Archive for 2004

I’m happy to announce that I’ve taken a new job: I’m working for href="http://www.jot.com/">JotSpot as a software engineer.

align="middle" border="0" height="71" hspace="5" vspace="5" width="291">

Like a lot of other people, I was first introduced to JotSpot during href="http://www.fettig.net/weblog/2004/10/web_20_jot">Joe Kraus’s
demo at Web 2.0. I posted a blog entry about it, and (through the
amazing connecting power of the Internet) within an hour had received
an email from Graham Spencer, href="http://www.jot.com/about/index.php">JotSpot’s CTO. That
lead to conversations with Joe and Graham, and a couple of weeks after
Web 2.0 they offered me a contracting gig. I took the contract, and href="http://www.fettig.net/weblog/2004/11/ongoings">quit my day job.

And, let me say, I was perfectly happy as a contractor. After almost
three years of working a 9-to-5 cubicle job, the freedom felt great,
and I wasn’t feeling any rush to find another salaried position. So
when Joe told me they were interested in hiring me, I had mixed
feelings. I told him that I’d be willing to consider it, though, and he
invited me to come back out to California and spend a few days at their
office.

Which I did. And by the time the week was over, it was clear: I really
wanted to work for JotSpot. Graham and Joe are talented, creative,
enthusiastic, and friendly. They’ve developed a powerful platform which
solves real problems that I’m personally familar with. They have some
very cool ideas in development. And they have a great team of people -
I can honestly say I liked everyone in the Jot office.

In short: I like the product, I like the company, I like the culture.
I’m flattered to have been offered the job, and I’m tremendously
excited to be part of JotSpot.

I spent the day at home today, not feeling well. Being stuck at home I
had a chance to catch up on some things, including cleaning out my
email Inbox, which unearthed a few messages that I need to reply to but
had forgotten about. So if you’ve sent me a message in the past month
and haven’t heard back, you should soon. I’ve got 19 pending items in
by Inbox at the moment, and I hope to get that down to < 10 before I
go to bed.

Also today I wrote a little program, href="http://fettig.net/svn/yarn/trunk/examples/yarncopy.py">yarncopy.py,
which uses Yarn to copy messages from one location to another. It’s
just a little script, 118 lines of code, but since it’s based on Yarn
it can work with lots of different formats and protocols.  You can do href="http://catb.org/%7Eesr/fetchmail/">fetchmail-type things:

yarncopy.py pop://abe@my.server mailto:abe@localhost

Or rss2email-type
things:

yarncopy.py http://fettig.net/rss/rss2.xml \
 mailto:somebody@example.com

And you can even do weird things like IMAP-to-blog:

yarncopy.py imap://mail.example.com/Inbox \
 metaweblog:http://example.com/mt.cgi#Weblog

It will prompt you for user names and passwords as necessary.  Right
now it copies every message, every time, which limits its usefulness,
but with some more work I think it has the potential to become quite a
powerful tool.

I’m moving my mail and web hosting to TextDrive. Hopefully the
transition will be complete by the end of the weekend.

Taking a moment to blog some random things from the past couple of
weeks…

1. Yesterday I got this message from the new iTunes 4.7:

src="http://www.fettig.net/weblog/iTunes Message.png" height="194" width="490">

“This will prevent… AirTunes remote control”? Does anybody know what
that means? A Google search on href="http://www.google.com/search?q=%22airtunes+remote+control%22">“AirTunes
remote control” returns a few instances of people wishing for an
AirTunes remote control, but no mention of such a product actually
existing. Is this a hint of things to come?

2. Spammers have been spoofing my email address as the source of
their messages. Not only that, but they went the extra mile and faked a
“Received” header so it looks like the message came through my
provider’s SMTP server. I’m getting a lot of messages bounced back from
non-existent addresses that they’re trying to spam. Two hundred fifty
eight bouncebacks messages in the past three days, to be exact. I’ve
set up a filter to deal with them, but it’s still extremely annoying.

3. I got a new cell phone, a href="http://commerce.motorola.com/cgi-bin/ncommerce3/ProductDisplay?prrfnbr=262351&prmenbr=126&phone_cgrfnbr=1&zipcode=">Motorola
v505. I like it. It has a camera, which is silly but fun. I’ve been
sending pictures to Hannah, pictures to Flickr, pictures to my href="http://fettig.net/sandbox/mt3/">test blog. There seem to be
a lot of questions, but no answers, regarding its compatibility with href="http://www.apple.com/isync">iSync. If anybody has some
definitive information I’d love to hear it.

4. href="http://homepage.mac.com/amake/shared/docs/software/firefoxy.html">FireFoxy
is a slick little OS X app for making your FireFox widgets look
better.  You wouldn’t think a little thing like that would make much of
a difference, but it does. Every form entry becomes a little more fun.

