httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Greg Stein <gst...@lyra.org>
Subject Re: cvs commit: httpd-2.0/server/mpm/perchild perchild.c
Date Thu, 07 Dec 2000 08:03:18 GMT
On Wed, Dec 06, 2000 at 07:20:43PM -0800, rbb@covalent.net wrote:
> 
> > > > Nope.  At this point, we are ONLY dealing with header data.  I'm not
> > > > convinced that this works completely, because there may be 
> > > some data on
> > > > in the bucket brigade that we haven't read yet, but the 
> > > data that I am
> > > > doing a strlen on is definately ASCII data.
> > > 
> > > Ah! Okee...
> > 
> > Sure this won't bite anyone implementing a protocol other than http/https?
> 
> By the time we are reading this data, it must be readable by the server
> itself.  If we are using some protocol other than http/https that passes
> data that is not ASCII, then the server is hosed in more ways than just
> this, so we are safe.  Am I sure this will always work?  Nope, but I am
> sure that this will work 99% of the time, so it is good enough. 

More importantly: it works for our current requirements. If some yin-yang
wants to use the perchild MPM for something other than HTTP(S), then they'll
have a bit of work to do.

Heck, they're already going to have a good chunk in terms of separating out
the request mechanisms and I/O handling from HTTP handling. Sure...
theoretically they are independent. But we have a ways to go on that :-) And
we don't need it for the Apache 2.0 release!

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/

Mime
View raw message