Note: I spent a bit of time this morning editing and cleaning up this post because upon re-reading it—which as usual was well after I published it—I thought there were a few things I could clean-up and clarify. That said, I haven’t changed any of the ideas or the overall argument which is basically: Y U NO MAKE SYNDICATING RSS EASY.
I don’t necessarily disagree with Alan Levine that ds106 shouldn’t “come in a box,” especially given the bulk of what makes that course/community great is the people—and there is no packaging that. However, in order for others to use the syndication framework that ds106 runs on—and that’s more important for folks not particularly interested in ds106 in and of itself—we still have to try and make automating blog syndication more accessible.
FeedWordPress has been the key ingredient for that in WordPress, and it has been pretty well automated using the premium plugin Gravity Forms and some other hacking by Martha Burtis. That said, without that plugin and hack any faculty member trying to create a syndicated blog for his or her course will end up spending the first few weeks manually adding and tagging feeds in FeedWordPress. That’s just not practical. Making this easier is crucial, even if it is not perfect. In fact, it’s probably more important than a new version of the assignment bank, which I think is a cornerstone of ds106. Providing a simple way for people to sign-up their blog feed (and Twitter, Flickr, etc.) and have it automatically syndicate is the basic model of eduglu we’ve been approximating for years. While it was never gonna be perfect (what is?), I sure as shit didn’t think we would still be floundering around manually adding feeds the way we are.
Without being able to automate adding feeds to a site like ds106, it takes a tremendous amount of labor on the part of a professor to make a syndicated course work—with or without a community. I don’t think this is “in-a-box” as much as it is a necessary streamlining of the syndication framework which is long overdue. Fact is, sometimes a basic structure for shelter, like a box, has its uses in a case where a faculty member wants to experiment with a syndicated course structure but doesn’t want to have to be farming feeds for the first three weeks of class to do it. There is a balance here when it comes to the idea of making a framework like ds106 more usable for a wide-range of people. To suggest it’s all about the community might miss some of the basic elements a faculty member needs just to get started with such an experiment.
How many of the ds106 partners are WordPress sites?
Of the ones I know doing this (ETMOOC, RMOOC, ds106, etc.), all of them are using WordPress. The real solution in my mind would be finding an abstracted tool that does the syndication on a pubsubhub level (I think gRSShopper is one example of this). This is kind of what I was trying to knock around with Kin Lane and Audrey Watters in April. I know you have been the beginnings of what Brian lamb has imagined as an edtech co-op, and you even started down that road. Imagining a way to abstract the syndication away from WordPress as a self-defined tool that could be pulled into any system that accepts a feed would be even better. If gRSShopper is quick and easy enough to do this great, but when I played with it a year ago I was lost (which speaks to my limits). That said, the monolithic presence of WordPress in the OSS world is not necessarily an advantage anymore, in fact, it could be argued it’s becoming an issue—and I would hope this model for syndication abstracted out from the publishing tool could move beyond one application. I’ve used WP because I’m stupid, that remains true, but I am also smart enough to know the more we consolidate around one solution, however slick and easy (and to its credit WP is that), the poorer the commons becomes in general.
Have you looked at Tiny Tiny RSS? I use this as a Google Reader replacement. It harvests feeds and can republish them too. It may be way off what you need, but it’s open source and can both aggregate and syndicate. It’s a brilliant feed reader if nothing else.
Some random (and somewhat brief thoughts) having played with WordPress/FeedWordPress for ocTEL it was interesting to hit the limits of RSS availability for 3rd party sites. We started down the rss aggregation route because it was easy to ingest student blog posts into a central wordpress site. The thing was we quickly discovered people were frequenting other spaces, getting up their own mini communities and still doing this in the open using things like G+. We felt it was important to tap some of this gold.
If I was starting from a blank sheet I’d be looking at creating a bridge that can pull data via json and rss with some common connectors to 3rd party apis (twitter, delicious et al.). For storage I’d be looking at NoSQL solutions just because of the messiness of it all and I would rule out 3rd party data storage using something like the Tin Can API. How you interface all this is critical and the option to self host should always be there (I recently looked at http://www.edufeedr.org/ which appears to do gRSShopper type things – all I wanted was an OPML bundle damn it https://twitter.com/EduFeedr/status/302113854920462336 …
Even abstracted out of WordPress you are dependent on some set of tools/code to wrestle with as you need something to poll requests, some RSS parser library, and some database management storage- so it could be gRSShopper, Drupal… or maybe I don’t see fully how this can operate at some data level you guys talked about.
That said my post was not meant to say that there should not be automation or things to simplify, just that I hate to see the mindset that we can boil down a complex system or design to a one button operation.
