Categories
Personal Publishing

On hybridised RSS feeds as evidence of a need for weblog refactoring…

Right then – I feel a bit like I’ve got the wind behind me and it might not last so I’m going to plough right on into another subject before the demons of fear crawl up my leg. Dave Shea’s written a really insightful little piece (after Haughey) on the current trend for hybridised RSS feeds merging del.icio.us feeds and Flickr photostreams with normal weblog posts. Here’s some of the best stuff:

The problem I have is quite similar to what Matt describes: when new items show up in my newsreader from people I enjoy reading, Iím often mildly disappointed when itís simply a new camera phone image, or a couple of sparsely-described links to stuff Iíve already seen. Iíll go one further though, and say this about the practice: itís really damaging the signal-to-noise ratio of content I otherwise love.

And all I can say is that I couldn’t agree more. My RSS feed at the moment is a monstrous atrocity. It’s vile and clumsy and ugly and infuriating. But it’s as vile as it is because that’s where the software and systems that I want and need to use have led me. I want to fix it. I’d love to make it better, but to do so I’d have to sacrifice something somewhere along the way.

When I started weblogging it was all about the links, and the little asides and the one-liners and guff like that. I believe fundamentally that weblogs are communicative and social rather than being publishing ((Weblogs and) The Mass Amateurisation of (Nearly) Everything) and this seemed like a natural register for that kind of thing. Everything was nicely informal and easy. But the systems had their problems – Blogger style permalinks were an ugly and clumsy way to reference particular pieces of commentary, so people moved towards using Movable Type with is individual archives and built-in comments. Page-per-post sites, though, require a different form of writing – people change their interactions and the sites become less agile and less cross-conversational. Shorter posts get lost, posting becomes more of an effort and many things that one might like to talk about in passing get thrown away. There are benefits, it’s clear – you write longer, better, more considered things. But they’re not the same things that we used to be writing. You can see some of the transitions that occurred when I moved to using Movable Type in the visualisations that people did for me last year: Visualisations lead to Self-Knowledge.

I think this shift was really what caused the desire for people to start link-logs. Sites on the internet that we responded to emotionally but couldn’t find time to write Movable Type-length posts about were getting no commentary at all. People who were busy found that they simply weren’t writing anything for their sites at all – the length of posts that using MT seems to inspire (in my case anyway) started to be incompatible with post-work energy levels. The concept of a simple, throwaway linklog seemed to present a cure to this situation. People could post things and get their thoughts out into public quickly and easily. Or – just as importantly – they could keep track of the links that they wanted to keep track of – back to the weblog as personal link-organiser. It seemed like the perfect solution.

Certainly it didn’t seem to matter much whether or not the links were unique or whether everyone else in the world had posted them too. This was the time that saw the emergence of what I call microcontent voting. The more people linked to something, the more people saw it, obviously – but now it was becoming an exponenial relationship rather than a linear one. This was because of the newly significant presence of aggregators like Blogdex, Daypop, Popdex, Technorati. Now there was an effective feedback loop – if something got the attention of a certain number of people in the ecosystem, it would be brought to the attention of even more. A site that got only a few links could be at the top of the aggregators within a day, and experience thousands of visits immediately.

The links had – in part – ceased to be just something you did individually and instead became something that you did as part of a community that one way or another helped information bubble up. I think this found its best expression to date in Hotlinks, which I really think needs only to be abstracted into a more generic service to take over the world.

But linklogs had their own problems – how should they be integrated into a weblog site? Should they be individually permalinkable or should they be aggregated in daily clumps? Should they sit in central bars or be relegated to little-read sidebars? Different weblog authors found different solutions to these design problems – with kottke and Anil probably being the groundbreakers in this area. But it still felt a little clumsy. The weblogs were capturing everything again – covering a whole range of content from long-form essays through to the smallest particles of link data – but it wasn’t sitting together well – the weblog softwares didn’t seem (and still don’t) to have found a way to really consolidate this kind of combination management / conversation / publishing role. I personally strongly felt that my link-logs should be posted in daily digests as part of my main weblog. I’d done groups of links as posts for years, and I didn’t see why that should change now. But these things were far from simple to perform.

