Guidance needed

[Henrik Vendelbo <hv...@bluprints.com>]

As I said a couple of days ago I would like to get to work on a configuration front end module, but first I guess I'd better get to know the Apache source.

I have NO CLUE though how far along the whole thing is. I've tried compiling for Win32. It bugged out in quite a number of places, so I figure that there's quite a distance to go. Should I forget about trying to compile the whole thing ?

I'd like to get to work with something related to config & logging ? Is there a todo list ? Who's coordinating ?

Thoughts ?

\Henrik

Re: Guidance needed

[Dean Gaudet <dg...@arctic.org>]

On Sun, 17 Oct 1999, Ryan Bloom wrote:

> is where our passion lies.  I happen to be one of those people.  I don't
> think, and I could be wrong, that Dean has committed anything to APR yet.
> It isn't what he wants to work with.  So, Dean is working in Apache.

actually dean is mostly pursuing other passions at the moment, he's only
checking in now and then to add his own particular brand of anal
commentary/guidance/whatever.

Dean

Re: Guidance needed

[Ryan Bloom <rb...@raleigh.ibm.com>]

> But there are those who have write access, and those who don't (no pun). So if 
> you don't like me, then I guess you find somebody also working with the same 
> stuff that you want to work on and exchange patches with them.

That isn't the way it works.  There are people who have write access, and
those who don't.  That is because to give EVERYBODY write access is just
crazy.  Write access is given based on merit.  In the last year and a
half, I have never seen a patch be rejected because of who sent it.  We
reject patches because we don't like how they were done, or because they
aren't a good idea.  Either way, we explain why the patch was rejected,
and hope that people re-submit with our suggestions if it's appropriate.
There is a group of people, because that way we can police each other.  If
I decide to reject a patch because of who sent it, Ken or Dean or anybody
else can give me a good whap over the head.  It also makes me look bad.
This is why we don't pay attention to who wrote the patch, just to what
the patch is for.

> I'm just lacking a mental picture of the status, that's all.

The status is all over the map.  APR on UNIX is very far along.  On
windows, it's not quite as far.  Apache itself is looking good, but it has
a ways to go.  Unfortunately, a good portion of our time is being spent
getting APR right, so it can be integrated into Apache.

> I'll stick to 2.0, and help out where I can.
> 
> I'm just trying to figure out what is the current focus, and which part are almost done.
> It seems that APR is the current focus, nes pas ?

The focus is where it always is.  Where ever each individual person wants
to put their time.  If you look at APR, you will see that it is really
about four or five people committing 80% of the code.  That's because that
is where our passion lies.  I happen to be one of those people.  I don't
think, and I could be wrong, that Dean has committed anything to APR yet.
It isn't what he wants to work with.  So, Dean is working in Apache.
Apache has a lot of work to be done in it.  A good portion of that, is
cleaning up the code, by putting APR into it.

I really think the way for you to progress from here, is to take a look at
a small piece that needs to be done, and submit a patch.  We will look at
it, and either give feedback or commit it.  I really can't describe our
process any better, so I really think the way to continue, is to dive in a
try it.  :)

Ryan

_______________________________________________________________________
Ryan Bloom		rbb@raleigh.ibm.com
4205 S Miami Blvd	
RTP, NC 27709		It's a beautiful sight to see good dancers 
			doing simple steps.  It's a painful sight to
			see beginners doing complicated patterns.	

Re: Guidance needed

[Henrik Vendelbo <hv...@bluprints.com>]

> From: Ryan Bloom <rb...@raleigh.ibm.com>
> well.  The 2.0 tree compiles most days on Unix, but there aren't as many
> Windows developers looking at that part of the tree, so it lags a bit
> behind the other ports.
> 
> As far as who is cooridinating the effort, that isn't the way Apache tends
> to work.  We are group of volunteers who do what they are interested on
> in their own time.  Nobody steps up to coordinate the effort, until just
> before we ship a release.  Then, the person who does step up, is only
> responsible for getting that release out the door.

But there are those who have write access, and those who don't (no pun). So if 
you don't like me, then I guess you find somebody also working with the same 
stuff that you want to work on and exchange patches with them.

> There is no one over-riding voice leading us in one direction.  There are
> a group of voices, suggesting the correct direction to head.

Obviously. 

> I don't believe the source is as far behind as you do.  You happen to have
> come in at a bad time, when I was trying to move around a lot of function
> parameters, and that broke the code on platforms that I was unable to test

I'm just lacking a mental picture of the status, that's all.

> on.  Apache 2.0 is NOT production quality, and it is missing a lot of
> features that 1.3 has already.

Well, that's development.... :-)

> My own suggestion, is that if all you want is to learn the code, use 1.3.
> If you want to help with the current code base, use 2.0

I'll stick to 2.0, and help out where I can.

I'm just trying to figure out what is the current focus, and which part are almost done.
It seems that APR is the current focus, nes pas ?

\Henrik

Re: Guidance needed

[Ryan Bloom <rb...@raleigh.ibm.com>]

The Apache 2.0 tree is progressing as quickly as possible.  If you want to
work with a tree that is stable, and is most likely going to compile and
run EVERY day, then you really need to work on 1.3.X.  The problem with
this tree, is that the 2.0 tree is substantially different from the 1.3.X
tree, and any work you do in 1.3.X will most likely not transfer over very
well.  The 2.0 tree compiles most days on Unix, but there aren't as many
Windows developers looking at that part of the tree, so it lags a bit
behind the other ports.

As far as who is cooridinating the effort, that isn't the way Apache tends
to work.  We are group of volunteers who do what they are interested on
in their own time.  Nobody steps up to coordinate the effort, until just
before we ship a release.  Then, the person who does step up, is only
responsible for getting that release out the door.

There is no one over-riding voice leading us in one direction.  There are
a group of voices, suggesting the correct direction to head.

I don't believe the source is as far behind as you do.  You happen to have
come in at a bad time, when I was trying to move around a lot of function
parameters, and that broke the code on platforms that I was unable to test
on.  Apache 2.0 is NOT production quality, and it is missing a lot of
features that 1.3 has already.

My own suggestion, is that if all you want is to learn the code, use 1.3.
If you want to help with the current code base, use 2.0

Ryan

On Sat, 16 Oct 1999, Henrik Vendelbo wrote:

> As I said a couple of days ago I would like to get to work on a configuration front end module, but first I guess I'd better get to know the Apache source.
> 
> I have NO CLUE though how far along the whole thing is. I've tried compiling for Win32. It bugged out in quite a number of places, so I figure that there's quite a distance to go. Should I forget about trying to compile the whole thing ?
> 
> I'd like to get to work with something related to config & logging ? Is there a todo list ? Who's coordinating ?
> 
> Thoughts ?
> 
> \Henrik
> 

_______________________________________________________________________
Ryan Bloom		rbb@raleigh.ibm.com
4205 S Miami Blvd	
RTP, NC 27709		It's a beautiful sight to see good dancers 
			doing simple steps.  It's a painful sight to
			see beginners doing complicated patterns.