fettig.net

I’m going to be on vacation (and travelling) from tomorrow through the 28th.  During that time I’ll be checking my e-mail infrequently, if at all, and I won’t be writing code or posting to my weblog.  So things are going to be quiet around here. Have a good couple of weeks!

Some people have reported a problem with Hep.  The problem goes like this:  Hep creates an e-mail in their Inbox that has a bad header.  Then when it goes to read the e-mail, it gets an error.  This makes all your messages unavailable until you either (1) find the bad  message and delete it, which is pretty much impossible since Hep doesn’t tell you which message it is, or (2) delete all your messages, and hope there wasn’t anything in there you really wanted to read. 

This latter "solution" has been used by a couple of people recently.  This makes me feel uncomfortable.  So I’ve released a patch that fixes the problem.  Actually, it fixes the problem twice: by not allowing multi-line titles on messages (which is why the e-mails were getting corrupted) and my handling unparseable messages gracefully.

To apply, download __init__.py and maildir.py from http://www.fettig.net/files/hep/patches/0.3.3/.  __init__.py goes in your Hep installation under /messaging, maildir.py goes in /messaging/sources.  Restart Hep, and you should be good to go.

TODO: put up a real 0.3.3 tarball.

Wari sent me an e-mail this morning explaining trackbacks and pingbacks, the key distinction being that trackbacks are client-server and pingbacks are peer-to-peer between servers.  So Hep could support trackbacks, but not pingbacks.

Wari even outlined how trackbacks could work in Hep:

Trackbacks are very possible, all a user needs to do is to enter a trackback url somewhere somehow in a message (mail header works nice for mutt) and hep will try to ping the entry when an outgoing mail

is sent.  There’ll be variables you need, so after sending the blog entry through XMLRPC, you need to keep the blog ID, request it again for it’s full URL, then you can really send the trackback, but you must make sure that when sending, the main message must be stripped off HTML and be less than 255 characters.

This is, as we say in New England, wicked cool.  If (and this might be a big if) people are using the Trackback RSS module, the flow of information could work like this:

1. Hep fetches an RSS feed.  It parses the feed, and converts each entry into an e-mail message.  It also saves the messages, and some metadata (including the messages’ trackback URLs) in its cache.

2. Joe User gets the messages in his e-mail client.

3. One of the messages is particularly interesting to Joe, and he wants to blog about it.  So he hits "Reply".  Because this message came from Hep, the reply-to address is already set to "MyWeblog@hep", which is a special address that tells Hep to send the message to Joe’s weblog.  Joe’s e-mail client automatically inserts an "In-Reply-To" header, containing the ID of the message he’s replying to, into the outgoing message.

4. Joe sends the message.  Hep sees that it’s an "@hep" address, so it knows that instead of delivering it as a regular e-mail it should be posted to Joe’s weblog.  It also notices that the In-Reply-To header references a message that it has in its cache, so it looks up the trackback URL for that message.  Hep posts the message to Joe’s blog, and retrieves the URL for the newly-posted message.  Then it sends a trackback to the website where the original message was posted, notifying it that Joe has replied to the post at such-and-such a URL.

The result of all this is that Joe is able to have a publicly linked, blog-to-blog conversation, all from his e-mail client, without doing anything different than he would in having a regular e-mail conversation.

So, is anybody using the RSS Trackback module yet?

Update: Sam Ruby writes:

I am now. Generally I wait until somebody finds a compelling use case for a feature, and you certainly have a humdinger. I kinda suspect that this is the type of thing that Jon Udell would like to see. ;-)

Today I was surprised to get a reply from Ingo Rammer regarding my post the other day about developer thought patterns.  (He saw a link from me in his referrer log).  Anyway, he explained his motiviations for using Outlook and Exchange as his platform for bridging e-mail and weblogs:

First, I’m a consultant, trainer and commercial developer. Whenever I don’t have a gig at a given time, I instead play with the technologies at hand (aka "learn"). This time, I wanted to play with ASP.NET (simply because I haven’t yet done so) and at the same time wanted to get rid of my closed weblog tool.

I took Ingo’s project out of context, and assumed that his motivation for creating it was just "to make a weblog-to-mail-client bridge", and wondered why he has chosen such a heavyweight platform.  But in turns out that the whole point of the project was to work with the platform.  Next time I’ll try to understand the context of a project before I voice my opinion on it.

Ingo, BTW, is a good example of an open minded Microsoft developer.  His response to my post was friendly and reasonable.  And he’s releasing his project under the GPL.

A wish of mine was granted today when I found a plugin for Eclipse that does Perl syntax highlighting.  It works great, and supports a number of other languages too.

Wari reports on CVS PyBlosxom.  Lots of cool stuff - plugins, calendars, pingback.

One of these days I’m going to take the time to learn about Pingbacks and Trackbacks, and the differences between them.  And then I’ll figure out how to make use of them in Hep.  Another thing for my ever-growing TODO list.

Greg Reinacker has written RSS support for Microsoft Outlook.  And Ingo Rammer is working on making Outlook and Exchange work as a blogging tool.

Both of these projects are cool, and will be useful to people who use Exchange and Outlook.  But I guess I don’t really understand the Microsoft Developer mindset behind projects like these.  As a developer, why is it rewarding to build on top of an expensive, proprietary platform, the next version of which probably won’t be backwards compatible and will break your extensions?  To me it makes a lot more sense to develop for the Internet, and open standards.

But I guess it’s just a matter of thinking patterns.   Developers are always looking to use existing technology platforms, to save the time and effort of implementing common functionality (or, in some cases, the time and effort of thinking about good designs for things) themselves.  For many Windows developers, these platforms are very large - .NET, Office, Exchange/Outlook. In the UNIX world, you’re used to picking and choosing - Python (or Perl) for your language, XML-RPC (or SOAP) for client-server communications, XML (or dbm files) for data storage, GTK (or QT) for your user interface.  Lots of smaller pieces talking to one another.

I’ve been thinking about this because of Charles Cook’s post "Renaissance Developer".  An excerpt:

If you only concentrate on what satisfies your immediate career needs, you’ll be living in a box with tinted windows. You’ll not only see everything in the box in a single shade of colour but worse than that you’ll be missing out on a world of other interesting and intellectually entertaining software ideas. So spend your evenings and weekends on something different from how you earn a living. It will be much more fun and you’ll incidentally end up a better developer, a Renaissance developer instead of an MSCE.

Anyway, in case there was any doubt, I’m a UNIX/open source/standards developer.  But maybe I should make an effort to understand the Microsoft Way.  A .NET project…?

Steven Noel has written up a nice tutorial on using Hep with Moveable Type, and tips on using Mozilla mail as a Hep client. Thanks, Steven!

James Strachan wanted to read RSS feeds in his mail client:

To read RSS feeds I’d far prefer to just use a single UI, that I already use much of the day already, an email client.

So what I’d like is an RSS to email gateway such that there is 1 email per post on each blog, coming from a spoofed email address named after the name of the blog.

After "much googling", he found Hep… and he likes it!

James’ writing about Hep has in turn inspired Steven Noel and Sam Ruby to mention Hep in their blogs, and a lot of new visitors today.

It’s cool to see all these people noticing Hep, and trying it out.  And most people who have written about it seem to like it.  Although I tend to think that for every person who writes about Hep in their weblog, or sends me an e-mail saying that they’re using it, there are 5 who downloaded it, glanced at the README, and thought "python… twisted… INI files… I’ll wait ’till version 1.0 comes out and there’s a nice installer."  Which, if you’re not a developer, is not a bad idea :-)