Quoting Bojan Smojver <[EMAIL PROTECTED]>:
> I'm using Mozilla 1.0.1, which support pipelining
Damn Mozilla... It didn't actually pipeline the requests :-(
So, I switched to telnet and discovered that there is a MAJOR difference in
results with and without the FLUSH filter (which is kinda the p
I'm doing some testing with and without this new filter that replaces EOS with
FLUSH.
Here is the test HTML file:
-
...
-
First, the bad news - I'm using Mozilla 1.0.1, which support pipelining, but I
can't make Apache send
> At the risk of racing too far ahead in this discussion, here is my
> suggestion... 2.0.43 becomes 2.1 and the MMN major does not change
> for subsequent 2.1 series releases (except for a compelling reason,
> eg a security fix -requires- a bump). Why 2.1? No technical
> reason; purely a PR tact
APACHE 2.0 STATUS: -*-text-*-
Last modified at [$Date: 2002/10/16 20:29:29 $]
Release:
2.0.44 : in development
2.0.43 : rolled October 2, 2002
2.0.42 : released September 24, 2002 as GA.
2.0.41 : rolled September 16, 2002. not rel
APACHE 1.3 STATUS: -*-text-*-
Last modified at [$Date: 2002/10/14 21:14:09 $]
Release:
1.3.28-dev: In development
1.3.27: Tagged September 30, 2002.
1.3.26: Tagged June 18, 2002.
1.3.25: Tagged June 17, 2002. Not released.
1.3.24: Tag
OK, this is my first "official" attempt at this. The basic idea is to replace
the EOS bucket at the end of the brigade with a FLUSH bucket, so that each
request is flushed down the pipe. I'm guessing this will completely break
pipelining if mod_logio is enabled, but, as someone mentioned, this is
My answer below still explains your situation. There is not way to
'restrict' requires. Each module has access to the SAME requires for a
given location. If no modules are authoritative, you probably will get
INTERNAL_SERVER_ERRORS for all unauthorized requests, right?
sterling
On Wednesda
On Wed, Oct 16, 2002 at 11:57:35AM +0200, Oliver Wulff wrote:
> We are using Apache 1.3.26 on Solaris.
> As I understood, an Apache child process will die after processing a number
> of request (defined in MaxRequestsPerChild directive). If there isn't a lot
> of traffic the processes won't die. I
On Wed, Oct 16, 2002 at 03:40:55PM -0500, Thomas Eibner wrote:
> > What I expect is a list that shows for every MMN change a short description why it
>was changed or better what has changed in the sources, and if it affects third-party
>modules and which, f.e. something like that:
> >
> > 20020
At 06:43 PM 10/16/2002, Greg Marr wrote:
>On Wed, 16 Oct 2002 19:24:22 -0400
> Joshua Slive <[EMAIL PROTECTED]> wrote:
>>I'm +1 for creating 2.1 and 2.2 trees as proposed by Bill.
>
>My one thought about this proposal is that it is unclear whether or not this is
>attempting to emulate the Perl ve
On Wed, 16 Oct 2002, Aaron Bannert wrote:
> > I like.
>
> I also like, but I also think we should stick with the "even numbered
> revisions are stable, odds are developmental" axiom.
Definitely agreed.
--Cliff
On Tue, Oct 15, 2002 at 10:22:46AM -0400, Jim Jagielski wrote:
> Bill Stoddard wrote:
> >
> > At the risk of racing too far ahead in this discussion, here is my
> > suggestion... 2.0.43 becomes 2.1 and the MMN major does not change for
> > subsequent 2.1 series releases (except for a compelling r
Quoting Greg Marr <[EMAIL PROTECTED]>:
> On Wed, 16 Oct 2002 19:24:22 -0400
> Joshua Slive <[EMAIL PROTECTED]> wrote:
> >I'm +1 for creating 2.1 and 2.2 trees as proposed by Bill.
>
> My one thought about this proposal is that it is unclear whether or
> not this is attempting to emulate the P
+1 for a 2.1 tree
Why? :
When a securityproblem/bug is found in 2.0.43, then I excpect to
get a new version who is just a "drop-in" replacement for it.