The introduction of moblogging caused another problem – the infrastructure for handling phone to weblog stuff effectively had never seemed to emerge in an elegant, simple and un-hacky way – clearly they’d need different templates for a start, and again decisions had to be made about how to integrate them with the design and layout of sites. Should the photos be in a sidebar? Should they be a different and separate weblog? Should they be posted each day, or immediately as the photos were taken and coming in?

The two sites that – for me – changed this picture enormously were del.icio.us and Flickr in that they both provided me with new tool-sets for managing stuff and they both also gave me mechanisms for posting to my weblog in ways that seemed to afford more benefits than they had costs. Firstly, del.icio.us gave me the ability to organise my links more effectively than my weblog had – and made the process of refinding my links much much easier – and some slightly provisional settings do exist for publishing a daily digest to a weblog. So I get the space to file my links in a way that makes sense to me, and get to expose this action back to people. Except of course it’s not a finished piece of functionality – I can’t give it alternative formats for the title, I can’t change how each link is formatted and I can’t stop it publishing everything to both my site and weblog with extra fields of information exposed that I don’t want people to see (like the del.icio.us tags that each post has). I can hide these tags on the web because I control the stylesheets. But it’s much harder (rightly) to do that kind of thing in an RSS reader. The consequence? The posts generated by del.icio.us in my RSS feed are ugly and feel clumsy – they’re functional, but they’re not how I would have them…

And the same is true of my Flickr photostream. The pictures that I take are aspects of my life, and I want them to be exposed to people in the same way as my overt posts are – but I have non flexibility. I can have them posted directly to my site, but then they don’t feel cleanly on my site in that they’re still hosted elsewhere. And I can’t aggregate them into clumps usefully – every photo appears as I take it, and I can’t make a daily archive of them to be posted into the body of my site. So feedburner becomes the best option for bringing these very separate things together – except it has design problems of its own. Titles of Flickr photos don’t seem to update and the integration of feeds – while beautifully elegant technically – does seem to create unbalanced or confusing feeds to experience… And if you’re asking why I want to keep everything together in one Feedburner feed at all, it’s because the functionality that feedburner affords me in tracking the number of people reading the damn feed is so incredibly useful to me. And I wouldn’t get that ease or accuracty of calculation by having multiple feeds…

Phew! So that’s the history of all weblog functionality in a nutshell, which wasn’t quite what I was expecting to write. But the point is that all through the history of weblogs, the technologies have opened up new doors and created new problems. Different functionalities make it possible to do one thing much more easily or effectively, but they come with a smaller cost elsewhere. We’re definitely moving in a positive direction, but each time we make a leap to a new level of functionality, things get more complicated and fractured and difficult for a while. Our feeds are ugly, and they don’t quite work right and neither do our sites. But this is because the technologies that we’re using to organise and collate our lives aren’t quite communicating perfectly and aren’t splicing themselves together in the way that we might like. And things are getting ever more complicated, and we need to do something about it.

And I’m beginning to think that the thing we have to do is start to reconsolidate and refactor the weblog concept itself. We need to take a step back for the first time in years and re-ask the question – what is it for? How do we find something hard and shiny in the middle of all these hybridised trends and make it the ideal shape to support all the other services that will grow upon and around it. In a whole range of issues – from the collation of our browsing to the handling of our photos, from the posting of our opinions to the way we’re relating to our social networks – the traditional weblog format is starting to buckle. So rather than concentrating on the specifics of clashing informational streams in our feeds and looking to fix them, I’m going to make the problem even larger and ask – are these clashes evidence of something more seriously broken? Does anyone really have any idea what we do next?

22 replies on “On hybridised RSS feeds as evidence of a need for weblog refactoring…”

The sleeping giant
Man, I didn’t know that my little late night blurb written mostly to myself would spark such a reaction. I guess I woke a sleeping giant or something. Dave is annoyed, Tom blames it on tools, and Andre wants more…

On the aggregation of personal feeds
I’ve been entertained by a conversation that’s been bouncing around the blogsphere the past couple days. On the 31st, I finally gave into feedburner and consolidated my feeds. A day later, Haughey posted a complaint about daily links polluting feeds,…

Stupid-Ass Feedburner
to manage their RSS or Atom feeds. This is all well and good; I think itís a cool service. What bugs me however, is that Feedburner lets you send the lame pictures you posted on your Flickr account as part of your feed. This is annoying. Most sites th…

Your perceived need to refactor weblogs is sure to be enabled primarily by the format and templates used in our publishing tools. The journey to NoFollow is an example of industry driven format change that should be effective because many bloggers won’t stray far from the default settings. So a call to action in this regard could be directed to the tool manufacturers, through lobbying and feedback.
The shortfalls you metion in del.icio.us and Flickr are mainly due to their default configurations not allowing publish/nopublish differentiation in their current form. It seems a few people have hacked together effective workarounds for the level of publishing control you desire in del.icio.us, and Flickr’s upcoming badge updates are sure to please. These enhancements will organically change the form of our publishing into yet another beast.
Change is the only certainty, hey.

