On Sun, 24 Jun 2001 17:33:17 +0100 esteemed Gervase Markham did hold forth
thusly:
> I know jrgm can produce stats about how Mozilla's performance is
> improving, but any line drawn on those graphs as "acceptable for 1.0"
> would be, to a greater or lesser extent, arbitrary. We can't compare to
> 4.x - because we do so much more. We can't compare to NS 6.0, because it
> sucked, perf-wise :-) We can't compare to IE, because it doesn't run on
> all the platforms Mozilla does.
I don't understand the reluctance to do performance comparisons. Certainly
Moz 1.0 can be compared to IE and Opera and NS4.x on all the platforms where
at least one of the others runs. That will provide a great deal of useful
info. Certainly as well Moz 1.0 on Win32 can be compared to Moz 1.0 on OS/2,
Linux, and the BSDs on the same x86 hardware. So lots of useful comparisons
can be done to get a good sense of where Moz 1.0 will stand performance-wise.
However, without all those comparisons I can tell you already what the answer
will be: Moz will be perceived as being the slowest browser out there. Its
slower to start. It takes more CPU to load and display the same page. Its
slower to open new windows. And so on.
Frankly, I come away from reading this thread wondering why Moz 1.0 is going
to be released in a few months. The reasons must be political and business in
nature. From a technical standpoint its hardly ready to go up against the
competition and be favorably compared.
If the preparation for doing the Moz 1.0 release is going to slow down the
overall rate of progress (eg no major changes for performance enhancement, no
added APIs to allow more outside developers to join in to make contributions)
then I'm rather unenthused by the drive toward Moz 1.0 in a few months.
Better it come out next year from a purely technical standpoint.
Of course there may be business reasons why a premature release is required.
If so then I can accept that.
