Scott Leslie recently wrote about using Rollup to put together a super feed of his favorite educablogger’s furl and deli.icio,us feeds:
lots of folks have separate Furl and del.icio.us sites/feeds. I’ve been subscribing to one or two of them in the past, but wanted to get all the ed tech bloggers’ bookmark feeds in one place. So off I went to Rollup.org, where I created a new RSS feed that rolled up the Furl or del.icio.us RSS feeds from Alan, Brian Lamb, James Farmer, Greg Ritter, George Siemens, Trey Martindale, Harold Jarche, Will Richardson, D’Arcy Norman and myself. I would have added more, but these were all I could find.
So the handy thing about this is that I can subscribe to one feed in my bloglines account and see all the URLs collected by all these brainy folks.
which is syndicating from http://rollup.org/rollup/rollup.php?id=495
So we can say Scott is doing some new ripping and mixing of feeds, stuff we like to see. I’m not very familiar with what RollUp does– it sounds like what Blogdigger spins out as well- the nifty thing Blogdigger does that no other web-based RSS tool does is to save a cache or history of past entries that have come through– do old items in Rollup fall off the edge of the Roll? I think Bloglines keeps a record as well (true?)The downside of Blogdigger is the lag sometimes 24 hours, for new content to appear.
The other thing I noticed peeking at the source is that Rollup seems to mal-parse the Furl pubDate fields (you can guess that is happening if dates appear as December 1969) and seems to mis the Furl descriptions tags… but wait a mjinute, I peeked at the source of my own Furl RSS and see that Furl lumps many things it;s own RSS 1.0 custom fields (which is perfectly XML valid) but note that odd </description> tags that have no opener:
<item>
<title>Radio X-Factor: Podcasting University</title>
<link>http://www.furl.net/forward.jsp?id=1374192</link>
<description />
<category>social tech</category>
<guid isPermaLink="true">http://www.furl.net/item.jsp?id=1374192</guid>
<pubDate>Fri, 26 Nov 2004 16:49:12 GMT</pubDate>
<furl:rating>3</furl:rating>
<furl:clipping>Think of Podcasting as something like a
Tivo for your iPod (though it isn't exactly like that).
</furl:clipping>
</item>
which looks a little less than kosher to me.
I am trying to make a point that mixing and recasting RSS can be a tricky thing, and have to admit that Feedburner does a rather elegant job of translating different RSS formats to a clean format. Actually my use of Feedburner was purely an experiment to what they did- I cannot say it really has a grand penultimate purpose.
That said, my using Furl has really not been in the interest of publishing anything important to others- it is my own source for tracking relevant sites that I can come back to and search later. It is purely selfish, It does seem a bit weird at first that others would be interested in my furling, as it is pretty much without pattern– weeks of intense furling to ones of no action at all.
And then I even use del.icio.us to a lesser extent, more out of a curiosity and some experiments to find links to some more off beat corners of the net. I am more interested in the mixing and matching of collective bookmarking both here, and the amazon-like recommending feature of Furl (“People who furled this site also furled….”).
So although Scott lumps me in the “brainy” category, I pretty much do a lot of tapping into other’s links and tagged sites as well. But there is also a huge amount of serendipity when you find interesting links one or two clicks off of these initial forays.
Heya, Alan! Glad you’re back safely from the realm of Weta.
The description tag is actually valid. That’s a self-closing xml element. It will parse and validate OK, but doesn’t contain any data, so it’s basically just a placeholder. Often, tools will automatically turn empty elements into self closing ones to save some characters in the file…
<description />
is the same as
<description></description>
Thanks Dr. D- That makes sense and shows my limited knowledge of the innards of XML.
Alan, thanks for mentioning Blogdigger. We’re working on improvements to our crawler that should speed up the indexing of posts (favoring sites in groups), and you can always ping Blogdigger when you post, which will significantly speed up the time to index.
Thanks Greg, always attentive…
When you say I can ping Blogdigger- is there an actual ping address for my group or or you just saying I ought to ping your blog??
Also, is there a limit to how much past feed content that Blogdigger will archive??
And is there anything in the works that might reduce duplicated items from multiple sources?
Lastly, your group login page really needs a means for recovering passwords 😉 I cannot find mine.
Regading pinging, for MovableType, if you go into your Weblog Configuration, under Preferences, I believe, you can set your blog to ping weblogs.com, blo.gs, and any other ping services. If you enter http://www.blogdigger.com/RPC2 in that box, whenever you enter a new post, an update ping will be sent to Blogdigger, and we’ll push your blog to the top of the crawl queue.
We have no limit to how much data we store. We save every post we crawl.
What kinds of duplicates are you seeing? We are taking measures to reduce the duplicates in our index.
Email me a new password, and I’ll update it. An update with password resetting should be coming soon-ish.
That approach to pinging is fine if I want to have Blogdigger index just my blog– but the approach I was thinking of is for a group collection of feeds from elsewhere- I cannot force them to ping.
Damn, D’Arcy beat me to the XML clarification. I was all ready to flex the XML knowledge, but I’m not going to get into a brainiac pose-down with D’Arcy, though — that’s a losing proposition. 😉
I will kvetch that although the empty-element tag is valid XML, it’s pretty gratuitous, since leaving the DESCRIPTION tag out altogether will have the same result as an empty DESCRIPTION tag . . . and leaving it out would conserve a few bytes of bandwidth!
I agree with you about Furl & Del.icio.us, Alan. I use Furl primarily as a personal web archive — Furl’s killer feature is that it archives a copy of every page you furl. I hate Furl’s categorization, though, especially since a good search obliviates the need for a taxonomy (cf Google’s dominance over Yahoo’s human-managed directory).
I’ve taken to using Furl keywords like I would Del.icio.us tags, and just searching against keywords or full-text. More effective for me, but less effective for people interested in what I’m furling, because keywords don’t show up for other users. Hence everything in my Furl archive is in the “General” category.
Unfortunately, Furl keywords (or even categories) don’t provide the social aspect of Del.icio.us tags. Someone who combines the social aspects of Del.icio.us with the archival capabilities of Furl is going to have a killer app.