As far as the WordPress situation it’s not complex– to add a feed to FWP you need something that creates a WP link item and modifies the notes field. I think Martin Hawksey did this in his octel site by adding two custom fieldsto user profiles. When they create an account as a Subscriber, a hook looking for changes in the blog URL field does this link creation… And I thought you also had that gong on UMW blogs with that widget that allows users to adda link (??).
I’m more than game to better understand how a standalone system might work.
Do check out Tiny Tiny RSS. It can publish RSS feeds of collections of feeds as well as publish a public OPML feed. It’s a lot like Google Reader and dead easy to install and run on a LAMP server.
I’m still in search of a good RSS reader, so I’ll check it out. How do you feel about the work Alan and Martin have been doing to integrate a reader into a course framework like this? That, too, is a part of this ecosystem, and if you are showing students how toc reate their own blogs on their own domains, might installing something like Tiny Tiny RSS be another element?
I love what you are saying about 3rd party APIs, because Kin lane was saying the same thing. Given how many services basically turned off their RSS feed at this point, we need a more sophisticated method at this, which is both a shame and a necessity it seems to me. The more everything is designed for an app, the more you need coders to bringing it back to the open web in an open format. If you would be interested talks we are having about this as part of the Shuttleworth foundation flash grant I have, I will bring you in. What’s more, I think you and Kin lane would have a field day. If I actually get the larger Shuttleworth Foundation grant, I think the idea is try and fund as much of this as possible and get people paid to build it. That is the other crucial piece. I guess Lockley will build he open source Tumblr, isn;t that right @Pat? 🙂
Your post got me thinking about just these things because the work you are doing with the assignment bank is crucial, and I was thinking so would this be. And while the in-a-box mentality can be damning, I am also thinking about the population i work with at UMW, that are engaged and ready to go, but would have to balk at the idea of managing 120 feeds (which spread over 4 classes is about right) for a syndicated framework. What’s more, if we are trying to push this out to an entire community at UMW, which this Fall we will be, we need a way to make it easy for faculty to pull in the feeds from their student’s blogs.
This is not going to be the extreme case at UMW anymore, but part of the fabric of how courses are taught as Domain of One’s Own comes online. Give that, to introduce the concept of syndication will be enough of a mental jump for some faculty, so making it relatively seamless seems essential. I know we are close to making it somewhat seamless in WP, particularly given 99% of our faculty and student here will be using WP, but at the same time I’ve never had enough skills to jump that last obstacle and make it seamless enough that I can show a wide range of faculty members how to do it on there own in some automate fashion.
WordPress is now effectively the default “I need a website” tool, so FWP works almost because people can get WordPress easy. This is almost the innate dilemma, build for WordPress as everyone has it, and your tool gets lots of traction as people can use it.
Drupal has native support (as part of core) for feed aggregation, but drupal is much much harder than WordPress to use (although it is getting better) and any alternative solution is reliant on some techincal issues to install.
This to me gets into a problem over hosting. Do you build a central service and remove the install problem, or create a distributed solution and leave those to find their own solutions?
I also think the FWP solution is reliant on RSS, whereas some of the time other platforms don’t make RSS, or even any subscription options. Perhaps some variation on scraping might be a better bet.
But you effectively cede power to WordPress in doing so, and I worry (and have talked a lot to Christina about this) that Open is great, but democractic might be better.
Pingback: Syndicate This: a Practical Dream Vision of Syndication for UMW’s Website | bavatuesdays
“In a box” is a bit of a red herring I think. What “in a box” means in practical terms is that I spend most of my time engaging in activities that are meaningful to me. So Reason 6, which I use for music production, is a “Music Studio in a Box” with which I can record a song very quickly without worrying about writing code to find a MIDI device, or having to futz too much with the auto-compression. At the same time I have to write the songs and record the vocals, and to get a sound that is unique I’m going to have to learn to tweak the dials. But for the simplest use case I spend my time making music instead of manually coding sidechain compression, debugging VST load crashes, or searching endlessly for third party components.
And I think that’s the point — when an activity is new, it’s really impossible to know what the simple use case would be. But as an activity becomes mature enough that we have a sense of what the most common use cases are, we work to make those easier, while trying to leave room for people that want to tweak the dials. And if I venture out to tweak the dials, I understand precisely that it’s because I’m looking to do something unusual — get a unique sound, fix a bum recording, alter a sample.
There are people out there that claim that not writing the tag in HTML divorces you from something important. In general if I’m doing something simple with bolding, blockquotes, and an embed I use the visual editor. I get that I have to go to source view if I want to make my data table pretty or put a non-standard border around an image. But I think Alan probably gets that, and in reality we’re not as far apart as you think. This thing you’re doing isn’t as mature as blogging, there’s wide variation in it, so it’s not going to feel quite as simple as setting up Tumblr. At the same time it shouldn’t feel like setting up Apache circa 2002. And the multiple feed management piece of that is a huge huge deal with that, because adding and subtracting feeds never feels like an activity that is adding value, it feels like drudgery.
