My name is danah boyd and I'm a Principal Researcher at Microsoft Research and the founder/president of Data & Society. Buzzwords in my world include: privacy, context, youth culture, social media, big data. I use this blog to express random thoughts about whatever I'm thinking.

Relevant links:

Archive

what i want in an RSS tool

When i first started using RSS, i was ecstatic. Rather than relying on going to each person’s page, i could just throw them all in one place and go through them. I’m a bit more disillusioned now.

I got all excited and started adding every blog that had an interesting thread. Almost humorously, i started breaking after about 150 regularly updated blogs. Worse: i miss half of the interesting posts that i want to read because i’m too overwhelmed.

This made me sit back and think about what kind of an RSS feed i want.

First, i want to be able to choose to watch an entry, a topic or a person. I don’t want to be forced into a person only; this unit of view is way too big.

Following a person should be like now – i see everything in their feed.

Following a topic means that i can specify things like “all entries by this person related to ‘echo chamber’.” As such, i can follow whatever this person has to say on something. This is particularly relevant for following bloggers who have a topic of interest to me, but whose entries are by in large, not of interest. Of course, i know that this means that all of you YASNS followers will never read my V-Day writings. But alas, i know you don’t care about my politics anyhow.

Following an entry is a bit more fun. Say that i find an entry that i think is of interest – either in my feed or out there on the web. I should be able to add/mark the entry so that the entry tells me when there are new comments and all new trackbacks get inserted into my feed as single entries too.

Personally, i’d like to tap into the graph of blogs. Technorati knows the linking structure. Forget blogrolls. We can see who links to who embedded in their blog. We can determine blog topology. Why can’t i have topic-based RSS requests. “Tell me anyone within 3 degrees of my network who is talking about ‘rape’ or ‘domestic violence’.” In theory, Google should help me on this but that’s overload! Plus, i can’t pull out just the blogs (a feature that i’m STARTLED they haven’t implemented after having purchased Blogger).

Finally, every day trusted friends of mine send me URLs. When i surface for air, i have to fish through thousands of emails to find those interesting tidbits. I love getting recommendations from friends. Why can’t they just drag a URL into my RSS feed? Why can’t i have a feed of “every URL that Ronen thinks i should read”? Frankly, this would be so much more efficient to reading things. Plus, my friends know what is of interest to me. Another thing is that it should be possible for me to have a public dump. Anything that people in the public think that i should read.

I don’t want automated recommendation systems. I want tools so that my friends can do what they already do – pass on information that they think is relevant. But i want to make it easy for them. And perhaps have a mechanism to say “THANK YOU!”

As more and more people blog, RSS is going to break on the social/attention level. In many ways, it already has for me. I’ve started interviewing bloggers and i’m fascinated by how hard it is for them to consider adding something to their RSS. Overload. Overload. If anyone wants to know why the early players get all of the attention, it’s because RSS feeds focus on people, not ideas, and the early players are too overloaded with following the other early players to consider new people.

Anyhow, just an idea… Although i’d love anyone’s thoughts on this approach. [Perhaps there are tools out there… So far, i’ve only used Bloglines, Shrook and NetNewsWire]

Print Friendly, PDF & Email

55 comments to what i want in an RSS tool

  • Agreed. To a certain extent…

    RSS aggregators need better management UIs. Grouping into folders is a start but far from optimal.

    Following “people” RSS feeds (weblog entries by Person A) would benefit from being tied into your AddressBook/Buddy List (hello FoaF?)

    Following individual entries is possible if everyone created RSS feeds for individual entries. In MTthis is done simply by adding an extra Individual Entry Archive template. Was never done “by default” so it never caught on. Shall we begin?

    Follwoing “topics”. Ahhh here we get tricky… Technorati/Feedster/Google could provide such a service…

    As for URLs, we’re talking Bookmarks essentially… http://del.icio.us/ is a good start… get your friends to all use it… (good luck)

    Myself have been thinking alot along these lines recently:
    It’s all about “people” -> nodes -> identities. If your URL is your identity, then we can go on and say every URL that is somehow related to me is part of my identity, be it an identity I put forth or one others construct of me.

