Top sources of del.icio.us links
Nivi suggested using the top sources of your del.icio.us links, in order to clean up your feedreader and throw out feeds you don’t find worthwile to bookmark links from anyway. It created some discussion on whether this was feasible/ethical/practical to build this as a service only requiring the (public) username (meaning you could probe any other del.icio.us user’s “attention stream”)
Reiterating my arguments:
- a person’s public rss feed is limited to 30 items: http://del.icio.us/rss/pascalvanhecke
- theoretically, you can screenscrape the html: http://del.icio.us/pascalvanhecke , but that would be against del.icio.us fair use (see the robots.txt file and Joshua Schachter’s mails about screen-scraping) and can be stopped by del.icio.us (see the discussion on populicio.us, another service built on del.icio.us)
From the del.icio.us point of view, that’s understandable
- every user can retrieve his or her own data in the best Web2.0 tradition and recent entries are published as an rss feed (for notification and reuse purposes)…
- but if it were allowed to query other user’s full data, the database would be leeched by the competition, and del.icio.us would loose its current lock-in (that lies in the accumulated data and the network-effect coming from the present userbase)
So the way to go was to use the api but that requires both username/pswd login for the http://del.icio.us/api/posts/all request… So if you trust me that I don’t store your password*, and take the risk of sending a (non-encrypted) password (hopefully not the same as your really important pswds!) over http… then this form might be for you:
This extra feature is on my to-do list:
- the del.icio.us export format is plain xml (<posts><post href = … ></posts>), which makes it more compact than RSS, Atom or RDF, but less readily reusable…
- an optional conversion to RSS or Atom would be handy tool for people seeking to import their linklist in other apps
Oh, BTW, it took me more than 30 seconds, but hey, I’m not a Ruby on Rails programmer…
* I do cache the fetched linklists for 2 hours however - but again, I promise I don’t have any plans with them!
Update:
I made the export tab-separated, so you could copy-paste in MS Excel… Results aren’t spread over cells as they should be however… is there a way to do this without having to save and going via the import txt data wizard?
Bookmark this post:
Subscribe to future posts:
October 5th, 2005 at 02:49
This rocks. I will write on it.
Send me your paypal, I owe you $50.
Some ideas:
1. Can you number the items on the list so I can see who is #1, #2, #382, etc.?
2. What percentage of my bookmarks are from a source with only 1 bookmark? i.e. how much area is under the long tail?
3. Use some kind of sparkline so I can see the time distribution of bookmarks for each source. i.e. have i found anything useful from this source lately?
October 5th, 2005 at 02:51
Another idea: beside each source, put links to each of the bookmarks from that source.
October 5th, 2005 at 07:30
I suppose it would be helpful if each of the sources was a hyperlink too… I think I have reached the “Land of What Must be Obvious to Pascal”
October 6th, 2005 at 22:27
Nivi, I will look into it coming weekend!
Thx for the 2-year subscription to Flickr!!
October 7th, 2005 at 03:08
Rank sources of your del.icio.us links
Top sources of del.icio.us links is a place to see the top sources according to all of your del.icio.us links.
This way you can see the sources that matter the most to you, although as said in a comment to the Nivi post:
“The problem is that D…
October 8th, 2005 at 23:06
Bloggersbijeenkomst in Brussel
Bloggers zijn doorgaans anoniemeriken. Ze leiden een tegenstrijdig leven. Ze pleuren dingen op het web die honderden (duizenden…) mensen onder ogen krijgen, maar zitten daarvoor wel uren achter een computerscherm in hun knusse huisje. Ze mogen nog zo…
October 10th, 2005 at 02:38
[...] to-be) famous Pascal Van Hecke has created the web service I seek. It rocks and I left him comments on how the service could be improved. You can find my t [...]
October 10th, 2005 at 06:12
[...] tware. Mapping Del.icio.us - Plots locations on the world map from del.icio.us geotags. Your del.icio.us top sources - Nivi had proposed a $50 bounty to who can m [...]
October 10th, 2005 at 09:12
Ouch…
So much attention and it isn’t even finished yet…
I didn’t manage to finish it during the weekend, but I’ll post an update with Nivi’s suggestions included in a new blog entry!
October 12th, 2005 at 15:04
I’ve been thinking about these security issues as well — it sucks that you have to give away your full login just to get read-only access to your personal database. This is not a del.icio.us problem, this is a problem with any web service that you want to connect to: Flickr, blog software, etc.
The problem is that there is enormous value in connecting to these services and getting/manipulating your data, and virtually all of the applications that currently are exciting today require you to give up yout login.
Service providers should give users ways to connect client software to their databases. I’ve written a bit about this here: http://dekstop.de/weblog/2005/07/connecting_web_services/ and propose a solution based on authentification tokens here: http://dekstop.de/weblog/2005/08/securely_connect_web_services/
If you’re interested in using web services as a developer or user, and most of all if you’re a service provider yourself, please take some time to read at least the second article (it’s short) and tell me what you think. We really need to fix this. We’ll have major security problems with this in the future if we don’t.
October 12th, 2005 at 15:08
Oh, and…
Pascal: looks like a nice service, but I didn’t actually use it ;)
It’s not that I don’t trust you, it’s just that I don’t give away passwords to my data. I’d try it if you let users give you the XML output of the del.icio.us posts/all API function instead, it’s simple to retrieve from within any browser. (Problem with that: it’s a lot of data, might be too much for a web form.)
October 12th, 2005 at 17:46
[...] that you’ve bookmarked pages from, so you can see which sites you find most useful. Top sources of del.icio.us links Popularity: unranked
[...]
October 13th, 2005 at 02:47
Hi there. I’ve been thinking about the cross-website authentication problems you guys are talking about. I’ve spoke a little bit about it on the mailing list. I’ll have a proposal and implementation done in a few weeks.
Joshua
October 13th, 2005 at 15:02
Joshua,
man, this rocks. I love it that you are the first to do it, because you have the freedom to do it right, and the exposure and popularity to motivate others to follow suit.
October 20th, 2005 at 23:15
[...] s data in a way not yet seen before. That was a bit the case for the really simple “Top sources of del.icio.us links” posting, it is also the case for t [...]
December 14th, 2005 at 13:58
[...] y more waiting to be added). My top bookmark sources (courtasy of Pascal van Hecke’s service): del.icio.us 14 ocw.mit.edu 13 wired.com 13 manageability.org 9 n [...]
February 4th, 2006 at 05:09
[...] Given such a cloud, I want to drop it into google/ig as a module, like this for my tags only with different data.I’m sensitive to the idea that del.icio.us doesn’t want to leave alltheir href=”http://pascal.vanhecke.info/2005/10/04/top-sources-of-delicious-links/”>datahanging out in the open - I’m not asking for databaseaccess. If they’d generate the cloud like they do the tag rolls, I’dbe just as happy. I’ll link back to them regardless. [...]
August 22nd, 2007 at 08:17
[...] Top sources of del.icio.us links (tags: tools rss delicious del.icio.us Web2.0 XML) Posted by roedward Filed in Uncategorized [...]
September 1st, 2007 at 05:15
[...] Your del.icio.us top sources- See your top del.icio.us bookmark sources. [...]