Alpha 8...delaying?

[rb...@covalent.net]

Unless somebody objects loudly, I would like to delay a8 until we get
mod_include.  I am not all that interested in releasing multiple alphas in
the next two weeks, and since the last we heard about mod_include was that
it should be ready at some point this weekend, I figure we are better
waiting a day or two on this alpha than in releasing now.

Any objections?

Ryan
_______________________________________________________________________________
Ryan Bloom                        	rbb@apache.org
406 29th St.
San Francisco, CA 94131
-------------------------------------------------------------------------------

Re: Alpha 8...delaying?

[rb...@covalent.net]

> There is one anomaly which I'm not sure is coming from mod_include. At various
> random times, when the browser has made a request it only gets part of the page
> back (no error, just thinks its at the end). I'm not sure if this is a content
> length problem or some sort of buffering/coalesceing problem (or a mod_include
> problem).
> 
> Anyway, sorry again. Please look for a post within the hour.

I'll be on the train into work when this comes in.  Once I get to work,
I'll test and commit (unless somebody beats me to it).  We roll a8 about
12:00 today.  :-)

Ryan

_______________________________________________________________________________
Ryan Bloom                        	rbb@apache.org
406 29th St.
San Francisco, CA 94131
-------------------------------------------------------------------------------

Re: Alpha 8...delaying?

["Paul J. Reder" <re...@raleigh.ibm.com>]

Please accept my apologies for the delay on mod_include. We had snow and
that caused me to have serious power problems (power browns @ every 10-15
minutes, power outages about once an hour for 5 minutes at a time for a
large part of Sun. I just love Duke Power) and when I did have power I
couldn't get at my mail. I'm sure you're all broken up about my problems
though. ;)

I did a lot of debugging and testing of mod_include this weekend and have it
working well except for one cross brigade case which I am fixing now. I will
be posting the diffs shortly.

The conditional (if, elif, else, endif) expression parser has not changed,
but that is completely self contained and can be dealt with later if required.
All of the rest of the work has been done. 

Mod_include is now fully bucketized. It can handle tags that span buckets or
brigades.

There is one wholesale copy of the directive into a contiguous buffer, and then
nearly no (p|m|c)allocs after that. Most parsing/processing work is done in place.

Simple tasks like echo, fsize, flastmod, and printenv are performed in line by
just creating and inserting buckets with the new content in place of the SSI
tag buckets (no extra brigade splitting/passing).

Execs and Includes are handled off line (causing the initial buckets to be split
and passed before the exec/include processing).

I have also fixed the quoted value processing for the directives. The docs all
indicated that if the value contained no whitespace then quotes were not required.
The code, however, required them. The code now does the right thing when processing
values. The cleanup of value processing also removed a core dump that I found.

I have created a mod_include.h file, but that content could easily be merged into
the mod_include.c code if preferred.

There is one anomaly which I'm not sure is coming from mod_include. At various
random times, when the browser has made a request it only gets part of the page
back (no error, just thinks its at the end). I'm not sure if this is a content
length problem or some sort of buffering/coalesceing problem (or a mod_include
problem).

Anyway, sorry again. Please look for a post within the hour.
-- 
Paul J. Reder
-----------------------------------------------------------
"The strength of the Constitution lies entirely in the determination of each
citizen to defend it.  Only if every single citizen feels duty bound to do
his share in this defense are the constitutional rights secure."
-- Albert Einstein