    In other words, what we are headed towards is essentially a giant URL I/O manager…

  • Danah it’s like you read my mind! I’ve been working on a little side project around topic-based viewing that doesn’t include the social network aspect (yet). I generally use NetNewsWire, but get frustrated by the shear amount of feeds I subscribe to and the lack of tools to effectively use them. I’m a big fan nice, rich client-side apps that leverage services so I’ve started working on a little something that leverages some nifty Apple technologies (like the SearchKit and language analysis tools built into Mac OS X) and back-end services. I’ve blown away my notion of feeds and have a large “hopper” that I just throw RSS links into. My blogsphere is then reshaped around my topical interests and the feeds are then grouped by relevance and entries pulled together into one view on-the-fly (basically just saved search terms executed locally). This way I can wing through many topics based on my current research agenda and increase my signal-to-noise ratio. Not for everyone, but if it works well enough, maybe I’ll throw it out for the community to bang on.

  • Oh and… er… most of this stuff is already being done by Winer’s Userland suite of tools. Feature and implementation wise, the man is miles ahead. (Without getting *how* it is implemented etc…)

  • joe

    Bloglines has the potential to do much of this… send them your thoughts and they might see it as strategic to build in some of the capability you mention.

  • What I Want in an RSS Tool

    apophenia writes, “When i first started using RSS, I was ecstatic. Rather than relying on going to each person’s page, I could just throw them all in one place and go through them. I’m a bit more disillusioned now. I got all excited and started adding …

  • Danah, I’m confused by your use of the word “person.” Do you mean an actual human being, or did you mean blog? On group blogs, like Crooked Timber, are you saying you want to be able to follow each individual writer seperately?

    As to keywords, on my end, as someone writing weblog software, it would be easy to output XML that included all the keywords an author had associated with a post. But there is no official recognition of such keywords in RSS. Do you have a suggestion?

  • Following individual entries is possible if everyone created RSS feeds for individual entries. In MTthis is done simply by adding an extra Individual Entry Archive template. Was never done “by default” so it never caught on. Shall we begin?

    I don’t understand what is being said here. What is an RSS feed for a single entry? One that doesn’t list any other entries? Are we talking about a unique address for the RSS feed for each entry? What address would that be? Would it be useful to have an addition to the url that is universally understood? What would be the right way to do this? For this entry, as an example, would it by useful to put an /RSS at the end? Like this:

    http://www.zephoria.org/thoughts/archives/2004/03/03/RSS/what_i_want_in_an_rss_tool.html

    Am I way off base, or is this what we are talking about?

    Follwoing “topics”. Ahhh here we get tricky… Technorati/Feedster/Google could provide such a service…

    Wouldn’t it help to have a place to put this in RSS?

    RSS tools

    browse by topic

    Again, is this what we are talking about, or am I way off?

  • Sorry that last bit didn’t come through. I wrote this:

    “Follwoing “topics”. Ahhh here we get tricky… Technorati/Feedster/Google could provide such a service…”

    Wouldn’t it help to have a place to put this in RSS?

    <keyword>
    RSS tools
    </keyword>

    <keyword>
    browse by topic
    </keyword>

    Again, is this what we are talking about, or am I way off?

  • Is Userland still Dave Winer’s? Does he still work on it day to day? I thought he was at Harvard now? I had the impression he was no longer working on Userland. He is a user now, not a programmer, yes?

  • The future of attention management

    danah boyd: What I want in an RSS tool .

  • Really, really, really good thoughts.

    Thank you.

  • RSS Needs – 2

    A while ago, i had struggled to articulate my needs from RSS feeds …

  • I’d recommend checking out Furl.net.

    Similar to del.icio.us, but with added search (full-text search!) capabilities and other cool features, you can have your friends add their favorite sites with just one or two clicks from their browser, and you can follow their choices via RSS feeds.

    Yes, this then requires your friends to actually ‘furl’ their favorite useful sites, but once you condition them that it’s easier to do this than to open up their e-mail program, start a new e-mail, address it, add a subject line, write something, add the URL, and hit send… then you’ve made progress ;).

