In a message dated 01-09-08 01:29:57 EDT, Ian wrote...

>  looking back at justin's original request.
>  the code was posted ~1 month ago.
>  he reviewed it, he thought it was OK.
>  and asked.. 
>  'guys.. should we have this module included?'
>  
>  a simple enough question.

Yes, it was... ( a question I have asked before myself a few 
times on this forum ) and no one really answered so obviously
folks were just a little 'too busy' to take up the gauntlet
when you first posted it. It wasn't a sign of rejection and 
usually never is at Apache... sometimes people are 
just too busy to 'go there' at the moment a question is posed.
  
>  it then degenerated into a flame war. 

Only after nothing happened with it for awhile and then
Justin suddenly said 'I want this all the way in the CORE RIGHT NOW'.
  
>  It is hard to read you long and flamebait ridden comments
>  Kevin,  alot of people stop reading after the third 'asshole'
>  reference. I seem to remember you mentioned that ZLib wasn't 
>  the best candidate for realtime compression.

I did... it's the reason that mod_gzip does NOT use it. I put
my code where my mouth was on mod_gzip 1.3.19.1a and
never looked back...

...but that doesn't mean I was right. 

Perhaps I know a lot more now than I did then ( over a year
and thousands and thousands of downloads and installations ago )
and the actual and/or imagined 'problems' with using ZLIB as a 
real-time data compression engine might have easy solutions
for anyone who really wants to use it that way.
  
>  This whole flame war has left me with a bad taste in my mouth 
>  and I don't think GZ is going anywhere for the time-being. so 
>  I have gone back to working on other things 

There WILL be IETF Content-encoding filtering in Apache, 
I assure you, no matter where it comes from. Ryan Bloom and 
Greg Stein and many others didn't bust their butts squeezing a 
complicated filtering scheme into an already attention-needing 2.0 
codebase just for SSL or mod_charset. Indeed... during the design 
phase of the filtering the only really good example anyone could ever
come up with for using filtering at all was 'compressing responses'.

It WILL happen... and if no one has said 'thanks' to you for
showing that it is at least possible to use filtering for such a
task then let me be the first to say so. Thanks. mod_gz DOES
seem to show that the filtering is ready to convert content.

The only question on the table really seems to be WHEN this
is all going to happen, not IF. Stay tuned.
  
>  My point to Jim J was really a question as to why. I submit code
>  regularly to the codebase, and commit patches. I do not see the module
>  I personally wrote as being any different. I do not claim to understand
>  how apache works, never did. 
>  
>  I really am sick of responding to you Kevin, because all you will
>  do is ridicule me and other posters.

I won't apologize for my passion on this topic but I will hereby
apologize for 'running off at the hand' recently. There's some 
history there that just keeps creeping back up.
  
> If you have anything constructive to say, then please do. Otherwise
> shut the fuck up and let everyone get on with it.

Fair enough.

There is no way that I, alone, can either cause or prevent 
Apache from doing anything. All I can do is send emails
when I have something to say and send up some code 
(if/when I am ready) and see if it flys. If the vote to include 
mod_gz or anything else goes down right now then that's the 
end of  it. Heck... I am not even a committer so my votes don't 
count for anything.

Like I said in previous message... I have been thinking about
'What's good for Apache' on this topic over last day or so and
while chaning my vote to -1 for inclusion of mod_gz AT THIS
MOMENT I am now leaning toward the conclusion that ZLIB
is probably the way for Apache to go for now if they want to
compress something.

It just starts to make sense. Apache is not in in the compression
business and there is hardly any other software in the world that
supports as many different platforms as Apache does. 

Dr. Mark Adler ( co-author of ZLIB ) is currently on
our payroll and we have switched our agenda to a specific 
discussion about whether using ZLIB in Apache as a 24x7
compression engine is really the right way to go ( for Apache ).

If that isn't 'trying to help' I don't know what is...

..but ( like you I suppose ) I have other irons in the fire at the moment
( I was actually on vacation last weekend when the broil-fest began 
and could not have uploaded anything even if I wanted to )
and Mark (Adler) himself is a very busy guy ( Mission director for latest 
Mars Rover Expedition at JPL in Pasadena ) so the real scoop on 
all this ( if anyone is still willing to wait for it ) might take a few days 
and a some more discussion to 'iron out'. I don't know. It's just a 
feeling.

Later...
Kevin


Reply via email to