fettig.net

I’m going to be  in San Francisco from October 2 – 7, taking a
mini-vacation and then attending the Web
2.0 Conference.  If you’re going to be there too, and would be
interested in meeting up, drop me an email.

I downloaded the href="http://www.mozilla.org/press/mozilla-2004-09-14-02.html">new
releases of Thunderbird and Firefox at work this morning.  First
impressions:

• Wow, Thunderbird has RSS support!  You can read RSS right in your
  mail client, without using Hep or another RSS-to-email converter.  For
  the record, I think this is great.  Clients should try to support as
  many standard formats as possible, and if they’re going to support mail
  and news, they should support RSS.  I expect that RSS/Atom support will
  be a standard feature for mail clients in the future.  If all people
  want is an easy way to read RSS feeds in their mail client, this new
  Thunderbird might mean they don’t need to use Hep, and that’s fine with
  me.  However, I don’t think it will ever be possible for every client
  (or
  server) to support every protocol and format that a user might
  need.  For a lot of people, a simple RSS-to-email bridge is not going
  to be enough, and that’s where Hep will continue to be valuable.
• As far as the RSS support itself, it’s definitely in the early
  stages of implementation.  Thunderbird can’t autodetect feeds, it can’t
  import or use OPML files, and in my few minutes of testing I ran into a
  couple of display bugs.  Also, I don’t like how it defaults to
  downloading the entire linked-to page, rather than just showing the
  text contained in the RSS feed.  But I’m sure all these things will get
  better with time.  For now, it’s just cool to see completely integrated
  RSS support in a mail client.
• src="http://www.fettig.net/weblog/thunderbird-show-images-button.png" align="right" border="1">Thunderbird finally lets you download images on
  a message-by-message basis.  So you can disable loading images
  globally, but still choose to see the images from people you trust. 
  This is the one big annoyance I had left with Thunderbird, so I was
  psyched to see that button!
• Wow, Firefox supports RSS, too! 
  
  The implementation works like this.  When you visit
  a site with an RSS feed, Firefox detects it and displays a little RSS
  orange icon in the status bar.  Then you click on the icon to bookmark
  the RSS feed.  The feed then acts as a dynamic bookmarks folder, with
  each linked-to item in the feed showing up as a bookmark.  Now, this
  doesn’t strike me as a great way to read news and weblogs.  But after
  thinking about it for a minute it struck me – this is a great way to do
  server-based bookmarking!  Keep your bookmarks on a server (like
  del.icio.us), access them from anywhere.  You can even have a bookmarks
  list being maintained by a group.
• Firefox also has a really nice little search toolbar that shows
  up on the bottom of your page when you do a search.  It lets you jump
  forward and back, or highlight your search term on the entire page. 
  Searching in Firefox was nicely dialog-less already, with the
  command-line style hit-slash-and-start-typing feature; now it’s even
  nicer.

Side note: See those images included in this post?  I’m playing with a
new Hep feature.  Send an HTML email with included images to your
weblog, and Hep will pull out the images, upload them using metaWeblog.newMediaObject,
and update the orginal message to point to the uploaded URLs.

Update, 3:06 PM:

More thoughts on RSS support in Firefox and Thunderbird from Simon Willison, Unclespam.

I should mention that there’s now some introductory href="http://fettig.net/wiki/HepProject/HepProject">Hep documentation
available in my Wiki.  My href="http://fettig.net/projects/hep/doc/README.txt">previous attempts
at docs were designed mostly for the administrator, but these are
written for end users.  I wrote them while sitting on my rooftop (via a
wireless link) which made the dull task of documenting my work a little
more fun :-).

I’ve set up a mailing
list for Hep users, using the Google Groups beta.  This will be the
place to ask questions about running Hep, share tips, and discuss
problems and solutions.  I encourage anyone using Hep, especially
people with accounts on hep.fettig.net,
to sign up.  There are also some Atom feeds for people who just want to
track the latest posts without signing up ( href="http://groups-beta.google.com/group/hep-users/about">see the
hep-users about page).

