WordPress.org

Make WordPress Core

Opened 8 years ago

Closed 8 years ago

Last modified 7 years ago

#2277 closed defect (bug) (wontfix)

Ditch RDF and RSS 0.92

Reported by: davidhouse Owned by:
Milestone: Priority: normal
Severity: normal Version: 2.0
Component: General Keywords: bg|dev-feedback bg|commit
Focuses: Cc:

Description

RSS2 and Atom are both competent feed formats. All feed readers these days are built to understand one or both of them. It's time we ditched RSS .92 and RDF, they've been bloating the WordPress core for too long. If they're desperately needed a plugin can provide.

Change History (18)

comment:1 skippy8 years ago

I second this proposal.

comment:2 ringmaster8 years ago

Sounds like a good idea to me. Especially if we beef up Atom support.

comment:3 TonyStreet8 years ago

I agree.

comment:4 davidhouse8 years ago

  • Keywords bg|commit added

comment:5 markjaquith8 years ago

+1
We need:
RSS 2.0, Atom 1.0, and MAYBE Atom 0.3... but probably not.

comment:6 Denis-de-Bernardy8 years ago

  • Component changed from Administration to General

+1

+1 to ditch Atom too. Atom is a dead end technology. MS, AOL and Yahoo all chose RSS. Google is providing RSS and Atom. Plus, having several formats confuses users, and I cannot name a single feed reader who cannot understand both formats.

comment:7 davidhouse8 years ago

People like and use atom. No point ditching something people use.

comment:8 fncll8 years ago

+1 assuming plugins are made available for users that do want/need them

comment:9 ear1grey8 years ago

I agree with the concern over bloat, but suggesting RDF should be dropped because "readers these days are built to understand one or both of [Atom or RSS2]" is akin to arguing that IMAP or POP should be dropped from all mail servers - the protocols serve totally different audiences.

Ditch RDF and WP loses the "semantic markup" part of "WordPress is a state-of-the-art semantic personal publishing platform" - having RDF available means that WP content can be more easily integrated with Semantic Web enhanced applications - it's a low hanging fruit which (IMO) should be kept available because it's user base is significantly different to the users of Atom or RSS.

Basically RDF makes wordpress more attractive as an experimental platform.

The suggestion that "they've been bloating the WordPress core for too long" is fair enough, so long as it's also levelled at RSS2 and Atom - should they not ALL be plugins?

comment:10 matt8 years ago

If it ain't broke...

I see no compelling reason given to drop support for this. The only argument has been "there was a bug that wasn't fixed for a while", but honestly I don't think dropping ~90 lines of template code we bundle is going to help that problem.

comment:11 ringmaster8 years ago

In an informal survey of users on #wordpress, I asked users to grep their server logs for RDF requests. On average, RDF makes up 2% of all feed requests. Most of those requests came from engines (like Yahoo) that can easily acquire the data from one of the other feeds.

It's a small sample size, but I think it'll prove reasonably consistent across WordPress blogs.

When we make a change that will affect all the feed scripts, why waste time updating this unused format unless it can be done in step with any other supported format?

More thoughts here: http://asymptomatic.net/2006/01/17/2216/no-more-feeds-for-wordpress/

comment:12 skippy8 years ago

Marginally compelling reasons to drop RDF from the core:

  • shave 90 lines of code, allowing us to (continue to) boast that WordPress gets leaner with each release!
  • one less thing to test with each release cycle!
  • it has yet to be demonstrated that RDF provides something not provided by the other feed formats

By announcing early that RDF will be removed from the core, those parties interested in RDF have ample heads-up of the pending change, and develop a plugin to be made available at or shortly after the next version's release.

comment:13 michelv8 years ago

There is no practical reason to remove RSS 1.0 support.

  • Bloat? 90 lines of template code that reuse code from other feed-related functions is hardly bloat.
  • Maintenance cost? 8 revisions in a year hardly constitutes high maintenance cost.
  • Outdated? Then the logical decision would be to support RSS 1.1? But there are very few tools ready to consume RSS 1.1, while many have been supporting RSS 1.0 for years.
  • Not updated in years? How about: it's _stable_ and that's something that matters, especially to aggregator vendors (whom, I bet, are not very happy with RSS 2.0 elements seeing various semantic changes as Dave Winer redefines it along, with no way to distinguish "RSS 2.0 from today" from "RSS 2.0 as defined in April") and big feed consumers.
  • RSS2 and Atom do the same thing? Then the logic dictates we go the whole way and stick to one true format. While Atom has been called "semantic", it's still a long way until there are enough tools to process Atom as there are to process RSS 1.0. There is a reason why Yahoo! and others fetch the RSS1 feed.
  • It could be made a plugin? Yes, I'm eager to see thousands RSS1 subscribers (and big feed consumers) waste everyone's time asking authors they read "can you pretty please add the RSS 1.0 plugin?"

Besides the bloat and maintenance arguments, I'm still wondering what little benefit would come out of removing RSS 1.0 feeds.

comment:14 garubi8 years ago

The point, IMHO, is not that "all the feed contain the same data": this is obvious :-)
The point is that RSS 1.0/RDF contains data tagged/marked/packed in a different way. A way that can provide to those interest more meaningful information about our blog content. And as this information can help my content to spread out to the world I don't want to loose it!

So ok dropping 0.9, but please don't touch 1.0.
(ah, yes, we should write some plugins that can provide this feeds... yes, this could be the right way...)

just my 2 c

Stefano 'garubi'

comment:15 Mr_Cynical8 years ago

So we 'save' 90 lines of code out of how many hundreds of thousands? I simply fail to see how removing functionality can possibly make Wordpress better (as opposed to lazier). If for some reason this is accepted, then the plugins must be provided before the feeds are removed from the core, and kept update by the main development team (to prove that this is a genuine belief that it improves Wordpress, rather than 'we can't be bothered maintaining this functionality anymore'

comment:16 raster8 years ago

I agree that plugins need to be provided before the feeds are removed from the core. I think it would be a disservice to WordPress and it's users to do otherwise.

comment:17 skippy8 years ago

  • Resolution set to wontfix
  • Status changed from new to closed

Closing, based on overwhelming opposition.

comment:18 Nazgul7 years ago

  • Milestone 2.1 deleted
Note: See TracTickets for help on using tickets.