>> I didn't have problem with static page.
>
>That would follow. The bug is generally only seen when subrequests are
>involved (eg, mod_include's include virtual and exec cgi are broken by
>this bug), because the file/mmap/etc buckets from the subrequest which
>should get passed back up to the main request go out of scope.
Could you explain me why lynx and netscape didn't have this sort
of problem ? When I look the packet using ethereal they seems
to be very similar. 2 packets (600 + 5)...
>Not yet AFAIK... there are still a few important bugs
>(including this one)
>that people are trying to sort out.
No problem, I'll test mod_jk by using NS 4.77