  • Lawrence – i did initially mean blog, but you’re right – i want to follow individuals on group blogs without the whole group…

  • Waiting for SyndiCon I

    So BloggerCon II is scheduled to take place on April 17 this year, again in Boston at Harvard.

  • Lawrence – i did initially mean blog, but you’re right – i want to follow individuals on group blogs without the whole group…

    When i say that i want to follow individual entries, i mean that i don’t want to follow an entire RSS feed. For example, i want to find an individual entry (with a unique URL in MT or the like) and add it into my RSS so that i can know when that entry gets new comments, trackbacks or frankly, just to have a record of it. I have a lot of entries that i’ve read that i keep around for later reference without blogging.

    I don’t want to browse all RSS feeds by topic… i want to follow certain people’s topical bits and the general community who is following that topic and their topical bits. Most blogs aren’t universally topical. For example, while a lot of the YASNS folks follow my blog, there’s a lot that i say here that is not YASNS related. Well, i want to see who all is saying anything about YASNS even if i don’t want to follow their political/personal musings.

  • The problem with a del.icio.us model is that i want it integrated into my RSS behavior. I don’t want lots of tools. I want one coherent tool where data is pushed, not pulled, and where i don’t have to follow EVERYONE nor do i need to announce what i am following.

  • User Friendly RSS

    Danah is talking about what she wants in an RSS tool. There’s a lot of great suggestions in her list…

  • tracking blog threads

    Danah’s “apophenia: what i want in an RSS tool” post reminded me of a system I was thinking about a

  • tracking blog threads

    Danah’s “apophenia: what i want in an RSS tool” post reminded me of a system I was thinking about a

  • tracking blog threads

    Danah’s “apophenia: what i want in an RSS tool” post reminded me of a system I was thinking about a

  • tracking blog threads

    Danah’s “apophenia: what i want in an RSS tool” post reminded me of a system I was thinking about a

  • Bloggers as ‘friendly knowledge animals’

    When people work, they leave knowledge traces by doing things, writing things and saying things. People may either intentionally (‘smell flags’) or unintentionally (‘foot prints’) leave strong and clear (i.e. precise place) traces or weak and vague (i….

  • Bloggers as ‘friendly knowledge animals’

    When people work, they leave knowledge traces by doing things, writing things and saying things. People may either intentionally (‘smell flags’) or unintentionally (‘foot prints’) leave strong and clear (i.e. precise place) traces or weak and vague (i….

  • tracking blog threads

    Danah’s “apophenia: what i want in an RSS tool” post reminded me of a system I was thinking about a

  • What RSS tools should be

    Scoble really is a gold mine. Since he reads almost 1300 blogs, he knows many people. Like him, I’m trying to see what I can learn from the blogworld, even though I’m nowhere near that number of blogs on my

  • tracking blog threads

    Danahs what i want in an RSS tool post reminded me of a system I was thinking about a few months ago

  • Bloggers as ‘friendly knowledge animals’

    When people work, they leave knowledge traces by doing things, writing things and saying things. People may either intentionally (‘smell flags’) or unintentionally (‘foot prints’) leave strong and clear (i.e. precise place) traces or weak and vague (i….

  • What RSS tools should be

    Scoble really is a gold mine. Since he reads almost 1300 blogs, he knows many people. Like him, I’m trying to see what I can learn from the blogworld, even though I’m nowhere near that number of blogs on my

  • One of the most useful things about Moveable Type IMHO is the “categories” it lets you create. As I note here: http://blog.org/archives/cat_weblogs.html#001012 you can then produce RSS feeds of each category thus allowing you to view your friends’ posts by topic (that they define). For example if you are interested in my thoughts about blogging they are here: http://blog.org/archives/weblogs.xml

  • I don’t want to browse all RSS feeds by topic… i want to follow certain people’s topical bits and the general community who is following that topic and their topical bits.