What can be accepted is
- To have to recompile all modules
- Make sourcecode changes to the modules if AND ONLY IF the api change
is directl
On Wed, 16 Oct 2002 19:24:22 -0400
Joshua Slive <[EMAIL PROTECTED]> wrote:
>I'm +1 for creating 2.1 and 2.2 trees as proposed by Bill.
My one thought about this proposal is that it is unclear whether or
not this is attempting to emulate the Perl versioning scheme. If so,
then it's backwards,
I'm +1 for creating 2.1 and 2.2 trees as proposed by Bill.
The current auth-docs problems can be fixed (and, in fact, André has
already gotten us most of the way there), but things would be much
cleaner with a new tree.
I also believe this would better communicate with users the current
state
On Thu, 2002-10-17 at 00:36, William A. Rowe, Jr. wrote:
> At 02:10 AM 10/16/2002, Bojan Smojver wrote:
>
> >Coming back to what William is proposing, the three pool approach... I'm
> >guessing that with Response put in-between Connection and Request, the Response
>will store responses for multi
On Wed, Oct 16, 2002 at 08:32:34PM +0200, G?nter Knauf wrote:
> Hi,
>
> > "William A. Rowe, Jr." <[EMAIL PROTECTED]> writes:
>
> > First, I'm pretty happy with what is going on in 2.0 HEAD now. I
> > don't think MMN is changed gratuitously, I don't think the code gets
> > destabilized a whole l
The follow patch will allow binds to a static zlib, but it's not entirely portable.
I'm wondering which ac/libtool guru might point me at:
1. a way to ADD to the objs list of just THIS module, instead of having
to second-guess the name of the mod_deflate object.
2. the portable 'shared
--disable-asis --disable-cgid --disable-cgi --disable-negotiation
--disable-dir --disable-imap --disable-actions --disable-userdir
--enable-so --enable-dav --enable-ssl --enable-maintainer-mode
$ make
$ make install
After that:
$ /opt/svnd/test-20021016-220313/bin/apu-config --libs
-ldb /home/data/j
>
> >From a quick read through the patch, it looks like the
> connection processing flow is:
>- listener thread accepts connection
>- listener passes connection to a worker through fd queue
>- worker wakes up, passes connection to event thread
> through event queue
>- worker g
Hi,
> "William A. Rowe, Jr." <[EMAIL PROTECTED]> writes:
> First, I'm pretty happy with what is going on in 2.0 HEAD now. I
> don't think MMN is changed gratuitously, I don't think the code gets
> destabilized a whole lot on a regular basis, I think that having some
> aspects of the config chan
- Original Message -
From: "Thomas Eibner" <[EMAIL PROTECTED]>
> On Wed, Oct 16, 2002 at 11:20:07AM -0400, Bill Stoddard wrote:
> > > Hi,
> > >
> > > Why url finishing by / are not cacheable ?
> >
> > muddled thinking ? :-) I was unsure how to handle caching default index
> > pages but I s
Hi Bill,
never mind: just got reply from Brad who told me he checked in a new mpm_netware.c
that fixes the problem.
Just downloaded latest cvs snapshot and compiles now fine...
Guenter.
PS: just saw that my reply address was damaged in this post:
Guenter Knauf <[EMAIL PROTECTED]>
should be:
Gue
I spent a little time on Monday and Tuesday looking at this in the
debugger and it looks like the behavior I am seeing is probably
related to compiler optimizations. My testing showed that the fixes
you committed (64 bit atoi and conversion to use seconds) seem to
have fixed the random behavior I
On Wed, Oct 16, 2002 at 11:20:07AM -0400, Bill Stoddard wrote:
> > Hi,
> >
> > Why url finishing by / are not cacheable ?
>
> muddled thinking ? :-) I was unsure how to handle caching default index
> pages but I see no reason why we can't just delete this check.
Not for negotiation reasons? (I s
At 10:08 AM 10/16/2002, William A. Rowe, Jr. wrote:
>At 07:16 AM 10/16/2002, Günter Knauf wrote:
>>Hi Bill,
>>on NetWare linking breaks with 'Undefined symbol: APR_TO_NETOS_ERROR';
>>seems its only defined outside the '#ifdef NETWARE' block...