5. I quit my job.  As of Friday, November 5th, I’m working for
myself again!

Inspired by href="http://www.fettig.net/weblog/2004/10/web_20_rael_dornfest">Rael’s
suggestion at Web 2.0, I added auto-linking to Hep yesterday.
Here’s how it works. Say you have a Wiki that you often link to in
your blog posts.  You add the Wiki to your list of Hep resources (via
an RSS feed, or Atom). Now Hep sees all your Wiki pages as messages.
Next, you tell Hep that you want to use the Wiki as a source of auto
linking.  From then on, when posting a message via Hep, all your
WikiWords will be linked to automatically.

But it’s even more useful than that.  Since Hep treats all message
sources equally, you can enable auto links from other sources besides
Wikis.  For example, I’ve enabled auto linking for my del.icio.us
bookmarks feed.  And here are some of the things I’ve bookmarked: Hep
Message Server.  Garrett Wilkin.  TextDrive. The Comment API.
See those links?   Hep created them automatically for me. For
non-WikiWords, or phrases with spaces, you just [enclose the phrase in
brackets].

This feature is currently live on hep.fettig.net, and it will be in the
next release.  Hep 0.7 is going to rock!

I’m back in Portland (that’s Portland, Maine – a fact which elicited
looks of confusion, dismay, and/or sympathy from many Web 2.0
attendees), recuperating from my week in San Francisco.   The pace of
Web 2.0 was so frenetic that, although I did a lot of blogging, I
didn’t have a much of a chance to mentally process everything that went
on until today.  So here are my post-processing thoughts.

The highlight of the conference was the people.  I suppose that’s true
of any conference, but it took me until the the second day to realize
that some of the most interesting things were happening outside the
sessions.   During the breaks, when there wasn’t anything happening on
stage, the volume of people and noise in the hallways made conversation
somewhat of a challenge (and made me afraid that I was losing my voice
from shouting).  So some of the best conversations I had ended up
coming while sessions where in progress.  I had the pleasure of meeting
a bunch of people whose work I’ve admired for some time, and a few
others who I really should have been familiar with, but (somewhat
embarrassingly) wasn’t.  I started to list everyone here, but it seemed
too much like name dropping, so I’ll just mention a few people by
name.  First, Jason Fried, who was
nice to me and gave me helpful advice.  Thanks, Jason.  Also href="http://kottke.org/">Jason Kottke (who in real life is still
dryly funny), Rohit Khare
(who seems to be up to speed, and ahead of the curve, on just about
everything tech-related), href="http://ifindkarma.typepad.com/relax/">Adam Rifkin, (who made
my day by doing the “I’m not worthy” bow),  href="http://ross.typepad.com/"> Ross Mayfield, href="http://aa.typepad.com/mobile/">Andrew Anker (who explained
the New York-San Francisco connection, or “why west coasters don’t like
New England”), Joe Kraus, and
Mark Wong-Van Haren.  It was great to meet all of you.

It was interesting to see a real tech conference in action.  Everybody
had laptops, WiFi was provided, and so there was a constant flow of
blogging, emails, and IM going on in parallel with the face-to-face
sessions and conversations at the conference.  The 15″ Powerbook was
clearly the laptop of choice, with 12″ Powerbooks and iBooks fairly
popular as well.  It was pretty amazing to see all the familiar names
and faces in my iChat Rendezvous window.

There are a few themes that stuck with me.  These are the things that
people were talking about the most:

• RSS.  Publishers want to know how to make money off it, which
  means not just advertising but being able to track readership.  Lots of
  people and companies, including the new Rojo.com,
  are thinking about the problem of being able to wade through hundreds
  or thousands of feeds and find the most important stories.  I’m not yet
  sure how I feel about this.  My instincts say that humans are
  incredibly good at filtering out unimportant information, and that many
  users will only want to subscribe to a few feeds.  So for a lot of
  people it might be perfectly reasonable to just skim through every new
  message, and ignore the ones they aren’t interested in.  I could be
  wrong on this, and time will tell.  In any case, I came away with the
  feeling that nobody is really sure how the average net user is going to
  use RSS.  Lots of startups are making bets, though.
• Wikis.  Web 2.0 had a wiki, Ross Mayfield from SocialText did a
  workshop on using Wikis in the enterprise, Joe Kraus introduced his new
  company, which is selling a WIki.  It’s interesting to see the life
  cycle of a technology like Wikis.  They’ve been around for href="http://c2.com/cgi/wiki?WikiHistory">almost ten years with a
  limited but enthusiastic following, gained credibility with big
  projects like wikipedia, became a standard tool for software projects,
  and now all of a sudden there are two start-ups selling commercial
  software, and betting that wikis are going to take off in the
  enterprise.
• Web APIs.  The theme of the conference was “the Web as Platform”,
  and a lot of the sessions touched on APIs.  All the big sites – Amazon,
  Google, Ebay, PayPal, etc. – are opening up some functionality through
  web services.  These companies seem to have put some thought into
  providing web services that enable developers to use their platform
  without giving away their sources of revenue (although it will be
  interestingn to see their reaction when people start to use the APIs in
  unpleasantly unexpected ways).  Amazon’s services in particular look
  cool – I’ll have to find come time to play with them.

