Vote summary

[David Robinson <dr...@ast.cam.ac.uk>]

Vote summary so far, pending any further vetos. (I've dropped 0.8.14.patch
from the filenames to save space.)

03_XBITHACK_restrict withdrawn.

                randy+rst robh  bb    ben  aram   ay    Jim        Total
01_http_config      +1    +1    +1    +1    +1    +1    +1       +8/accept
02_DBMGroupCoreFix  +1    +1    +1     0    +1    +1    +1       +7/accept
04a_ExtraPath       +1(1)  0(2) +1    +1    +1    +1    +1       +8/accept
05_NoKill           +1    +1    +1    +1    +1    +1    +1       +8/accept 
06_SCODocFix        +1    +1    +1    +1    +1    +1    +1       +8/accept
07_SocketMemLead    +1    +1    +1    +1    +1    +1    +1       +8/accept
08_license          +1    +1    +1    +1    +1    +1    +1       +8/accept
09_HP_comment       +1    +1    +1    +1    +1    +1    +1       +8/accept
10_mutual-failure   +1    +1    +1    +1    +1    +1    +1       +8/accept
11_fd_removal       +1    +1    +1    +1    +1    +1    +1       +8/accept
12_del_max_security +1    +1    +1    +1    +1    +1    +1       +8/accept
13_error_fd         +1    +1    +1     0     0    +1    +1       +5/accept
14_rlimit           +1    +1    +1    +1    +1    +1    +1       +8/accept
15_urlchars         +1    +1    +1    +1    +1    +1    +1       +8/accept
16_alias            +1    +1    +1    +1    +1    +1    +1       +8/accept
17_const            -1    -1          -1    +1    -1    +1       +2-5/reject
18_geteuid.apache   +1    +1    +1    +1    +1    +1    +1       +8/accept
19_redir            +1    +1    +1    +1    +1    +1    +1       +8/accept
20_score            +1    +1    +1    +1    +1    +1    +1       +8/accept
21.escape           +1    +1    +1    +1    +1    +1    +1       +8/accept
22.spawn            +1    +1    +1    +1    +1    +1    +1       +8/accept
23.mmap             -1    -1          -1    -1    -1    -1       -7/reject
24_imap             +1    +1    +1    +1    +1    +1    +1       +8/accept
25_startserver      -1(3)  0(4)       -1(5) -1(6) -1(7) -1       -5/reject

I suggest that the 17_const, 23.mmap and (maybe) 25_startserver patches
be carried forward for the next version. It might help if we start
numbering new patches from 26.

 David.

Notes:
04a_ExtraPath:
(1) (Randy,rst,ben): after Ben's correction
(2) (robh): I'm prepared to be convinced it's a good idea, but at the moment I
    don't think it is. It might break well established URLs, so isn't worth
    the risk. I'll change my -1 to a 0 for the sake of harmony.

25_startserver:
    [Andrew's comments]
    Don't get me wrong.  Is there gonna be any fuss if I -1 this because it's
    a functional enhancement which changes the semantics of the configuration
    file.  Wouldn't sensible documentation (ie examples) obviate the need to
    play with the code, in an area that is sure to be changed again post 1.0.

(3) (rst): See Andrew's comments.
(4) (robh): Looks like a feature that has zero effect. I couldn't get 8.14 to
    behave any differently to a 1.0 candidate with p25 added.

    As much as I'd like improvements made to this part of Apache, this
    doesn't seem to do anything or fix anything... I'm willing to
    reconsider if it can be shown that this patch does something useful,
    I've no idea if it's meant to be a feature or a bug fix... the 
    former probably. "StartServers 0" worked fine before I think.

    MinSpareServer 0 is a different matter... no warnings, host connects
    but no response... "what did you expect?" - no "spare servers", just
    fork servers when necessary. Another day..

(5) (ben): If 25 is going to be fixed, I think it should be done like this:

    if(daemons_to_start < daemons_min_free)
        daemons_to_start=daemons_min_free;
    if(daemons_to_start > daemons_max_free)
        daemons_to_start=daemons_max_free;

    Or by eliminating daemons_to_start altogether.

(6) (aram): It looks like it fixes a bug, but horribly.  Why did we do this
    again? I thought this was our name to fame... that we had available servers
    as soon as you needed them.

(7) (andrew): changes conf file semantics, albeit to make things more intuitive