Slashdot Mirror
Novell Dumps the Hula Project
asv108 writes, "On the Hula general mailing list today, it was announced that Novell is no longer providing full-time developers to Hula. While the project will continue, it appears that Novell is not committed to developing a viable open-source alternative to MS Exchange. The Hula project was announced in February 2005 with much fanfare."
Didn't Zimbra beat them to the punch anyway?
Take your tinfoil hat off.
Hula was a piece of software that deserved to die. It did the things that Exchange does but didn't interoperate with Exchange. And since not many people in the Linux/Unix world are interested in running a mail server like Exchange/Hula that is a jack of all trades, master of none, people didn't really use Hula. So Novell was pouring all the time and money into a project that they thought everyone wanted, but no one really did.
Honestly, have you heard of anyone actually using Hula? I certainly haven't. I'm sure the Novell sales reps made a few sales based on Hula, but really anyone with a clue was running Sendmail, Postfix, or whatever plain Jane email server your prefer, and the few people who really did want an Exchange like solution just ran something like Zimbra that does the same thing but runs Postfix underneath.
If dumping Hula means that Novell will spend money on things that are actually useful to the open source community then this is great news. Because I'm sure no one is really sorry to see Hula go.
What odd timing. I literally swapped out Hula this morning with Really Simple CalDAV Store. The only reason I used Hula was for it's CalDAV support, so that Evolution clients can work on a shared calendar. It worked fine for a while, but it started eating up 99% of the CPU on the server, so I had to dump it for something else. So far RSCDS seems to do the trick, but I haven't tested it extensively yet. You'd think a shared calender server wouldn't be very difficult to implement, but there doesn't seem to be many stable options in the Open Source world. Evolution's CalDAV support does seem to be a bit lacking, however, so that could be the bulk of my problem I imagine.
Thus far I've tried Hula, RSCDS, Cosmo, and Apple's CalendarServer and none of them seem to be the perfect solution. I'd love to see a package that acts as both a CalDAV server, but also gives you the ability to view and edit the calendars via a nice looking web-interface as well. I'm thankful for the projects that are currently being worked on however, and I guess I should stop complaining and start coding...
--It's Pimptastic!--
the open source project to build an exchange alternative Theres like, 5 different projects trying to achieve the Holy Grail of replacing Exchange. And Hula was far from the leader of the pack.
Short answer:
No, believe it or not, Microsoft wasn't in on this.
Longer answer:
I work at Novell, and for about a year, I was on Hula. I loved it. I still run it on my home server, and it still bothers me that I didn't get to finish and polish the bits I was hacking on. An insufficient degree of planning and management led to the magic "1.0" getting pushed farther out and being less clearly defined. Inside Novell culture (and elsewhere, I would think), that's a bad sign. Other projects were in the spotlight, some Ximian modus operandi kept a lot of Hula's exciting stuff secret, and a few months back, the already-thin team was cut back dramatically. At the same time, its release deadline was moved up, and Hula was still without what I'd call a manager. The writing was on the wall well before the Microsoft deal came around.
I made the mistake of getting pretty emotionally attached to Hula, so this has all been pretty rough for me to watch. I worked weekends and wee hours on that code, and I'd do it again. I can't blame anyone for using this news as fuel for the fire and/or shouting "Novell just doesn't get it", and I can't blame anyone for being highly suspicious given the recent Microsoft deal (I'm still not sure how I feel about that, by the way). But I can say, and you can take with as much salt you want: No, this was the result, a long time coming, of numerous mistakes, and of other decisions that truly didn't seem like mistakes at the time. As much as I love to blame Microsoft for stuff, the facts say otherwise in this case.
Its death as a Novell-sponsored project is unfortunate, but Hula's not dead - it's grown a small community and a bunch of us still have commit access. Read the mailing list message, take a breather, and if you still feel like being pissed off at Novell or Microsoft, fine. I tried. But at least check out Hula. It still has a ton of promise and is surprisingly useful today.
Much of what Novell was promising for Hula was ideas that have been either implemented or planned in the Citadel project [http://www.citadel.org] anyway. (We pitched Citadel to them about six months before the Hula announcement ... and they said they weren't interested, and then they announced their project. Draw your own conclusions.)
... very easy to install, and just a joy to use.
Anyway, do try Citadel -- it is a very well-integrated collaboration server with an ajax-style web user interface, built-in data stores, lightweight implementations of all relevant protocols (POP, IMAP, SMTP, etc.)
Tired of FB/Google censorship? Visit UNCENSORED!
>An open source alternative to Exchange is the single most important project the open source community could develop to allow IT managers to migrate away from Microsoft. This comment was so insightful it motivated me to create a Slashdot account just to say so.
you had me at #!
As far as the microsoft angle goes, I don't think it is nearly as open and shut as that. Hula had a variety of problems that were difficult to overcome. Almost all of those problems are centered around the underly platform.
Anyone who thinks that Hula had any kind of momentum at all before this announcement is ignoring the fundamental architectural problems that killed the project months and months ago. Something may emerge from the ashes. Zimbra has proven it can be done, but it will have to be a firefox to this convoluted and bloated Mozilla.
Hula wasn't able to do anything without a huge amount of duct-tape and bailing wire. Even if you did manage by some mircle to get it working, replacing all of the sendmail, httpd and ldap servers, you could not integrate with ActiveDirectory, could not add functionality to the stack, had a web interface that was just plain ugly (dragonfly is much better) and could not easily integrate with most of the anti spam systems out there.
More to the point, they had a core crew that looked at JWZ, declared him God, read his statements, proclaimed them good, and then immediately replayed every single mistake that Mozilla ever made.
Hula needed to be simple, clean and functional. Even know a year later, it is none of those things.
That conspiracy theory, while entertaining, is just totally untrue.
The Hula team decided not to go forward with the project because the project wasn't working. It had been nearly two years since we launched Hula and during that time a lot of other people entered the space (Zimbra, Google Calendar, etc) and implemented many of the innovative things that we had planned to do with Hula. This took some of the wind out of our sails, and we had some execution problems too; I don't know if you've noticed, but the project has essentially gone two years without a release, and if you've ever done any significant software development before, you know that's not a sign of a healthy project.
Now, there is some great work in Hula and we sincerely hope that some of it will be useful to the community. The AJAX-based dragonfly web interface for mail and calendar is gorgeous and open source and could be turned into a nice replacement for SquirrelMail or the other web mail/calendar interfaces. The Hula store and the former NetMail agent code are also both open source and other companies are using them now as well.
The guys who worked on this stuff (Jacob Berkman, Peter Teichman, Dave Camp, Cyrus Dolph, Rodney Price, and others) are extremely bright guys, did fabulous work, and really enjoyed the project -- but unfortunately it's one of those things that didn't work out the way everyone hoped. So it goes.
Novell customers of NetMail and GroupWise and other products can rest assured that they are unaffected and will be supported and carried forward -- I'm sure Novell will have things to say about that, so stay tuned.