fettig.net

Jeff Bezos Introduction:

Web 1.0 was about making the internet useful for humans - Web 2. 0 is
about making the web useful for computers.  Amazon has web services - href="http://amazon.com/webservices">http://amazon.com/webservices. 
Also just announced Alexa web services, with data from Alexa’s web
crawl (is site slow/fast? adult content? etc.)

Sites using Amazon APIs such as musicplasma.com,
scoutpal.com (use barcode-enabled
scanner to find out how much a used book costs on amazon.com, so you
can decide if you should buy it at a store).  Amazon also uses its own
web services for internal apps, such as the new href="http://a9.com/">a9.com.

Now Tim O’Reilly is interviewing Jeff Bezos

Tim: Web Services let other people “Rip, Mix, Burn” web content - will
site owners feel like the music industry?
Jeff: “There’s got to be business models for these things”.  Don’t give
away key assets for free, but figure out which assets might be valuable
to others (and possibly charge for use).

Tim: (after a brief discussion of rich people funding space startups):
Are you planning to get up there?
Jeff: Definitely.

Q (from audience): How did you convince pubs to allow
search-inside-the-book?
Jeff: Offering samples at point of sale always increases sales - all
book stores let you flip through the book.

Ross Mayfield from href="http://socialtext.net/">SocialText, on Enterprise Social
Software.  True story:  While sitting in the “Web OS” workshop, I check
the weblogs of a couple people listed on the Web 2.0 Wiki (powered by
SocialText) under “Blogging Web 2.0″.  One of those people is href="http://buzzmachine.com/">Jeff Jarvis, and when I check his
RSS feed the first post I see is href="http://www.buzzmachine.com/archives/2004_10_05.html#008125">this
one, where he says “I’m at Ross Mayfield’s SocialText session…”. 
I think, “Hey wait a second!  I wanted to be in that session!”, and I
leave my workshop and head over there.  (Turns out they had rescheduled
Ross’s workshop, without really announcing it).  Wikis and weblogs to
the rescue, getting me the info I needed in time.

Anyhow, my notes from Ross’s talk:

Social Software - software that involves social rewards systems -
blogs, open source.

Important thing to enterprises is dealing with exceptions to
processes - how to deal with breakdowns.  This is where most innovation
occurs - goal should be to capture what happened in exceptional
situation.

Unlike content management systems, Wiki doesn’t erect barriers (other
than social ones).  Anyone can edit page.  Make it as easy as possible.

How to get people to contribute?  Let them email in, let them read with
RSS - don’t make them change behavior by coming to the page.  (How is
email authenticated?)

“Occupational spam” - mass emails within a company, reply-to-all.  Give
people alternatives for making information available to the entire
company.

PhotoNotes standard for embedding comments in an image (related to
specific parts of the image).  Interesting.

Notes from the Web 2.0 workshop “Dialing the App Tone: How the Early
Web OS is Shaping Up”.

Speakers are Rich Skrenta, Topix.net,
and Stuart Butterfield, Ludicorp.

Rich wrote a paper about how Google was actually building a Web OS, a
brand new computing platform.  Stuart then explained how the “Web OS”
idea comes from the concept of an OS as a set of underlying tools that
developers build their apps on top of.  Today open APIs (such as those
of Stuart’s flickr) let
developers build applications on top of existing web services (although
Google’s platform is for the most part only available to Google’s
internal developers).

Next up in the Lightweight Business Models workshop, a talk by href="http://marc.blogs.it/">Marc Canter.  Marc is kind of a crazy
loud guy, quite a contrast from the mild-mannered Jason :-).  His talk
was about open standards:

• Blog posts - where we are today.  RSS is a standard for
  transferring author, title, timestamp, contents, url.
• XML schemas - new kinds of microcontent.  iTunes has itms:
  namespace that they use to add music-specific information to their
  feeds.
• Marc thinks more people will write reviews than blogs - their
  opinion on their hairdresser, a movie, etc.
• FOAF - XML standard
  for information about people.  Marc is invovled with a project called href="http://www.socialtext.net/foafnet/index.cgi?FOAFnet">FOAFnet,
  for sharing FOAF info.
• Creative commons - standards for releasing copyrighted
  information.  Combined with open APIs, anybody can upload/find/download
  freely available multimedia.  (see OurMedia)
• OpenEvents - sharing
  information about events
• Open standards for reviews - currently reviews are locked in to
  single sites - Epinions, Amazon.

These standards provide an open infrastructure, which companies can
then build business models on top of.

Good talk from Jason Fried of 37signals
on building products with a small team.  Jason is a good speaker.  Main
points:

• “Say no by default” - avoid adding features before you really
  need to.  One good reason less features are better - people will figure
  out creative ways to use the system, rather than having all processes
  dictated to them.
• Design the interface first, and use it as your functional spec
• Do a quick 1.1 release soon after 1.0
• Iterate in the wild - enable beta features for existing customers
  to get feedback
• Scaling - can handle increasing number of users in several ways;
  for example, raise prices to increase profits while keeping number of
  users from growing beyond what system can handle.
• Double duty - use a single interface for admins and regular users
• Hype features other people want to talk about - build features
  that enthusiastic communities will latch onto.

I’m going to try to blog my Web 2.0 Sessions. To start, here’s some
notes form the workshop on RSS and Syndication:

• Lots of discussion around publishers working with RSS, monetizing
  feeds, tracking users with cookies, etc. Publishers are really into
  tracking readers so they can tell their advertisers how many readers
  they have
• Big players talking about routing everything through their services as a
  solution
• Steve Gillmor: attention.xml will let big RSS services work together
• Talk about Ebay, Craigslist and buyers finding sellers. Current
  architectures are centralized - people are thinking about solving the
  same problems with distributed architectures.
• “We’re Still in the Kum-ba-ya period of blogging” - bloggers aren’t
  trying to game the system by lying about what they’re really writing
  about (other than comment spam). Systems for tagging are potentially
  open for abuse.

Ah, so maybe it wasn’t the best idea to post the public email address
for my href='http://fettig.net/sandbox/mt3/'>sandbox
blog… the spammers href='http://fettig.net/sandbox/mt3/archives/2004/10/your_ref_512906'>have
arrived!

Whew, this has been a crazy week. Getting ready to leave, and wrapping up a lot of work. I’m ready for a break.

Hep continues to come along nicely, though. I’m testing some new features on hep.fettig.net, including giving users the ability to selectively share resources with the general public. For example, here’s me sharing my RSS feed. Not much to get excited about, since my feed is already public. But you can also share private resources, like a weblog (via the MetaWeblog API). This page and this email address provide read/write access to this weblog.

I’ve also added support for the CommentAPI, and Atom comment service.post links, meaning that for weblogs that support web-services commenting you can read an RSS feed in your mail client, hit reply, and have the reply get routed back to the originating weblog as a comment.

My plan is to let these features stabilize for a week or two, and then release Hep 0.7.

Now I’m off to the west coast!