This has been a crazy week. I spent a lot of the week talking to people on the show floor and in the .org pavilion about a lot of various issues, both for Red Hat and for the Mozilla Foundation. It was nice to put some faces with some names. I live in the world where you talk to people over email and irc, but rarely get to meet them. This was one of those times when you get to meet up.

On Wednesday I got to do a 20 minute song and dance at the Linux in Government Breakfast. I went about as fast as I could, given the time constraints and I hope that I didn't lose anyone. It was a group of around 30 government folks, many from local towns and states but a few from as far away as Virginia. Giving quick talks from the open source community were Fernanda from Brazil, talking about her work with local organizations and government, myself talking about the success of firefox and some folks from the LTSP talking about their project. On the government side, Jim Willis talked about bringing open source software into the Rhode Island state government. He talked about some of the challenges he faced as well as some of the successes that they enjoyed. I hope that it was interesting to everyone involved.

On the sidelines both before and after the meeting I got the chance to talk to people about what we're doing in the Mozilla community right now and over the next few months. At least in that crowd there was a lot of interest in both the mail and calendar problems. Many had deployed Thunderbird and were using it successfully. There was a huge amount of interest in Calendaring as well. On the client side Mozilla is currently working on Sunbird and we're really looking forward to what that project might produce over the next few months. It's worth keeping an eye on.

In a coincidence of amazing timing, Novell has started the Hula Server Project, based on Novell's NetMail product. It's a mail server and calendar server which fills a huge gap in the open source world. There have been mail and calendar servers in the past, but not with the kind of focus that we've got here. (A collection of random pieces of software doesn't count.) And there really hasn't been a working calendar server until now. It doesn't yet have support for CalDAV, but I think it's only a matter of time before it does. And they have the usual nat-and-miguel-supplied energy behind the project, so they will probably end up with developers and a good healthy project. Nat and I sat down and talked for a half an hour or so, with me trying to prod into describing why they were doing what they are doing, what they want to get out of it in the long term and what they will do when and if the Hula server ends up competing with Novell's larger groupware products. I was pretty satisfied with his replies - it sounds like they are doing it for the right reasons and Nat and his cohorts really understand how open source works (that is, you either work together or you die) so I have high hopes for the Hula project.

On that note, I should talk a little bit more about the Netscape Servers that Red Hat recently purchased. I haven't posted in a while and I don't think that anyone else has either. I got a lot of questions from my previous post on the subject. People wondering what our plans are and what the timelines are. I can't speak as to the timelines, but I can say that a plan is coming together.

For the Directory Server we've selected a license for the project and have started working on getting the source together for a release. We have pretty solid buy-in from inside of the company to make this happen from a political standpoint, so most of those bases are covered. From the strictly technical standpoint, we've got some large hurdles. The build system that it uses right now is pretty far from autoconf. That means that it's not as simple as running ./configure and having an ldap server come out the other end. It's also has a history of only being built in a single environment. This means that the locations for a lot of the libraries that the Directory Server expects to build against are hard wired into the build system. Simply put, there's a huge amount of work that needs to happen in order to get this code into a buildable state for people outside of our build environment. I'm not sure if we'll be in a buildable state at the time the source is released, but I'm hoping that we can get there. It's just a long road. (As a side note, this is one of the things that I really liked about the Hula project - you could just pull and build with familiar tools. I think it's one of the reasons why they've gotten so many people interested in such a short amount of time.)

Also of note is the fact that as part of the Netscape purchase, Red Hat also came into possession of the old Netscape Mail Server and Calendar Server code. We've actually been trying to figure out what to do with it. Talking with some of the old developers that built that code, it was apparently amazingly scalable. They were doing a quarter million imap clients on a single machine - but that was four years ago. Think about the machines back then vs. the machines that they have now. Most of the mail code is still in the tarballs that we got, but it's not in a buildable state. It's got many of the same problems that the Directory Server has with its build system, except that it's built against really old versions of the components in the build tree. It will need a good bit of porting before it's usable again. The code, however, is still quite complete in the sense that the really important parts - the mail store, the imap server, the pop server, and everything else that was at the heart of the technology - are still there and would still be considered modern even though they haven't been built in some time. I haven't looked at the calendar server source code so I'm not sure what state it's in - I suspect a similar state.