You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by Chuck Murcko <ch...@topsail.org> on 2000/06/09 22:17:55 UTC
1.3.13 mod_proxy
Are there still objections to adding the HTTP/1.1 patch to mod_proxy in
1.3.13, even though we're adding PUT and therefore have to test the
thing anyway?
Just curious.
--
Chuck
Chuck Murcko
Topsail Group
chuck@topsail.org
Re: 1.3.13 mod_proxy
Posted by Graham Leggett <mi...@sharp.fm>.
Greg Stein wrote:
> On Fri, Jun 09, 2000 at 04:17:55PM -0400, Chuck Murcko wrote:
> > Are there still objections to adding the HTTP/1.1 patch to mod_proxy in
> > 1.3.13,
>
> I'm -0 here. It would be great if the patch could be broken down into chunks
> that can be reviewed, however.
That would be a bit of a problem.
The patch is essentially a rewrite of the caching handler behind the
proxy code. The original proxy suffered a design flaw in that it didn't
handle the updating of cached objects as required by HTTP/1.1. Basically
the original code said "oops, we really should do this, but we don't".
What I can do is break the patch up into the "core code" and the
peripheral bits, those bits being subroutines that parts of the main
code were broken into to make the flow easier to follow. These
peripheral bits can then be looked at as they stand, it will be easier
to digest that way.
Regards,
Graham
--
-----------------------------------------
minfrin@sharp.fm "There's a moon
over Bourbon Street
tonight..."
Re: 1.3.13 mod_proxy
Posted by Greg Stein <gs...@lyra.org>.
On Fri, Jun 09, 2000 at 04:17:55PM -0400, Chuck Murcko wrote:
> Are there still objections to adding the HTTP/1.1 patch to mod_proxy in
> 1.3.13,
I'm -0 here. It would be great if the patch could be broken down into chunks
that can be reviewed, however.
> even though we're adding PUT and therefore have to test the
> thing anyway?
Huh? "adding PUT" ... what does this part mean?
Cheers,
-g
--
Greg Stein, http://www.lyra.org/
Re: 1.3.13 mod_proxy
Posted by Greg Stein <gs...@lyra.org>.
On Sun, Jun 11, 2000 at 09:33:58AM -0400, Chuck Murcko wrote:
> "William A. Rowe, Jr." wrote:
> >
> > I'm confused too about your PUT comments... and I walked back
> > through the mod_proxy commits to try to figure it out...
> >
> So am I. 8^)
>
> > I wouldn't mind passing through PUT if that is a key request,
> > I'd like to see 1.3.13 roll with the obvious improvements
> > that users have grown impatient to see. But I'm sure HTTP/1.1
> > will cause all sorts of new problems...
> >
> > And for the same reason as I decided to incorporate the Win9x
> > services, 1.3.13 might be a good target, since it allows us to
> > discover if there are side effects out there, and roll the
> > best danged 2.0 server we possibly can.
> >
> > But can we assure that the feature can be disabled for less
> > adventerous administrators :-?
> >
> Actually it would have to be disabled by default. But I've got to scan
> the archives myself to see if that's really the commit that went in. I
> know I saw *something* go into the thing, and I would also rather
> concentrate on 2.0.
This are a bit confusing here :-) ... but if the point is that mod_proxy is
being patched to allow a PUT to pass through, then I would also want to see
the WebDAV commands allowed through. No sense in doing a teeny subset.
Cheers,
-g
--
Greg Stein, http://www.lyra.org/
RE: 1.3.13 mod_proxy
Posted by "William A. Rowe, Jr." <wr...@lnd.com>.
You saw me commit two fixes, and two others go in as well...
One patch was to fix BeOS, one to fix Win32 threads, one to
fix write->send and read->recv for Win32 SSL proxing, and finally
one to fix the stat() against directories on Win32.