>
>Günter, simply cvs up and rebuild clean. That ma
At 07:16 AM 10/16/2002, Günter Knauf wrote:
>Hi Bill,
>on NetWare linking breaks with 'Undefined symbol: APR_TO_NETOS_ERROR';
>seems its only defined outside the '#ifdef NETWARE' block...
Günter, simply cvs up and rebuild clean. That macro was used
only within the netware mpm, and should be go
> Just an opinion..
> Lots of people trust you lot. Next time there is a security issue and
> you do release 2.0.x, if there is a change/new functionality that is
> beta, alpha or worse then that is extremely bad for a GA product. Start
> the 2.1.x branch!
I'll state this a slightly different way
> Hi,
>
> Why url finishing by / are not cacheable ?
muddled thinking ? :-) I was unsure how to handle caching default index
pages but I see no reason why we can't just delete this check.
>
>
> /* DECLINE urls ending in / ??? EGP: why? */
>
> if (url[urllen-1] == '/') {
> return
Hi,
Why url finishing by / are not cacheable ?
/* DECLINE urls ending in / ??? EGP: why? */
if (url[urllen-1] == '/') {
return DECLINED;
}
I delete this code and it's caching it really well
Is it possible to think a directive to enable/disable this ?
regards,
Just an opinion..
Lots of people trust you lot. Next time there is a security issue and
you do release 2.0.x, if there is a change/new functionality that is
beta, alpha or worse then that is extremely bad for a GA product. Start
the 2.1.x branch! Agree rough timescales bearing in mind you may need
At 02:10 AM 10/16/2002, Bojan Smojver wrote:
>Coming back to what William is proposing, the three pool approach... I'm
>guessing that with Response put in-between Connection and Request, the Response will
>store responses for multiple requests, right?
No, I'm suggesting a single response pool f
On Wed, 2002-10-16 at 17:54, Brian Pane wrote:
> I think you could achieve this result by appending a bucket of type
> apr_bucket_flush to the output brigade. I think that would be a
> reasonable workaround for 2.0.
This is what I wanted to do initially:
---
Hi Bill,
on NetWare linking breaks with 'Undefined symbol: APR_TO_NETOS_ERROR';
seems its only defined outside the '#ifdef NETWARE' block...
Guenter.
Revision 1.97 / (view) - annotate - [select for diffs] , Sun Oct 13 23:55:46 2002 UTC
(2 days, 12 hours ago) by wrowe
Branch: MAIN
CVS Tags: H
Hi all
We are using Apache 1.3.26 on Solaris.
As I understood, an Apache child process will die after processing a number
of request (defined in MaxRequestsPerChild directive). If there isn't a lot
of traffic the processes won't die. Is that right?
We had already about 180 child processes runnin
On Tue, 2002-10-15 at 10:31, Bill Stoddard wrote:
> Something I've been hacking on (in the pejorative sense of the word 'hack'.
> Look at the patch and you will see what I mean :-). This should apply and
> serve pages on Linux, though the event_loop is clearly broken as it does not
> timeout keep
Quoting David Burry <[EMAIL PROTECTED]>:
> I know that I'm ultimately wanting to log bytes per request pretty
> accurately to add up and tell if the whole file was transferred or not even
> if it was broken up into several byte range requests. I have given up on
> the possibility of this happeni
Quoting Brian Pane <[EMAIL PROTECTED]>:
> I think you could achieve this result by appending a bucket of type
> apr_bucket_flush to the output brigade. I think that would be a
> reasonable workaround for 2.0.
Cool, thanks. This should be easy - I can simply do it in mod_logio's output
filter, w
I know that I'm ultimately wanting to log bytes per request pretty
accurately to add up and tell if the whole file was transferred or not even
if it was broken up into several byte range requests. I have given up on
the possibility of this happening in 2.0 for a long time to come due to the
archi
On Wed, 2002-10-16 at 00:10, Bojan Smojver wrote:
> Coming back to what William is proposing, the three pool approach... I'm
> guessing that with Response put in-between Connection and Request, the Response
> will store responses for multiple requests, right? Then, once the core-filter is
> done
41 matches
Mail list logo
