The Open Source Application Foundation [LWN.net]

One might be forgiven for a certain sense of déja vù: a group of longtime industry people, with names like Andy Hertzfeld, gets together with a pile of money to redefine the desktop experience. The story is a little different this time, so a quick look at the Open Source Application Foundation (OSAF) is worthwhile.

The OSAF has actually been operating since the summer of 2001, but it has only recently made its existence known to the rest of the world. The Foundation has been funded by Mitch Kapor, the founder of Lotus and a co-founder of the Electronic Frontier Foundation; its mission is to "Create and gain wide adoption of Open Source application software of uncompromising quality."

The Foundation differs from the venture-funded exercises of the past few years, however. It is a non-profit organization, funded by donations. Thus far, it appears to be working mainly from a big donation from Mr. Kapor; there is a donations page for those who would like to add their support as well. The OSAF thus looks more like the Free Software Foundation than a company like Eazel, but there is no confusing the two. There appears to be no political agenda to the OSAF's activities beyond the production of high-quality free software. The Foundation also foresees ways of revenue generation ("fee-based license for proprietary developers who do not redistribute source code, the fees fund our core development") that the FSF would not approve of.

The first project is ambitious, the creation of an "interpersonal information manager" which will handle email, calendars, contacts, etc. It will be built on top of a number of established free software technologies: Python, wxWindows, the Zope Object Database, Jabber, Mozilla, etc. The calendar component, it is hoped, will be released by the end of the year. The project is seen by many as an alternative to Outlook, though its backers see it as something entirely new. Rather than try to clone Outlook, the OSAF people want to try some different approaches. From a design description posted by Mr. Kapor:

Recent open source groupware products & projects (Evolution, Kroupware) use Outlook as the baseline for design and functionality, an approach which benefits users by being familiar, but doesn't take design risks which could have big pay-offs for users in power and simplicity. We're trying to re-think the PIM in fundamental ways and expect to be judged in terms of our success in achieving that goal.

It is, frankly, a relief to see that the project is trying to do something new, rather than chasing the taillights of proprietary application vendors. As they say, we will have to see what they come up with to see if they succeed, but the goal is right.

The days of high-flying companies using venture capital to take over the world with a great new free software platform are done; the likes of Eazel or Zelerate will not be seen again anytime soon. Much of the excess of the dotcom boom will not be missed, but it would be nice if we could retain some of the focused (and funded!) development that those companies created. With luck, the OSAF will do that, at least for one piece of the application space.

(Log in to post comments)

New? What's that?

Posted Oct 24, 2002 10:22 UTC (Thu) by leandro (subscriber, #1460) [Link]

> the project is trying to do something new

Please don't make new such a misused word to be as meaningless without heavy qualification as innovation, inventor, or God. What they are trying to do isn't new, but a logical development on the very old Lotus Agenda. As good as it might be (as far as it doesn't compete with a full RDBMS), it is not new.

> rather than chasing the taillights of proprietary application vendors.

There's nothing wrong with that, because that's very seldom the whole story. Proprietary vendors seldom can match the efficiency, reliability, portability and security of free software, so even "chasing the taillights" is usually a big improvement in itself. Moreover, there is still proprietary software that needs being emulated. We still don't have equivalents of Alphora Dataphor as a truly, fully RDBMS, or of IBM DB/2 as a scalable, standards-compliant SQL DBMS.

Evolution is more than an "Outlook clone"

Posted Oct 24, 2002 15:40 UTC (Thu) by hazelsct (guest, #3659) [Link]

I disagree with the assertion that Evolution is an "Outlook clone". From Miguel's first posts outlining its design, the goal was to build around an "email database" concept, where messages can be quickly retrieved and very rapidly searched. This enabled the vfolder concept, in which search terms/conditions are stored for later use as a "virtual folder", much like a database query, allowing for cross-referencing far beyond what folder hierarchies can provide. The indexing and searching algorithms used to make this lightning-fast are quite innovative, and the concept of vfolders is brand new (at least in PIM/email application space).

The UI may look very familiar, but there are significant new features behind which lie an engine which is very innovative and new.

(Can't comment on Kroupware, I know almost nothing about it.)

Kroupware has a solid concept for scalability

Posted Oct 26, 2002 15:51 UTC (Sat) by ber (subscriber, #2142) [Link]

Being a commercial endeavour the Kroupware project can only jump as far as the client enables it two. It is true that the interface on the user site needs to be as familiar as possible. However there a more interesting design decisions elsewhere. The Kolab server and its idea of scalability might actually carry further than currently widespread approaches using fat databases. It is an attempt to overtake proprietory solutions.

Taking the long view stance this bring up the question on how innovation in software and especially with Free Software really works. My current theory is that it will not be the big leaps that progress the art for the users. They have the drawbacks that refinancing is mandatory and that they might be a jump in the wrong direction. Smaller steps seem to be more user driven and financially stable. Maybe we need a new way to look at the advancement of software in the future.