If many people aren’t happy with reading combined feeds, isn’t the obvious solution to provide alternatives: a combined feed, a feed of just writing, a feed of just links… (like, ahem, some people do already 🙂 ). Sure, you won’t get Feedburner stats, but if getting statistics about your readers is more important than pleasing them, that’s up to you.
One other thing to think about is the differences between weblog front pages and their feeds – traditionally the contents of these have pretty much reflected each other. But given that they’re read in very different ways perhaps it would be helpful to think about them separately. As an example — there’s more of a case for keeping the types of content in feeds distinct (separate feeds), but combining it all in one stream on your weblog front page so it doesn’t get cluttered with patches of stale content if, say, you’re only posting links for a while.

But the point is that fundamentally I don’t think of them as separate entities. I’ve been doing posts that consist of nothing more or less than a list of links since 1999. The only reason they’ve been disconnected from the main weblog at all is because we’ve been looking for a better way of managing and templating them. The individual linklog seemed to provide that, and now the del.icio.us posting interface seems to me to provide a better version of it. Except of course (1) these interfaces are not posting content that works well inside an RSS reader, and it all has the same slightly mechanical feel because we can’t reformat our titles or our posts to reflect our style elsewhere (2) the MT weblog and the del.icio.us feeds are not meshing particularly elegantly. This is why I talk about refactoring – we broke up our weblogs across different platforms to improve our ability to manage certain types of content but still pump it out in an easy-to-archive, well-organised consistent weblog-style format. But at the moment our aspirations and the reality of splicing these systems together are far from the same…
Flikr’s another great example here. Realistically what I’m looking for in a post with photos in it is something that looks like this old gallery of a weekend in Norfolk. I want Flickr to be able to do that for me. I want the photos on my server. I certainly don’t want each and every photograph separately posted to my weblog, or appearing in my RSS feed – but hey – you make do. So the question becomes – how much of this is to do with sorting out the relationship between the various components and how much of it is to do with refactoring the central chunk – re-examining our model of how a weblog should function? That’s the question I’m asking.
Do we need different input fields for different kinds of posts, do we need mechanisms to turn input content into more complex templates at will (linklog as bulleted list posted once a day or posted individually with comments attached to each one, fully integrated with our weblogs among posts that work differently – should that be a function that MT can support?), do we need distributed social networking functionality among and in between weblogs to really help us formalise the relationships that are at the heart of the weblog? Where do we go next?

Great post, Tom. I share many of your experiences with the transition betweeen various tools (although perhaps one further as I no longer blog).
I notice you didn’t even introduce the problems of podcasts (or future videocasts) to the problem. Now imagine what fun that will be.
In regards to the feed, part of me thinks this would be a clue for semantically structured feeds (ie. what RSS 1.0 should’ve been) where aggregators could filter different content based on readers’ preferences.
I think some granuality in the definition of content streams from personal sites would be helpful.
It would certainly let me avoid the inane photography of people I don’t know whilst enjoying their links and commentary.

That last but one question is interesting – I nicked Ben Hammersly’s method of using the keyword field in MT to provide links to relevant del.icio.us tags for each entry on a weblog, and that’s a little step towards the sort of social networking functionality (I think) you’re talking about. It would be nice to have some sort of del.icio.us-like weblog-specific service that let folk tag and categorise their posts – automatically, from within their weblogging app – so each weblog post could point both to on-site archives of similar posts, and outward to collections of posts by others on similar topics. (Come to think of it, maybe I should get around to checking out the tagging system at Technorati?)
I’m not sure a full-blown rethink of the way we ought to understand weblogs in the light of all this splicing is needed, though – I’ve found that picking up the pace by posting a daily del.icio.us digest has prompted a return to the old Blogger style, because I can’t bear to see an index page full of link lists, so, aside from the difficulty of presenting different types of content, this is all a welcome step back from the shift to the more considered, less frequent posting style that Movable Type inspires.
One thing’s sure though: bunging absolutely everything into one feed is infuriating for readers – Phil’s suggestions above seem the right way to deal with that problem.

