1. 19
  1. 6

    I used Mozilla as my daily browser since M17. It was a lot of fun watching it come together. There were plenty of bugs; I remember things like one build where all forms were invisible, and another that corrupted my mailbox. The Netscape 6.0 release was just horrifying, because I remember M18 implementing a lot of things and including a lot of bugs. Netscape 7.0 was based on Mozilla 1.0 which was, for the most part, “ready.”

    But, the great thing about that time (which I now miss) is there was a “Netscape” product for consumers and a “Mozilla” thing that was really aimed at nerds who could submit detailed bug reports. Mozilla itself was free from commercial considerations. A few great things were included in Mozilla that never made it to Netscape; one killer feature I remember was per-domain image blocking, which was great for blacklisting obtrusive advertisements. IIRC you could have it prompt on each domain or just right click an image to block all images from that domain. Per domain cookie blocking I think lasted much longer, but both are now removed from Firefox. Seamonkey still supports them…

    1. 4

      Shh, keep it quiet about seamonkey, lest the software ruiner gnomes hear of it

    2. 1

      Netscape 2.0 already quite usable, 3.0 a beast with java and js. I find it a worthwhile exercise to include both in my test suite. BTW, Mosaic 3.0 has a tree history which none of the mainstream has reproduced since.

      The Web has so much amazing history and it,s accessible to anyone capable of using Wine or VirtualBox.

      1. 2

        Didn’t Netscape 2.0 introduce JS? I seem to remember it being a bit of a mess because it was unusably buggy, but the browser thought it could do it and was always willing to try.

        For me there was a killer feature buried in there: tables whose cells could contain arbitrary background color. I thought that arrived in NS 3.0, and it was the last feature needed to support arbitrary layouts. Everything after that has been people trying to invent new ways to achieve an existing result.

        1. 1

          You’re right, I just was reminded of this testing with it. It is indeed very basic.

          Agree on table layout, it,s still the easiest, best supported, and most predictable, IMO