Adam at Darwinian Web is calling for a memetracker community in the form of a group blog authored by memetracker developers. I definitely want to understand better how the memetrackers work and why they often don’t work, as in the case of many domains that interest me. But what I want even more than an understanding of the current set of closed possibilities is an open-source news/conversation/topic/meme tracker platform, because I have lots of custom things I want to do to track conversations and topics, none of which are well supported by the current tools. Yet like I said before, I don’t want to have to start from scratch. I like to mess around with templates and rules and filters and I’ll write a plugin module or two, but I don’t want to architect an entire system from the ground up.
I think I may display a whole lot of cluelessness in this post, but Kathy Sierra says it’s a good thing, so I’ll press forward.
I’ve played around a bit with the OPML Editor, and I can’t say I’ve accomplished much or even enjoyed it. I find it frustrating to use a tool that keeps me away from the code (i.e., the OPML) itself. I don’t fully understand its architecture–it seems to be written in C using something called the Frontier kernel. Perhaps I just need to spend more time with it. I plan to watch Lisa Williams’ video on building a reading list with it. Hey look! She put in a bunch of mom blogs. Thank you, Lisa. Seems I’m not the only one who sees more to the Web than tech, politics, and celebrity gossip.
Happily, the OPML Editor community looks to be thriving. Even though I’ve been too scattered lately to spend much time with it, other people are more focused. Like Les Orchard who’s building an OPML community server using PHP. Yay! I’ve been wanting Les to write all his stuff in PHP. I hope this means that extensions to the OPML server/editor combo can be written in PHP rather than Frontier… Now, Les, please add in your popular links code and the Bayesian classification capability. How about some keyword filtering too? I like the sound of the WordPress integration; transitioning to WordPress for my blog is on my to do list.
Could the OPML Editor and Server become an open-source conversation tracking platform?
Now to some cluelessness, if I haven’t already displayed enough by my assumptions and suppositions: isn’t it a shame that OPML requires a whole different server than HTML? This reminds me of Danny Ayers’ comment on the Alex Barnett-Joshua Porter reading list podcast, where he suggested we don’t need OPML to capture lists of feed subscriptions because HTML could do the job just fine. However, the lingua franca of news aggregators is OPML, so I guess we’re stuck with two servers instead of just one, or at the very least two XML languages. Does the OPML Community Server serve HTML too? Or just OPML?
I’ve been spending too much time with closed-source solutions these days, without much success. I’ve made stabs at getting filtering and aggregation I like with Megite, TailRank, Findory, and Personal Bee. They all have their strengths but none of them meets my needs and I don’t have much ability to customize them or influence what features they implement. I think I need to do some custom development. I think I’ll spend some more time with the OPML Editor starting tomorrow.
7 Comments
Anne, just wanted to say: Don’t underestimate your ability to influence the features we implement. As to customization — no, we are not there yet.
Hi Anne,
You know I’ve been quietly reading your posts about OPML and thought “she gets it and I don’t” but thanks for this post. I’ve found OPML Editor really frustrating and confusing and found it so much faster writing the OPML code myself. I wondered if I was missing the point or something.
I’ve just (finally) published a post about this. Would be good to talk more!
R
Dan: I’m not giving up yet. PB is the closest thing I’ve found to what I want, as it offers a platform for building a newsletter-type presentation of hot/interesting articles meant for a wider audience than just one. I don’t want merely a page of articles just for me; I want a set of articles that would be interesting to anyone who wants to think and read regularly about a particular topic, like balancing work and family.
Though PB gives me only the barest hint right now of what I want, I can envision it going where I want to go.
Since I’m thinking about OPML… it would sure be nice if PB could work off an OPML file hosted at a URL so I could edit the reading list e.g., in Blogbridge and have PB automatically see my new feeds.
Rachel, that makes me laugh that you were thinking I got it and you didn’t, because I’ve felt so dumb when I used the OPML editor. I left a longer comment on your post… thanks for it, because I felt so much better after reading it. Woke up this morning thinking “I can’t believe yesterday I admitted how ignorant I am about the OPML Editor.”
I was insanely frustrated the first night I tried using the OPML Editor. I finally got the RSS reader working a couple of days later (after taking a break to cool off) but now I don’t really see the point. It’s just XML, right? May as well use Emacs or TextMate.
I also don’t understand what the advantage is to using a special OPML server. Wouldn’t XSLT cover any display needs well enough on a regular webserver? It seems unnecessarily confusing to introduce a new platform for a single purpose.
I think OPML is very useful as a way of describing lists of links (especially since it’s become the common format for exchanging data between newsreaders). But I’ve been reading a lot of things lately that seem to put it in some special category or box (requiring special handling), and I don’t see the point in that.
Audrey, I think you’re right, although my understanding of how XML is handled by standard web servers is limited. I didn’t even consider that aspect. Yes, OPML is just XML so why all the special handling? I admit, I don’t totally get it.
Anne: That’s a great suggestion re: the “remote” OPML file. That would actually solve an issue for us, too, because we know people want to work with their existing files. Earlier, we offered the ability for users to import OPML files, but backed off that because frankly, we’re not sure we can handle the aggregating and parsing chores implied by having people dump in 1,000 new feeds at a go. In any case, I’ll pass this on and get back to you on what the prospects are. …