On the plus side, this multiple feed management thing is something people like Jon Udell and Tony Hirst have tackled before — and is seems like something of wider applicability than student blogging. I wonder if we could think a little more generally about an aggregation tool that doesn’t reside as a plug-in, but as a standalone visual feed aggregator that starts out with a broad term search. Other people might be willing to help with that?
I couldn’t agree with you more in this, and given that Udell and Hirst have done this on a wider stage makes me think we need to be pushing for just what you say—a broader solution that stands outside of any one publishing applications, but can hook into any of the open, supported ones.
One of the things that has struck me when I was talking with you and Udell in Keene more than a year ago was how this stuff needs to be more broadly applicable than just education, or just ds106. I think you can provide the models, but then it takes resources and a real sense of commitment to actually build open, easy to use tools that can help people see the value of this stuff. More and more, that’s what I want to do.
@Jim I tried tapping Shuttleworth for this area last year but the ramblings of a mad man didn’t get very far 😉 http://youtu.be/FJGs35DAsqY
Happy to chip in (if you are also looking for named partners I’m doing some more work with the Association for Learning Technology)
Knowing you didn’t get that grant makes me even more concerned about my possibilities. That said, whether or not this is a Shuttleworth grant, like I said above to Micke, I want to start getting some momentum towards framing out a set of resources/tools that help people syndicate, archive, and visual the distributed networks they work within. And this is just what you were pitching to Shuttleworth, I have to believe there is someone, somewhere who is interested is funding this work and putting some people together who can begin to make it happen.
@Jim – I don’t think you should be that worried – it was a lonely voice of a desperate man about to lose his job. Given the diversity of people starting to congregate around this idea I’m sure someone will see value in funding
Pat — I completely agree with you (I think) — FeedWordPress is reliant on RSS, and I think that will increasingly limit its applicability. My guess is any solution that is going to survive is going to have to take the If This, Then That approach of making a unified interface that actually hides wildly different API calls out to different systems on the back-end. RSS, for a number of reasons, is going to be reduced to a bit part player over the next four or five years, whereas API calls are likely to stay around longer.
One way to think and pitch this is from the workplace backwards. How are our tools getting in the way of what we do at work (or in our creative collaborations, or socializing)?
That’s a hard question to answer. The rise of the twitter/tumblr model of life-streaming and the fall of website hubs is a complex phenomenon which happened both because it addressed deficiencies of the website hub system and because it was more attractive to an advertising based model. Broadly, the twitter/tumblr model benefits from:
* Integrated publishing and reading. Where I go to publish my thoughts and links to my stuff is where I also read people’s stuff, and I move fluidly between reading and reacting.
* I own my comments. Which means all the comment drama ends, and also I control and own my reactions to things. It also means I can’t ride on the circulation of a blog to trumpet my ideas to people I haven’t built a relationship with
* Meeting new people is easier. Retweeting/Reblogging as reputation propagation is actually a better solution than a blogroll in most cases.
These things — which are part of what is killing the syndication dream — are actually good.
The big tradeoff is that you’re stuck in their system, which for various reasons makes your comments somewhat ephemeral, takes control over your material away from you, and makes communicating with people in other spaces difficult. And for personal use this tradeoff isn’t so bad really so bad really. My daughter is fine belonging to the tumblr subculture of Superwholock kiddies (it’s a thing, look it up) and that’s fine. She doesn’t really need to connect to the Superwholock people on Twitter, or the Superwholock blog carnival (remember blog carnivals? Good times.), or whatever. It’s like hanging out at your local pub. Who’s there is who’s there, and the exclusivity of that is often more feature than bug. In fact, when Katie says she’s going to go do Tumblr for a bit, I hear it the way I’d here “I’m going down to the bar, be back in a bit.”
So it’s really not bad for socializing. But for research or business or education that seems insufficient. You can’t really be an academic that just exists in the “Tumblr water policy subculture.” I mean, I suppose you could. But it seems problematic to me. I meet you at a conference and we’re fast friends, and potentially great colleagues. But I say OK, find me on Tumblr. And you say, oh actually I’m on G+. And the chances of us maintaining that fruitful relationship drop precipitously.
Back in mailing list days there was a solution to this. I had my email, and I could use it to subscribe to listservs or newsletters of multiple platforms and flavors. Like twitter and tumblr it was a read/write tool — in the process of catching up on news I was a reply button away from participating. If I said to a colleague — oh, you can find me on SHAKESPEAR-L, they wouldn’t say — oh I don’t do SHAKESPEAR-L, because the email harness made it easy to add that, maybe initially as a digest. (It also made it easy to feed a post sent to SHAKESPEAR-L to the LINGUIST list and propagate posts across systems. If faculty had to sign up for an email based forum for support, they didn’t say “Another fricking community I have to join?” — because again, it was all in the same harness.
