If RSS scales badly for sites like Gizmodo, don't use it for sites like Gizmodo. RSS is great for small-scale personal publishing; in fact some sort of aggregator is practically a necessity for such a thing to work (this is why blogging has drifted to tumblr, facebook, and twitter: publishing integrated with aggregated consumption is a powerful thing).

This leads to a chicken and egg problem: when someone is first introduced to RSS, the only sources they know about are ones that don't work well with RSS. It's hard to cultivate the kind of network that makes RSS valuable to you until you've already bought in to the concept.

Personally, at my peak I was subscribed to about 400 feeds, but they only published about 100 items a day.

The other issue is that a lot of these sites gate the content fed through RSS to make you visit the site for the whole thing (plus, you know, ad impressions).

RSS works best when the content can be fully digestible in the reader. If you can read most of it there, great--you get a preview. What I hated was when I'd get like a line or two, or couldn't even see an image load in an image-only post without visiting a sloth of a site.

Oddly enough, my Reddit homepage is pretty decent now due to pruning subscriptions and using RES.

Yahoo had a very good approach to the problem of scalability with Yahoo Pipes, that allowed nearly anyone to filter and mix different RSS feeds (say, for instance, get from Wired only those articles wrtten by a given author and containing a certain keyword), but it got axed very recently.

Google Reader had social too, and served your purpose very well. Plus you didn't have to filter out your uncle's political rants or your cousin posting her new baby's every action.

rss had two major hurdles that made it difficult for people to pick up.

1) there was no collaborative space for teams to work together to create feed mixes. No service I know of became popular that allowed team of 7 people to moderate the member feeds of a metafeed. This harmed discovery.

2) Deduplication was usually a human task. Techmeme really spoiled me. It was impossible to see what other people had already marked as duplicate "aggregations" of an original story.

these two together basically meant that EVERY person who used rss reinvented the wheel. You couldnt latch on to other peoples work like you can subscribe to a subreddit. Twitter has a similar problem, it lacks collaboratively maintained lists. It kind of blows my mind how neglected the idea of a "run your own slashdot/techmeme with friends" is.

i probably sound like a shill by now, but redef.com seems to be moving in the right direction.

Google Reader's public labels had the capability to create a group-edited metafeed, but this advanced use case was never advertised. The one use that I know of for it was on the sidebar of the Google Reader blog: Each team member had a public label, and there was one person who had a label which subscribed to each of the others. Any feed that any team member subscribed to with their public label would show up on the sidebar of the blog; people would do this for their own blogs or their shared items.

I guess my vision of a metafeed is something that autoingests rss feeds, presents them to editors as "pending stories" and the upvoted ones become visible to non-logged-in-users/subscribers after a certain threshold. Maybe editors/moderators/curators (whatever we call them this week) would have some kind of superupvote to use sparingly if an important article was justified.

I imagine this place being like reddit, where I can view a giant list of rooms, and see how each room and its members are mixing content. it should also be easy to clone/fork (github/yahoopipes) a configuration if your team wants to start where another team began.

I also like how counterparties had a list of rejected stories, stories they acknowledged but refused to follow.

really I am describing a piece of technology that lets any group of people maintain their own techmeme plus a reddit like index of all the different rooms, that allows you to subscribe to multiple rooms and see them in one feed.

RSS had the concept of OPML, which could have been used for that purpose, but I only saw it used in very isolated cases.

The curation angle does feel like a space that the likes of Skimm, NextDraft et al have picked up.

I'm not in love with the idea of another daily email newsletter, and Redef often feels like a long list of pieces that I "ought" to read rather than stuff I get genuine pleasure from, but it's the best we've had in a while.

OPML was an export of lists, it didnt let you and me edit a list together. maybe opml+github+humandeduplicationofstoriesliketechmeme

the thing that makes techmeme great is that they CHOOSE the best story, and then link to the duplicates. It might be the first, or it might be the most researched source, and they may change the link if a better one appears. Hackernews does similar, but I cant easily click "other sites with variants of this story"

as an aside, I quite dislike the email newsletters, it seems very antiweb. nextdraft not having a website is infuriating.

That was always my policy: reader for those feeds I want to read every post of, visiting the websites for anything like the NYTimes.

I feel like we cold solve this with better feed threading. Take every article with certain URLs in common and present them in a threaded view.

This way the massive content mils would get written down to a single chain.