httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "William A. Rowe, Jr." <>
Subject RE: PR 6973 / 1.3.15 showstopper
Date Sun, 17 Dec 2000 02:29:04 GMT
> From: Jim Jagielski []
> Sent: Saturday, December 16, 2000 6:14 AM
> I understood this as meaning between the commas, but not
> around the "dashes".
> Tony Finch wrote:
> > 
> > ISTM that ap_getword in util.c is a very poor implementation of the
> > #rule. Didn't this come up with reference to cookies?

All three of us are on the same page here.  (Notice the guru who wrote
the spec+code hasn't chimed in :-)

Here is the patch that ends all discussion, perhaps?  The only issue
was the number of modules that would have to be patched (two, ours, but
also 3rd parties perhaps) to be convinced not to return the body if
ap_set_byterange determines a 416 is appropriate.  Since we didn't
want to go to that much pain, ap_set_byterange sets r->header_only.

Also decided to handle a single range spec as a special case of multiple
range specs, since the outcome of only one of multiple range specs being
in-range was effectively the single case.

Appropriate decisions?  Other comments?  This is one that -really- needs
three +1's (including mine, which I hold off for further testing) before
we commit ... that way, this time around, the blame is shared :-)


View raw message