For instance: I tag a picture in my flickr stream with "yfd eat sandwich" and "yfd restaurant burger king". Then YFD acts as though those tags were tweets.
FlowingData Forums » your.flowingdata
data from rss feeds
-
-
i've given some thought to this, and i'm definitely planning on bringing in other data sources. the first though will be your public twitter stream. the second could be flickr - or maybe delicious.
-
If you made it RSS in general then you'd cover all three of those at once - and more.
Could just use title/description elements to carry the payload like the existing DM does.
Would be most bang for the buck Nathan.
-
yes, once i have rss down, it'll be a matter of parsing the xml. what i meant was that while they all might use rss, the title/descriptions/dates could mean different things across different feeds.
-
I'm not sure how it would work to input general RSS. Some RSS streams wouldn't permit us to change the title. Importing the movies that we mark as viewed in IMDb would be great for instance. Also, music we listen to on Last.fm, books we read on allconsuming.net, etc... Some yfd users might be dedicated enough and sophisticated enough to fiddle around with the RSS feeds those sites generate and make them conform to some YFD standard. I've done something similar with Yahoo! pipes. Of course, it was a huge pain in the neck.
I'm very glad to hear that you're specifically considering importing from flickr. That would be my first choice. My second choice would be LivingSocial. That would make YFD automatically awesome. "Automatically" because it would immediately add every book that I read and movie that I see to YFD. There's a lot of incentive to add this information to LivingSocial (it shows up in Facebook) and importing from there would add a lot of good information to our YFD without any extra effort.
-
Me-trics currently allows some tracking of from RSS-type feeds - I don't think they are RSS feeds, but it does allow you to track tweets, tracks played via last.fm and bookmarks from delicious. Or at least it claims it can. There is an option to track these automatically, but I can't seem to get it to work. And it should just track counts. (this is how it works when you enter the data in manually.)
-
i think me-trics stopped developed a long time ago, no? or at least that's the impression i got from their blog and twitter feed.
-
I feel bad for digging up this thread again...but...
I've been thinking about trying to automate more of my data. For instance, a lot of my online activity in google reader, delicious, and facebook can (via some manipulation in some cases) be accessed via RSS.
I'm not a programmer, and have essentially no knowledge of any languages etc., but I did mess around with Yahoo pipes and managed to get data from last.fm to yfd - via pipes and Twitterfeed. I reckon this was fairly inefficient.
I was thinking of doing something similar for my delicious bookmarks - all I really want is a count which I can track over time. Having tagging details would be a bonus. I figure I could do this again with Pipes and Twitterfeed, but can anyone suggest a better method?
I'm keen to learn a language such as python, but time and limited experience are barriers - any ideas/suggestions?
-
@transformolly - I know Yahoo pipes can sometimes be a pain to work with, but I actually think that being able to mash up Pipes, Twitterfeed, and YFD is awesome - especially for non-developers.
I'm a developer and I considered building something to pull data from other places and send into YFD, but that would take a while to build and then I'd need to find somewhere to host it, etc. And then I saw this thread. I'm definitely going to start using the same process that you're using.
It may not be the most efficient use of computing resources, but I think it's great to have a workaround like this available (and free). That way @nathany can focus on using his time to build other features.
-
I'm definitely interested in general importing facilities for YFD. I think of YFD as a "hub" for personal data; some things (like how you're feeling) can be easily and naturally entered directly into YFD, but more specialized data might be better collected in other ways - specialized UIs, devices, etc. Then they can feed their data into YFD to take advantage of its ability to visualize, correlate, etc., without bloating YFD.
Two examples I can think of are time tracking apps (like Bubble Timer) and calorie tracking apps (like LiveStrong's "daily plate"). Both of those offer features that are above and beyond what YFD would probably want to have, like the ability to show your food intake broken out by nutrient (calories, fat, sodium, vitamins, etc). Pulling high level data from that into YFD would give you a single place to view your data, while letting the more detailed data live elsewhere. Of course, the interfaces would have to be carefully designed, but having server-level integration of these other services (and many others) into YFD would make a pretty compelling addition.
-
thanks, ivarley, i've been looking how i might integrate other sources in a way that's useful. certainly something i'd like to work more on this year.
Reply
You must log in to post.