    Do you (or anyone reading this) have suggestions about the best way to do this? Several separate issues come up, some of which are easy, some of which are not:

    1.) Each blog should produce a separate RSS feed (at a unique address) for each keyword that the site owner has created. If the site owner comes up with a keyword phrase “Howard Dean”, then every post to which the keyword phrase “Howard Dean” is assigned should show up in the RSS feed for that keyword. This is easy. In fact, with my own software, I could accomplish this before the end of today.

    2.) Get all posts, from multiple weblogs, that use a particular keyword phrase. Not so easy. Could be done if a central directory was updated by every post, like the listing over at Userland’s http://www.weblogs.com. Perhaps keywords could also be sent? Perhaps Dave Winer could be talked into accepting keywords at that site? Lots of bandwidth issues come up. It would easy to solve on the client end if only there was a place to put keywords in RSS. (This leads to a bunch of dumb questions from me: is ‘category’ in RSS 2.0 what we are supposed to use for keywords? Are most people now using RSS 2.0, as opposed to all the previous versions? Does anyone know where I could find some stats on usage patterns and versions of RSS? There was no place for keywords in .91, are people still using .91?) Then we could just screen out posts with unwanted keywords. Another issue is how to standardize the keywords in use. This could be done on a voluntary basis, or possibly the weblog industry could do what the advertising and photographic industries did a long time ago, which is come up with a standardized listing of all allowed keywords. (Or is the ‘domain’ attribute of the category tag supposed to allow one to specify which official listing of keywords one is using? Is there an official list of keywords that weblog software is already using? I don’t see one with my TypePad account, and my TypePad account is my only experience of MT).

    3.) In multithreaded comment systems, should each comment have a unique RSS feed (since it can spawn its own thread)?

    4.) Possibly a dumb question: Should the RSS feed for each unique entry also include all comments and trackbacks (I don’t know how MT handles it)? If yes, where do comments get their keywords? I ask only because if someone is screening for keywords then comments need keywords or they’d get screened out. My first hunch is that comments should simply inherit their keywords from the post to which they are attached. Does anyone see any problem with that?

    5.) Danah writes “Why can’t i have a feed of “every URL that Ronen thinks i should read”?” Easy enough to have a form on your site where friends could type in addresses they want you to look at, but then what happens to that information? You want it to go into an RSS feed to which you already have the address? Should the form be password protected to keep strangers out? Is the RSS feed public or private? I’ll assume private from what you already said. You said you wanted push, not pull, an RSS feed is pull but I assume if it is a standard one that is at an address that you already know (your software is already pulling from it), that would be close enough to push to satisfy you? Or did I misunderstand what you wrote? The ‘cloud’ attribute in RSS 2.0 is there to facillitate push, but what software is using it? Userland?

    Sorry for all the questions. I don’t use it much myself, so there is a lot about RSS use that I don’t understand.

  • Correction: it was stupid of me to say that the gathering of posts from multiple weblogs (on a related keyword) could be done on the client side. Not as things stand now. If all weblog software offered a web service that understood the ‘cloud’ tag in RSS, then searching across multiple websites by keyword would be like tracing keyword-sensitive trackbacks. And that could be done from the client (the client could register that it wants to be informed of every keyword sensitive trackback to a given post). But nothing like this exists now, that I’m aware of.

  • I tried to follow David Brake’s link to this article, but I get a “404: File not Found”. Are your archives down, or did he mislink?

    http://www.zephoria.org/thoughts/archives/001402.html

  • RSS Development

    “In the beginning was Yahoo and AltaVista…” I wonder if anyone else laments the loss of a feature that the Alta Vista once offered?

  • RSS Development

    “In the beginning was Yahoo and AltaVista…” I wonder if anyone else laments the loss of a feature that the Alta Vista once offered?

  • RSS Development

    “In the beginning was Yahoo and AltaVista…” I wonder if anyone else laments the loss of a feature that the Alta Vista once offered?

  • Tell me anyone within 3 degrees of my network who is talking about ‘rape’ or ‘domestic violence’.”

