May 31, 2003

Looking for bloggers who went looking for audiences

Today I got a call from a NYT reporter who is looking to find people who have started weblogs, written for a while, and had a hard time finding an audience.  Are you one of those people?  Ideally, you should have tried a variety of methods to gain an audience, and after trials and travails, you now have a regular readership.  If you're interested in being interviewed to find out your experiences, leave a comment below describing why you would be right for the article, some of the things you did that didn't work, and some of the things that worked.  Don't forget to leave a way to contact you, as the reporter will want to follow up with a number of people.  No promises that you'll get a call, but this does sound like an interesting upcoming article.

May 14, 2003

Welcome to the world, Noah!

dscn5286.jpgYesterday, May 13th, at 7:21am, my son Noah Joseph was born.  He weighed 8 lbs 10.2 oz, was 21 inches in length, and I'm about as happy as can be.  Mother and son are resting comfortably.  More pictures are forthcoming, but here's one of his foot right after birth, check out those wrinkles!  Big sister Melody has already met him, and she's already planning tea parties.  If you've tried contacting me recently, I'm gonna be somewhat unresponsive for a few days.  More later...

May 12, 2003

Technorati API 0.9

I'm proud to announce the first public release of the Technorati API, the application programming interface to Technorati's weblog index and search engine.

Over the past few months, I've gotten a lot of requests from people who wanted to be able to use the Technorati database for a variety of purposes - everything from social network research to mini-applications that would send them a page or an IM whenever someone posted a link to their website.  I created the Technorati API order to help foster these creative ideas and developers. I've also got a whole bunch of other work going on, so I didn't want to become a bottleneck between you and the data.

One thing to note: This is beta-level software right now.  Nothing is set in stone, but the code is working, and a number of people have already given some great feedback that has led to changes in the API that make it significantly more portable and forward-compatible than I originally designed it.  I'm planning on adding additional attributes and functions to the API.  Things may change again if someone finds an important structural issue or bug, so you may want to wait a bit (until the 1.0 version comes out) to build big enterprise applications with it. :-)

First off, regarding licensing:  I'm making API calls free for personal, noncommercial use, and you'd be capped at a certain number of requests per day, currently 500 queries a day. The draft terms of service is up on the website.  This is pretty much what some other folks are doing with their API.  I would also like to get some attribution added that points a hyperlink (and maybe a small picture) back to Technorati for people who use and republish the information, sort of like a Creative Commons non-commercial and attribution license.  Of course people who pay money will get to use the data without these restrictions.  If you're interested in licensing Technorati data for commercial applications, or if you want more than 500 queries per day, please send an email to

Skip the next 3 indented paragraphs if you're not a super-geek, and read on, starting at "So here's the initial API calls that I've built"...

        The API calls would be built (at first) with a REST-ful interface - a standard web-browser-like program (curl, wget, lynx, etc)would pull down the info, something like: and so forth.
        If enough people clamor for it, I can do a SOAP or XML-RPC interface, but REST-ful is easier for me to do, and I don't think people need the multiple-way communications channel that something like XML-RPC or SOAP gives them in addition to the REST interface.  In other words, each API call will be stateless, AFAICT. Besides, I have never written a WSDL file :-)
So here's the initial API calls that I've built:

API call #1: Cosmos

GET format: ...

POST format:
with the variables and values returned in an HTTP POST call.

This returns the Link Cosmos for the URL you specify.
Set the following variables to get the results you want:
url: Set this to the target URL you are searching for.
type: Set this to 'link' and you'll get the freshest 20 links to your target URL.  Set it to 'weblog' and you'll get a reverse blogroll - the last 20 blogs who linked to the target URL.
start: Set this to a number > 0 and you'll get the start+20 freshest items (links or blogs), e.g. set it to 21, and you'll get the second page of rankings - items 21-40.

format: This allows you to request an output format. I anticipate we'll have multiple output formats, like XHTML, RSS (different flavors), and whatever else comes along.  For now, the only valid format accepted is "xml", which adheres to this DTD.  The format variable is optional - if you don't specify it, it will default to "xml".

version: This allows you to specify a particular Technorati API version response.  Right now, there is only one Technorati API, called version 0.9, but I anticipate newer versions, some of which may not be backwards compatible with earlier versions.  By implementing this variable across all Technorati API calls, application developers can feel assured that they will always get reliable results if they specify a particular API version for requests and results.  The version variable is optional - if you don't specify it, it will default to the more recent version of the API, which at this time is "0.9".

key: Put your Technorati API key in here.  You can get your API key by going to:
If you're already a Technorati member, just log in and your key information (along with some accounting) is listed for you.  If you don't currently have a Technorati account, setting one up is easy and free, just follow the instructions on the signup page.
For example: This call (copy the link location to see thesyntax) will get you the Link Cosmos for the website (including my blog, Sifry's Alerts).  That API key is my key, so please don't abuse it, get your own!

