there.

Have you checked out BicBucStriim? It’s a PHP based OPDS server that uses a Calibre library as its database. I’ve set it up on a Synology drive as that was most convenient, but it’ll run pretty much anywhere.

Reply on news.ycombinator.com