    This might be a stupid question, but what is “3 degrees of my network” in the context of blogs? When you wrote this, were you thinking of some kind of integration with other social software, like the social network stuff? YASNs? Or did you mean that, starting from a blog post that you liked, you wanted to trace all trackbacks, but only to 3 degrees? Or did you mean that you wanted to use the links in people’s blogrolls to trace people who connect to you, and people who connect to them, and then people who connect to them? Perhaps, if you meant it in this sense, something like Technocrati could be used to figure out who is within 3 degrees of you? What did you mean to be the starting point to determine degrees of seperation on blogs?

  • the unified human channel and the universal index of categories

    seb linked to a great writeup by danah boyd about what she wants in a blogger tool: If anyone wants to know why the early players get all of the attention, it’s because RSS feeds focus on people, not ideas, and the early players are too overloaded with…

  • the unified human channel and the universal index of categories

    seb linked to a great writeup by danah boyd about what she wants in a blogger tool: If anyone wants to know why the early players get all of the attention, it’s because RSS feeds focus on people, not ideas, and the early players are too overloaded with…

  • the unified human channel and the universal index of categories

    seb linked to a great writeup by danah boyd about what she wants in a blogger tool: If anyone wants to know why the early players get all of the attention, it’s because RSS feeds focus on people, not ideas, and the early players are too overloaded with…

  • hi there…I’ve been following your discussion as I have with all “problems” with RSS readers. I totally agree with your broad criticisms and they come at an opportune time. Jyte, my company’s new software, is in Beta version right now but does everything you ask, except for the “3 degrees” portion (and that’s a great idea and similar to one of the advanced features on my list).

    I’ve discussed it in length on my blog, http://www.cafemama.com – but, in brief, Jyte uses push technology combined with a search metaphor to deliver headlines and article summaries to your desktop. Our engineers have been adding RSS feeds to our database all day and by Monday the .exe at http://www.jyte.com will include an RSS reader (or so they tell me ;). Please check it out!

  • Topic based rss stuff is real cool. If anybody is interested I’ve put an experimental page you can search on at http://www.boogieplay.com/KeywordsView.jsp?keywords=bush|5:pub,kerry|5:pub,yahoo!%20news|3:pub . Basically, this page displays results for ‘bush’, ‘kerry’, and yahoo! news. You can limit the number of items returned per page and whether you’d like to share your keywords with anyone to promote knowledge sharing if you will (pub = public, pvt = private) Also, Scott’s MyFeedster may already have something like this too. Just something geeky to share 🙂

  • Bloglines offers an email address that resolves as a feed on your Bloglines list. You can use that for your trusted friends who send you links. Or, alternatively, you can forward such emails to your Bloglines address to archive them.

  • Jyte – new searchable RSS aggregator

    I’ve just been enjoying test driving Jyte

  • I’ve never used it myself, but my impression is that k-collector offers many of the features you are looking for.

  • I agree with you. Bloglines has a feature for “saving a search”. So you search all blogs for a topic and then save it. The next time you want to find entries on that topic, just click on the search link. Now, I think they also give you RSS feeds of the searches, and *that* is interesting. Even if they don’t (since I’m not sure), Bloglines is basically a feed reader so the “saved search” pretty much serves the purpose. Worth evaluating. (I’m in no way affiliated with Bloglines.)

  • RSS *readers* should serve, not just client, says I

    organize redundancy to provide both much more fully collapsed (aggregate) and much more fully exploded (multi-faceted) context-views of information (and, also, as a social network, the people / groups related by information)

  • I take it bloglines is taking the centralized server approach? Do they actively go out looking for blogs? I just went and looked, they seem to only be tracking 90,000 blogs. Out of an estimated 3,000,000? Assuming that 90% of the 3 million blogs don’t see much use, that means the core of the blogosphere is 300,000 very active blogs. To watch only 30% of the core, and none of the periphery, seems like pretty weak coverage, but that seems to be what bloglines is doing. I take it the sign up is voluntary? That doesn’t seem like a plan for getting it done. It seems like either MT needs to expand what info can get past through trackbacks or Dave Winer needs to expand what info can be passed to the “Recently posted” page at http://www.weblogs.com. Those are the two quickest routes to reaching a space where some of what Boyd is talking about would be possible.

    Of course I’m open to other ideas.