Hmm, I have lots of thoughts on this, but not enough time to write them now.
One thing though – I do think that there’s something a bit wrong with a hybridised feed containing of all your del.icio.us and Flickr posts. I don’t mind photoblog posts if they’re accompanied by a bit of commentary and have something interesting to say, but a photo on its own which has arrived straight from your mobile phone with not even a title (other than ‘you have received a new message’) can be a bit dull (sorry).
By the way, the titles of your Flickr posts in the RSS feeds don’t link to anything (http://www.plasticbag.org/tag:flickr.com,2004:/photo/3941878)

Photos and feeds is a tough topic. I tend to break them down into two categories: event-based and non-event-based. Event-based photos would be pictures of a wedding I might go to, or a conference, or a party. They’re easy to categorize and it makes sense for them to all be grouped together into one entry. It’s the random moblogging that’s tough. Maybe we need to have a way to tag photos as we send them out that can instruct the handling agent (flickr, feedburner) on what to do with them. All photos tagged “wedding” that come in today would get appended to the same entry. All photos that come in untagged go into a “misc” entry. You could update the feed each time, but at least you’d only have one entry per new type of photo, and people who want to see all the photos from that event could easily find them in one entry.

Tom, I’m glad someone worries about these things and seeks to provoke improvement. You make digital a better place. I simply make it a more cluttered place.

Feed splicing = content dilution?
Matt Haughey complains about stinky links in other people’s feeds, even though he himself has been splicing up his feed for a while:
After five years of doing this, I never thought I’d be reduced to handfuls of interesting links sprinkled with a dab of…

Online Vermin and Refactoring Weblogs
Thereís a lot going on out there and I wish I had more time to take it all in. Iím about to leave town tomorrow morning for a bachelor party so Iím posting the bare minimum here. A couple of things that caught my eye recently: Amy Gahran has been posti…

Can one not simply use e.g. WordPress’s ability to send out feeds for each *category* of posts and then selectively subscribe to them?
If an author were really concerned about this, he or she could devise maybe three major categories (blog; photo; links) and provide feeds for those. Subcategories in each would be handled through the cascade effect; you the reader would have to choose only from the three main categories.

It’s awful at at the moment. The short term
solutions I can see are:
a) good headlines – too many people are trying
to be “clever” with their headlines.
Do a search on “good headlines rss” on google
and my little piece comes up 4th among 3.5
million searches. Why? One of the reasons is that I used a good headline.
b) RSS needs to be simple again. It’s heading the
same way as HTML in the early days.
Bad habits are developing – from users,
software developers and big sites.
My pet hate at the moment is people
putting the whole article
into the “description”. If you have a
long article a description might be needed.
If not let the headline take care of business.
Just because your
chosen RSS-reader always expect a
“description” (NetNewsWire ahoy!) does not
mean it is right to do so.
If a user selects a headline in a
feed and there is no “description” it should
show the article (guid or permalink): that’s
what an RSS-reader is for. In length:
RSS frustrations and the case for lightweight feeds.
c) maybe some trickery with “f:” for flickr,”l:”
for linkloggery and so in front of
the headline to indicate that this is not a
complete piece/”home-written” article

Online Vermin and Refactoring Weblogs
Thereís a lot going on out there and I wish I had more time to take it all in. Iím about to leave town tomorrow morning for a bachelor party so Iím posting the bare minimum here. A couple of things that caught my eye recently: Amy Gahran has been posti…

The Thrill is Gone
Last week I was instant messaging with Brent who’s started blogging again after I think about a year. One of Brent’s favorite reads shut his site down because he felt the thrill had gone out of blogging. I’ve felt it…

I certainly agree that there needs to be a change. I had been toying around with an idea for personalized feed subscriptions. As readers we should be able to create our own feed which contains only the posts from the categories we’re interested in. Why this isn’t the norm, I have no idea. I posted more in-depth about it here:
Personalized RSS Subscriptions
Hopefully we’ll see a change soon.

Comments are closed.