I’m happy to announce the availability of a public Hep server: href="http://hep.fettig.net/">hep.fettig.net.  It’s is running the
latest code from my subversion
repository, serving up HTTP, POP, IMAP, and SMTP on the standard
ports.

The purpose of hep.fettig.net is to give people a way of experiencing
Hep without installing it themselves, and to provide me with a pool of
testers to hammer on the latest code before releases.  Accounts are by
request only (for now), so if you’d like an account, send me an email.

Robert Scoble says that MSDN is turning off full-text RSS feeds because they eat too much bandwidth. I left a comment suggesting they turn on support for conditional HTTP GET, and I wanted to give some real-world numbers for the amount of bandwidth this saves, so I checked my current access log:

abe@fettig:~$ grep "GET /xml/rss2.xml" \
 /var/log/apache/fettig.net-access.log \
 | cut -d ' ' -f 9 | sort | uniq -c
    763 200
   5069 304

That’s really impressive. Out of 5543 requests, 4831 resulted in a 304 ‘unchanged’ response (meaning that the actual file wasn’t downloaded): an 87% bandwidth savings. The majority of feed readers are behaving like they should, checking to make sure the file has changed before they download it.

There’s a lot of interesting thinking being done on managing messages
in different systems.  First, here’s href="http://www.windley.com/2004/09/01.html#a1396">Phil Windley,
(via this
post from Jon Udell):

type="cite">

One problem with moving from a single general purpose tool
like email to multiple special purpose tools is split focus. To
understand what I mean, think about RSS. RSS has reduced the number of
mailing lists I subscribe to and consequently reduced my email traffic.
Perfect application, except that now I have to remember to fire up my
feed reader in addition to my mail client. … What happens when there
are a dozen special purpose tools managing my workflow instead of just
a linear email list?

Phil is seeing the future here: messaging means more than just e-mail. 
Weblogs, RSS, and web services are going to be first-rate citizens in
the world of messaging soon, if they’re not already.  So how do you 
manage your information now that it’s not all native to email?  Running
“a dozen special purpose tools” is cumbersome to say the least.

This is exactly the problem href="http://www.fettig.net/projects/hep/">Hep aims to solve, of
course, so it’s really exciting for me to see other people thinking
about these things.

Jon also quotes from his book ‘ href="http://safari.oreilly.com/?XmlId=1-56592-537-8/ch16-10321">Practical
Internet Groupware‘:

Messaging is at the center of all groupware activities. We
need to be
able to deeply customize our messaging environments. There are two ways
this can happen…

The two ways he gives are to add more power to web browsers, or make
messaging clients more programmable so they can be adapted by users. 
I’d suggest there’s a third way (one that I think Jon would agree with
based on href="http://weblog.infoworld.com/udell/2004/04/19.html#a977">this post):
allow for message routing and transformation at the server level, using
an intellegent proxy to help the user customize their message flow.

Note to self: Introduce Mr. Udell to Hep soon!

Continuing in my effort to try using other people’s services, I attempted to sign up for a TypeKey account today… and their server is down. Come on, people! I’m trying to give your services a chance, but so far I’m thinking that a central point-of-failure isn’t the best architecture.

I was able to get started with del.icio.us, by the way – if you’re reading this via RSS, swing by my site and check out the Links section in the sidebar. I’m pulling those in from my del.icio.us RSS feed via a cron job, xsltproc, and this XSL file.

CNET does something interesting with their URLs: they inject story keywords into the URL, and then their web server ignores the keywords in looking up the story. So these URLs all point to the same story:

• One
• Two
• Three

This is on top of the “news.com.com” thing, which makes me wonder what kind of mad scientist webmaster they have designing their URLs. Anyway,
I assume they’re doing this for Google’s benefit, which got me to thinking – what happens when a URL contains keywords not found in the actual page? So I’m conducting an experiment. Here’s a link to news.com.com, containing a set of keywords that currently returns 0 results from Google:

Blah

I’ll keep an eye on the search results to see what happens.