API call #2: Bloginfo

GET format:

POST format (see the cosmos call above for explanation):

Give it any URL and it'll tell you what blog, if any, that URL came from, and all the info it has on that blog, like cosmos stats and RSS feed, etc.
For example: This call(copy the link location to see the syntax) will tell you information about the weblog that created the following permalink: (BTW, the answer is Sifry's Alerts)

API Call #3: Outbound Blogs

GET format:

POST format:

This call will list for you all of the blogs that are pointed to by a blogger (well, most of them anyway).  Simply give the URL of the weblog and it will return for you a list of weblogs that are linked to by that blog. (The database is currently a bit old on this one, I'll move it into production status over the next few weeks)
For example, it is a relatively simple call to iterate and get a 2 degrees of separation function from this, which I'm working on... ;-)

Finding out more

The place to go to find out more information on the Technorati API is the api-discuss mailing list.  You can subscribe by sending a message to or by going to the mailing list page above.  There is a web site that I created just for developers, called but right now it is pretty sparse. I'm looking for volunteers who would like to keep that site updated.  Right now I just don't have time.

If you've gotten this far and you're still interested, come join us!  Subscribe to the api-discuss list.  Get an API key, and start playing around.  Leave a comment and let us all know how you're using the API.  Have fun!

UPDATE: The API hasn't been out for a full day yet, and already there exist plugins and drivers for Radio Userland(Dave Winer), Python (pyTechnorati (Mark Pilgrim), two different's - Phil Pearson's and Aaron Swartz', and at least three Movable Type plugins - Ben Hammersley's unreleased plugin, Joi Ito's SSI-based plugin, and Adam Kalsey's publicly released plugin!

Amazing. So here's my next thought: Dave Winer and others brought up the idea of Link Cosmos as trackback - and here's the simple equation that popped into my head:

Technorati Link Cosmos on each of your blog permalinks = instant trackback... Of course, you'd have to pull the information several times a day rather than being event based, but it wouldn't require any agreement between the linker and the linkee, and no messy XML-RPC calls involved... I'd love to see that in an MT plugin, I'd put it up on my blog!

May 8, 2003

Google to create blog search engine?

Reuters and The Register are reporting that Eric Schmidt, Google's CEO, announced at the JP Morgan Technology and Telecom conference that Google will roll out a new blog search engine.  Could this be the "surprise!" that is currently in the greyed out box on the home page?

If so, then welcome to the club, folks.  Jump in, the water's fine.

May 7, 2003

Hilary Rosen to rewrite Iraq's copyright laws?

It is being reported in the My 6, 2003 Harper's Weekly Review (damn, no permalinks!) and The Register that Hilary Rosen, the CEO of the Recording Industry Association of America, is going to help rewrite Iraq's copyright laws.  Whoa!  I thought we were fighting this war for the freedom and liberation of the Iraqi people.  Hmmm, real, or hoax?  Rosen has announced her resignation from the RIAA by the end of 2003.  The only article I can find on this story is Andrew Orlowski, who isn't exactly known for his journalistic ethics or fact-checking ability.  He references investigative journalist Gregory Palast, who has written for Harper's in the past.  Orlowski also has quotes from Palast from the Democracy Now radio program.  Can anyone verify this?
Posted by dsifry at 8:48 AM | TrackBack | View blog reactions

My new!

I'm sitting here in my new home office, woo hoo!  Given the upcoming release of Sifry 3.0 (any day now), this new office downstairs in my basement under my stairwell (remember Harry Potter's room under the stairs?) is actually a cozy spot.  It also gives me some balance so that I can go to a quiet place to work while also having the baby monitor nearby.  And all the tech infrastructure that was necessary was an electrical outlet, a couple of lamps (14000K Fluorescent for the cool spectrum, Halogen desk lamp for the warm light add up to a nearly full sunlight spectrum) and a UPS.  Connectivity is completely wireless (via a handy-dandy Sputnik AP120 upstairs) and my dd has packed productivity enhancers in every possible nook and cranny.  He carefully measured, cut, and adjusted the 7' long desk to fit around the exposed studs almost perfectly, and then put shelving between the studs, making use of the space above the desk that would otherwise be lost.  There's a lip under the desk where the house foundation meets the staircase, and it has been transformed into a utility shelf, holding my UPS and computer speakers.  I also plan on putting a small computer underneath to further take advantage of the space.  All in all, I'm a happy man with a clean desk.  I wonder how long it'll stay that way.