4 ms·

Rob Helmer on twitter reminded me to make the separate point that Netscape was a hindrance over time, not a help. At first, it endowed hundreds employees with

Rob Helmer on twitter reminded me to make the separate point that Netscape was a hindrance over time, not a help. At first, it endowed hundreds employees with CVS commit access, while volunteers -- some of whom were much better hackers -- had to submit patches and beg for reviews to get access. As time passed and we rewrote the old codebase to do XUL-based apps (a multiyear mission, by definition), Netscape management suffered via a series of misadventures culminating in Netscape 6 in Y2K, a terrible release decision, made against mozilla.org and grass-roots engineer advice. After that, AOL replaced the VP in charge and repeated the process in 2001, with another VP decapitation that fall. However, we at mozilla.org who were inside Netscape/AOL by then had the evidence from Netscape 6 to show new management that Mozilla was the better way to develop and test code, and (as important) to recruit talent. We leveled the CVS access playing field after Netscape 6, forcing Netscape to put new-hires on the same footing as volunteers and people who came fresh from Red Hat or other companies to hack the code. I've given some of the history here in other podcasts, as well as in this one. Bottom line is Netscape, while it paid lots of the contributors (all, on day one), was in short order a problem. Money does not equate to success; it has downsides. Too much money, with a crappy agenda and bad hiring legacy (and the code those people wrote, and the design costs they sunk), can and does hurt. Money from Netscape and then AOL indeed helped pay the top engineers with whom I worked to rebuild the code starting in October 1998, and I'm grateful for that. Money from Netscape/AOL did not help Mozilla gain market share or build product -- we did that in spite of Netscape, it was a total hindrance on us doing products intended for end users. We created and distributed our own binaries at first for QA builds to testers who did not or could not compile, as I said in this 'cast. Netscape was offended by our success, and this (in part) drove Dave Hyatt to quit and go work for Apple in 2001. I could go on but I'll save it for "the book".

davidgerard 10y ago
I was just on the volunteer documentation side for 1.0 (working under Gerv), but I do recall how very annoying the corporation could be. Was the basic fact of paying devs better than not, though?
- BrendanEich 10y ago
  Paying devs was critical up front. For infrastructure, paying the IT costs mattered until after 2003 July when AOL laid everyone off except a few of us, who spun out as Mozilla Foundation by September. It's really a trade-off curve, between the benefits vs. costs of everything, from the top paid talent such as Dave Hyatt (and others), vs. the middle tier, vs. the counterproductive cohort. Netscape/AOL hiring improved by hiring proven performers from the Mozilla community, to add to the complexity. My view now (I'm running a startup) is that leaner is better. But it's hard to account for all the costs and benefits. Too many hindsight what-ifs and post-hoc conclusions that could not have been assumed or even hedged at the time. I doubt something like Mozilla Firefox from the ashes of Netscape will be repeated in this way. Mozilla remains _sui generis_. For all the drag of Netscape and the consequent costs to the Mozilla project 1998-2003, we -- meaning Firefox the browser -- made it. I won't linger on the problems, but I will observe that "indie" open source funding seems to yield better core-team governance than Mozilla's centralizing tendencies have produced by now. These centralizing tendencies go back to pre-git/hg CVS days and so to the inevitable tool/culture co-evolutionary process. I'm in large part to blame for not addressing them while I was involved, so I am not pointing fingers, or else I'm blaming myself first. But I must say that in certain crucial respects, Mozilla is the new Netscape.

davidgerard 10y ago

I was just on the volunteer documentation side for 1.0 (working under Gerv), but I do recall how very annoying the corporation could be. Was the basic fact of paying devs better than not, though?

BrendanEich 10y ago
Paying devs was critical up front. For infrastructure, paying the IT costs mattered until after 2003 July when AOL laid everyone off except a few of us, who spun out as Mozilla Foundation by September. It's really a trade-off curve, between the benefits vs. costs of everything, from the top paid talent such as Dave Hyatt (and others), vs. the middle tier, vs. the counterproductive cohort. Netscape/AOL hiring improved by hiring proven performers from the Mozilla community, to add to the complexity. My view now (I'm running a startup) is that leaner is better. But it's hard to account for all the costs and benefits. Too many hindsight what-ifs and post-hoc conclusions that could not have been assumed or even hedged at the time. I doubt something like Mozilla Firefox from the ashes of Netscape will be repeated in this way. Mozilla remains _sui generis_. For all the drag of Netscape and the consequent costs to the Mozilla project 1998-2003, we -- meaning Firefox the browser -- made it. I won't linger on the problems, but I will observe that "indie" open source funding seems to yield better core-team governance than Mozilla's centralizing tendencies have produced by now. These centralizing tendencies go back to pre-git/hg CVS days and so to the inevitable tool/culture co-evolutionary process. I'm in large part to blame for not addressing them while I was involved, so I am not pointing fingers, or else I'm blaming myself first. But I must say that in certain crucial respects, Mozilla is the new Netscape.