httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Aram Mirzadeh <...@luers.qosina.com>
Subject Re: Lattest list of patches & my votes
Date Sun, 17 Sep 1995 13:52:39 GMT
> 
> > 
> > 
> > 
> > Here is the latest list of patches for monday's vote.  I doubt I'm going
> > to be around a connection tommorrow to do any kind of patch work so this
> > list should hold, unless a mojor bug is found. 
> > 
> > List & my votes:
> > 
> > 02_http_main_cleanup.0.8.13.patch	+1
> > 04c_SCOPort.0.8.13.patch		0
> > 05_AddDescription.0.8.13.patch		-1
> > 05b_AddDescription.0.8.13.patch		+1
> > 06a_bsd_next_cleanup.0.8.13.patch	+1
> > 08_Wall_htpr.0.8.13.patch		+1
> > 09_log_setgrpprivs.0.8.13.patch		+1

I was working othe 01a patch, there seemed to be a problem with it and how
my server was getting very slow.  It tuned out to be the 04c patch.  I've
not been able to figure this out why yet... so I don't think 'm gonna
veto the patch as of yet.  I'll 04c at 0 and add:

01a_log_config_integer_time.0.8.13.patch +1

> > 
> > 
> > 02 .. ehh... not much to say... 
> > 
> > I don't have access to a unix SCO system right now so I can really test
> > out 04c.  It compiled okay, and the code ( although a little messy ) doesn't
> > look like would cause any serious trouble with any other system. 
> 
> Messy?? But seriously, if you vote 0 because you don't have SCO then
> the SCO patches are liable not to be accepted. I assume most people don't
> have SCO on this list, or I wouldn't have had to port it?

I will have access by the end of the week, not that I would be able to 
support it though.  Anyway, I mean there should be ways of doing the
patch a little cleaner, right now it's changing too other variables that
I can't account for, and see my comments above about the slow down. I've
a little script that hits a site and reports how fast the responces were.
It doesn't have a cache to taint the output so... anyway I had a drop of
12% between a clean-0.8.13, and 0.8.13+04c. 


> 
> > 05 & 05b are both doing the same job, 05b is alot cleaner in my book. 
> 
> I'm glad to hear that not all my patches are "messy"   ;-)
> 
> > 
> > the rest are okay here... 
> > 
> > -- 
> > Aram W. Mirzadeh			     http://www.qosina.com/~awm/
> > MIS Manager	To err is human, to really screw up you need a computer.
> > Qosina Corporation			          http://www.qosina.com/
> > Email: awm@qosina.com				            awm@liii.com
> 
> My votes:
> 
> 02_http_main_cleanup.0.8.13.patch	0
> 04c_SCOPort.0.8.13.patch		+1
> 05_AddDescription.0.8.13.patch		-1
> 05b_AddDescription.0.8.13.patch		+1
> 06a_bsd_next_cleanup.0.8.13.patch	+1
> 08_Wall_htpr.0.8.13.patch		+1
> 09_log_setgrpprivs.0.8.13.patch		+1

you forgot 01a as well...

> 
> My problem with 02 is that h->h_name should already be a char *, so why
> is the cast necessary? Sounds more like a bug in Linux's headers to me (or
> the compiler).

Don't get me wrong it's a cosmetic fix, but why 0 instead of -1?

> 
> (SCO still produces one warning because of a similar thing ... I don't like
> to fix warnings which are wrong!)
> 
> Cheers,
> 
> Ben.
> 


-- 
Aram W. Mirzadeh			     http://www.qosina.com/~awm/
MIS Manager	To err is human, to really screw up you need a computer.
Qosina Corporation			          http://www.qosina.com/
Email: awm@qosina.com				            awm@liii.com

Mime
View raw message