All said and done, it was a great experience.  I’ll be looking forward
to next year’s edition.

I had a really interesting hallway conversation with href="http://www.raelity.org/">Rael Dornfest ( href="http://www.oreillynet.com/pub/au/35">CTO of O’Reilly, author
of blosxom), here at Web 2.0. 
Rael, as it turns out, actually started learning Twisted just so he
could write a Hep-like app, since Twisted has pretty much the only
easily-extendable open source IMAP implementation out there.  So he was
glad to see that I was already working on it, and that it was open
source.

Rael has some really cool ideas around having the message router do
smart things with messages as they pass through.  For example, he wrote
an SMTP server (or maybe proxy server) that would look for specially
marked-up paragraphs in his outgoing emails and post them to his weblog
(allowing him to cc his blog without including the personal parts of
the message).  He also mentioned the possibility of having the server
be aware of Wiki works, so it could automatically insert relevant links
into messages as they passed through.  Brilliant.

Lessig is cool.  His powerpoint slides are white-on-black, typewriter
font, one or two centered words a slide, and they look awesome,
emphasizing and complementing what he’s saying rather than just
mirroring it.

His talk is about remixing, he gives some cool and funny demos of music
and movies based on remixing other sources.  Today (since 1978)
everything is copyrighted by default, and companies use copyright to
prevent people from saying things they don’t want them to.  The need
for permission reduces adoption – if you needed permission before you
could take a picture of something, there would be far less photography.

Presentation from Brendan Eich, lead architect of Mozilla.  He demos
preliminary support for the <canvas> element, based on Safari’s
apple:convas.  E4X is a simple API for xml processing.  Brendan’s last
slide is on the future of open source platforms – mono, cairo (which
they’re using for their <canvas> support, href="http://irrlicht.sf.net/">Irrlicht (a 3D engine), the href="http://www.bbc.co.uk/rd/projects/dirac/">Dirac video codec.

Panel discussion with href="http://web2con.com/cs/web2con/view/e_spkr/416">Tim O’Reilly, Andrew Anker, Brian
Behlendorf, Bob Morgan, Allan
Vermeulen.

Introductory comments: Brian talks about how Apache’s open source model
encourages participation by exposing lots of APIs.  Andrew says that
RSS was the giant leap that made weblogs better than other forms of
self-publishing – suddenly it was possible to track lots of
infrequently-updated websites.  Bob Morgan says that Kodak is trying to
open more opportunities for photo sharing. 

Andrew says that a lot of the value of MT comes from the plugins.  Tim
asks, “Whats the difference between open source collaboration and what
SixApart is doing?”  Brian: Sometimes users contribute to a collective
work owned by a complany – Moveable Type, Amazon reviews.

Tim: Napster made an architectural decision, and then users contributed
all the data.  How do you design for participation.  Bob from Kodak
talks about how they encourage sharing, but I don’t see anything
revolutionary in what he’s saying – specifying sharing with a specific
group vs. sharing with the whole world, etc.  Seems like flickr is much
more tuned in to the social interface thing.

Brian tells the story of how Blue Note deleted all the anti-Norah Jones
posts by the hard-core jazz fans on their website, which changed them
all from being anti-Norah Jones to anti-Blue Note.  Tim asks if there
are other examples of things that can destroy a community.  Andrew:
today people care about links, some people try to game the system to
boost their rank.  Allen Vermeulen says that Amazon is pretty liberal
with what they allow in their reviews, and they haven’t had a big
problem with spam.  The only discussion they’ve had to give up on and
allow to collapse is the Swift Boat Veterans book.  He says the goal
should be to think long term – focus on the benefits that come from
giving people more information, rather than the short-term bad things
that might happen.

Tim: How hard is it for people to adapt to a community-centric, shared
mindset?  Brian: there are tendencies to shield people from blame.  But
these days software has names attached to it, people rather than
companies, “which I think is a good thing”.

From the audience: How can we take the architecture of participation
into the enterprise?  Andrew says that companies like SixApart and
SocialText are bringing social software to the enterprise, and that
these tools are successful for knowledge management because they’re
lightweight and simple enough to apply to many situations.  Bob from
Kodak says that many real estate people and other professionals use
ofoto.