And consider this — you had a hard drive based copy of everything you published and received in a standards-based format.
One of the reasons email survives in business is because it’s still the only way to do this. The problem is that it hasn’t incorporated features of social media that help us filter , manage, and discover connections and content. I suppose this is why I was embarrassingly supportive of Google Wave, because it seemed an attempt to carry forward the strengths of email into the world of social media.
Anyway, I’ve started to ramble (wow,, how did I get here?) but I do think the pitch is broader than education, and the issue is how we benefit from many of the good things that twitter/tumblr do while retaining ownership and not needlessly proliferating harness applications or balkanizing effort. That’s the question that syndication answers that both businesses and education are interested in, and the question I think you’d be good at raising, Jim.
Although API’s may well be the direction everything is headed, they also bring to the table the notion of control in a way that RSS doesn’t. Some of the API’s are designed and restricted so as not to facilitate sharing of content across platforms. Some API’s exist so walled garden operators can say they care about interoperability with a straight face.
A system which, invisibly to the user, makes calls across API ecosystems is also going to be very much not DIY. well not for anybody but Kin Lane, perhaps. It almost makes me want my c 2002 Internet back. Almost.
Oh, agreed, I very much like the RSS scenario more. But for precisely the reasons you cite I think it will cease to be the dominant scenario. To an extent that to do RSS only will be yet another barrier between parts of the web world.
As far as DIY — I see any tool to do this as the tool that makes DIY possible, just as IFTTT allows me to DIY crossposting without knowing a thing about APIs.
IFTTT and Yahoo Pipes type services are a great way for people to start creating their own flows. One issue is they become easy targets for services like Twitter to start putting pressure on what services you provide e.g. Twitter search pulled from IFTTT. One of the features I really like about ds106 style communities is they are distributed. It’s like guerrilla ed., rip and run. So for me it’s less about having a central service and more about a box of tricks anyone can deploy on a webhost or free services (tricks like generating RSS feeds from your gmail inbox 😉
This is a really timely post for me. As you know, we’re running our own Domain-style pilot this fall at Emory, but for a variety of reasons we at least started out without being able to get the server space worked out (though Reclaim Hosting is shifting our paradigm some, I hope). So, we’ve got some faculty using WordPress but most of them are creating sites for themselves and expecting to have their students create sites in the third-party service Weebly. It’s a really nice system for designing static web pages and had blogging capabilities. It will do some cool things for our faculty.
However, now they are starting to try to really wrap their heads around how to let the students publish to their own individual sites but then syndicate those posts to some central course site as well. I wish that there was a way to do this native in Weebly. For now, the solution I’ve come up with is to create a WordPress.org subdomain on my own little multisite installation just to run FeedWordPress for them to aggregate the posts their students put up. This solution is less than ideal b/c most of them are already trying to learn their way around Weebly and are creating websites for the first time and they get a little antsy at the idea that they will have to also use WordPress at the same time.
I’ve only just glanced at Tiny Tiny RSS but that requires server space too. I’ve thought about IFTT too, but I think that will seem even more difficult for the faculty to set up? I’ve only just barely used IFTT myself though.
Sorry for the bump on this, but I’ve had the tab open for ages and so deal with it.
I think (admittedly the blog post is about syndication) – but if you built a new system then I think you need it to be verb free – WP, Drupal, Tumblr, Twitter and facebook have shown how the system defines the language, and that defines the system, and what it can and can’t do.
Now this means two things. Let’s take this blog as an example – the comments are now longer than the blog, and some of the comments make sense, perhaps inline or added to the blog. But in WordPress there is a blog. Imagine if this post had been a wiki and each comment was an edit. But then that’d be wikipedia right. So the system shapes the content.
So syndication means the system needs to know what it gets. But what if say, you syndicate all the flickr pics on a tag, and make a gif or movie out of them. Then content can be crazy. Each blog post becomes a PDF page which is emailed to people.
So you start to see content as a bit more interesting, so the comments are content as the blog as the content, as are the pictures as are the links.
To conclude, the revolution will not be televised, because in the future, no verbs.
Pingback: What I’ve starred this month: August 28, 2013 MASHe
Pingback: Horses for open courses: Making the backend of a MOOC with WordPress #altc2013 MASHe
Pingback: Where Did I Put That Conversation? -
Pingback: TIny Tiny RSS | bavatuesdays
Pingback: ds106: Will Work for Feed Syndication Framework | bavatuesdays https:///ds106-will-work-for-feed-syndication-framework #reader #openpedagogy #plumbing