All were each of a very focused scope, I don't have the proxy
expertise to commit a significant patch to that module
Bill
> -----Original Message-----
> From: chuck@keel.topsail.org
> [mailto:chuck@keel.topsail.org]On Behalf Of
> Chuck Murcko
> Sent: Sunday, June 11, 2000 8:34 AM
> To: new-httpd@apache.org
> Subject: Re: 1.3.13 mod_proxy
>
>
> "William A. Rowe, Jr." wrote:
> >
> > I'm confused too about your PUT comments... and I walked back
> > through the mod_proxy commits to try to figure it out...
> >
> So am I. 8^)
>
> > I wouldn't mind passing through PUT if that is a key request,
> > I'd like to see 1.3.13 roll with the obvious improvements
> > that users have grown impatient to see. But I'm sure HTTP/1.1
> > will cause all sorts of new problems...
> >
> > And for the same reason as I decided to incorporate the Win9x
> > services, 1.3.13 might be a good target, since it allows us to
> > discover if there are side effects out there, and roll the
> > best danged 2.0 server we possibly can.
> >
> > But can we assure that the feature can be disabled for less
> > adventerous administrators :-?
> >
> Actually it would have to be disabled by default. But I've got to scan
> the archives myself to see if that's really the commit that went in. I
> know I saw *something* go into the thing, and I would also rather
> concentrate on 2.0.
>
> If the proxy changes already committed require full testing, then I
> think the HTTP/1.1 patch would not add too much to that headache. If
> really nothing much has changed in the proxy, then I'm willing to hold
> off putting it in, since there are only two +1 for it so far.
> --
> Chuck
> Chuck Murcko
> Topsail Group
> chuck@topsail.org
>
Re: 1.3.13 mod_proxy
Posted by Chuck Murcko <ch...@topsail.org>.
"William A. Rowe, Jr." wrote:
>
> I'm confused too about your PUT comments... and I walked back
> through the mod_proxy commits to try to figure it out...
>
So am I. 8^)
> I wouldn't mind passing through PUT if that is a key request,
> I'd like to see 1.3.13 roll with the obvious improvements
> that users have grown impatient to see. But I'm sure HTTP/1.1
> will cause all sorts of new problems...
>
> And for the same reason as I decided to incorporate the Win9x
> services, 1.3.13 might be a good target, since it allows us to
> discover if there are side effects out there, and roll the
> best danged 2.0 server we possibly can.
>
> But can we assure that the feature can be disabled for less
> adventerous administrators :-?
>
Actually it would have to be disabled by default. But I've got to scan
the archives myself to see if that's really the commit that went in. I
know I saw *something* go into the thing, and I would also rather
concentrate on 2.0.
If the proxy changes already committed require full testing, then I
think the HTTP/1.1 patch would not add too much to that headache. If
really nothing much has changed in the proxy, then I'm willing to hold
off putting it in, since there are only two +1 for it so far.
--
Chuck
Chuck Murcko
Topsail Group
chuck@topsail.org
RE: 1.3.13 mod_proxy
Posted by "William A. Rowe, Jr." <wr...@lnd.com>.
I'm confused too about your PUT comments... and I walked back
through the mod_proxy commits to try to figure it out...
I wouldn't mind passing through PUT if that is a key request,
I'd like to see 1.3.13 roll with the obvious improvements
that users have grown impatient to see. But I'm sure HTTP/1.1
will cause all sorts of new problems...
And for the same reason as I decided to incorporate the Win9x
services, 1.3.13 might be a good target, since it allows us to
discover if there are side effects out there, and roll the
best danged 2.0 server we possibly can.
But can we assure that the feature can be disabled for less
adventerous administrators :-?
Bill
> -----Original Message-----
> From: chuck@keel.topsail.org
> [mailto:chuck@keel.topsail.org]On Behalf Of
> Chuck Murcko
> Sent: Friday, June 09, 2000 3:18 PM
> To: Apache Development
> Subject: 1.3.13 mod_proxy
>
>
> Are there still objections to adding the HTTP/1.1 patch to
> mod_proxy in
> 1.3.13, even though we're adding PUT and therefore have to test the
> thing anyway?
>
> Just curious.
> --
> Chuck
> Chuck Murcko
> Topsail Group
> chuck@topsail.org
>