Milestone: Phoenix 0.1 released, first version of Firefox

The start of Firefox (originally called mozilla/browser, then Phoenix, then Firebird, then Firefox) happened because various key hackers in Mozilla (some of whom were Netscape employees) were fed up of the control Netscape exercised over the Mozilla suite – particularly in terms of the UI, but also the design-by-committee kitchen-sink feature list. The original group as I remember it were: Dave Hyatt, Blake Ross, Pierre Chanial, Ian Hickson and Asa Dotzler. Checkin rights to their part of the tree were not available to anyone else, although patches would be considered. Hixie wrote an extremely unapologetic FAQ on this point (which I’d like to find a copy of, but can’t at the moment), which led to me referring (perhaps unfairly) to their development process in an email to Brendan as “arrogant cathedral-style”.

Our habit of choosing already-taken names for our products began around this time. Phoenix was the name of an open source Java Server framework and of a mail client . (Chimera, which became Camino, was actually the name of an existing _browser_. which lived at http://www.chimera.org at the time, and might have been the same as this one: . It has been suggested that Dave Hyatt’s naming research was a little… perfunctory.) Amusingly enough, I have an email from shaver which says “It’s way, way [too] late to be changing the name of our Phoenix”. This was before we changed it – twice. 🙂 The final nail in the coffin was an approach from the Phoenix BIOS people because of their Phoenix FirstView Connect product . By December it was clear that a rename was going to be necessary.

Here’s a document that Mozilla staff wrote for a CNet reporter (although she never saw it before writing her article) about Phoenix, which we then agreed to publish somewhere. It’s interesting because of the insight it gives into the direction Mozilla was moving in at the time.

Mozilla.org is hosting a small, experimental project exploring development of a cross-platform, XUL and Gecko-based stand-alone browser application. The project is known as Phoenix. Phoenix is roughly analogous to the Chimera project, which mozilla.org is also hosting. Chimera is a new browser for the Macintosh platform built using Gecko and a native toolkit for the Macintosh to give Mac users the native look and feel they so desire. Phoenix is our stand-alone browser experiment for Linux and Windows. We don’t yet know how much effort will go into Phoenix or whether it will produce interesting results. We’re interested in this project because:

1. Phoenix exercises the Mozilla application framework in an illuminating way. We now have an application toolkit which has reached a 1.0 status, and which was created with browser-related projects in mind. What better way to test it out than to iterate once again a build a focused browser application. Our current application suite showcases what can be done to promote integrated applications. A project focusing on using Mozilla technology to create a single, stand-alone browser application may teach us new things. Perhaps we’ll find shortcomings in our XUL 1.0 capabilities. Or perhaps we’ll find that it’s an even better toolkit than we expected.

2. Phoenix explores the idea of decoupling the various applications which create our current application suite. We’ve received requests for a stand-alone browser for quite some time. Now that Mozilla 1.0 has been released, we can accommodate this type of experimentation.

3. Phoenix aims to provide a “layered” approach to building a web browser. In other words, allowing mozilla.org to ship a simple stable base with core functionality, and provide a means for managing extensions and layering in add-ons, so that a user could build up the browser to be as complex as he or she wants. This allows some users to have the range of features found in today’s Mozilla releases (or even more) while also providing a convenient path for those who want a lean, quick, simple browser.

4. It has been proposed by a group of XUL experts who have been leaders in the development of Mozilla’s browser application, and whose creativity we want to encourage.

To do this, we’ve created a separate browser partition in our CVS tree. This will allow the cohort of hackers who proposed this project some room to experiment without affecting either the seamonkey branch or trunk. This is a restricted partition, meaning that it is open only to its designated owners and peers. In other words, CVS write access to the seamonkey tree does not include checkin access to this partition.

Development of the browser application and suite in the CVS tree will not be affected at all. Review, super-review, check-in access, involvement of drivers and other mozilla.org policies will continue without change.

In addition to Phoenix and Camino, there were a number of other efforts going on around this time to experiment with different approaches to creating browsers based on Mozilla code. Other browser efforts include Galeon, K-Meleon, SkipStone, Q.Bati, Beonex Communicator and Aphrodite.

I think there was a sense that if ‘one hundred browsers bloomed’ then the best of those experiments could help shape the direction of development and some of these other browser efforts did help inform decisions made with Phoenix.

There’s an article with more information about several of these different browsers at

http://www.oreillynet.com/mozilla/2002/09/12/mozilla_browsers.html