Re: webdav wagon changes

[Jason van Zyl]

On 20-May-08, at 9:17 AM, Brian E. Fox wrote:

I agree a single provider is less confusing. What if jackrabbit is  
the trunk and we make slide a branch? If the api hasn't changed,  
then it should be ok.


It's not the API that matters, that has to be the same. It's the way  
it behaves. You lose nothing have two providers and not changing the  
way everyone uses it now while having the option to use something new.  
You have no idea what the behavior may, or may not, push on to users  
of WebDAV. Just avoid the potential problem of calling it the same  
thing, finding out there is a problem, trying to roll it back, and by  
some weirdism in maven-artifact users get the wrong provider. Don't  
make it more complicated then it has to be. People can choose the new  
implementation when they want.

We have enough problems with users complaining about the auto-magical  
update. Do we really want to explain "Oh we called it the same thing  
but we completely changed the implementation while it was in beta."

-----Original Message-----
From: [EMAIL PROTECTED] [EMAIL PROTECTED] On Behalf Of Nathan  
Beyer
Sent: Tuesday, May 20, 2008 10:26 AM
To: wagon-dev@maven.apache.org
Subject: Re: webdav wagon changes

I'd prefer a single 'webdav' provider. If there are concerns of  
passivity or
correctness, then I'd suggest a call out for help testing it, which  
could
identify any possible regressions. The slide-impl is pretty bad and  
I'd be
willing to live and work through any bumps for a long-term win.

If there's a SNAPSHOT or a way of building one and some quick  
instructions
on how to replace the slide impl in 2.0.9, I'd certainly be able to  
start
some larger-scale integration testing.

-Nathan

On Tue, May 20, 2008 at 12:43 AM, Brett Porter <[EMAIL PROTECTED]>  
wrote:

Hi,

I just finished some work to migrate the webdav provider from slide  
to
jackrabbit using the patch from James Dumay. Slide leaked some file  
handles
and is an unsupported project. In the process I added some more  
tests around
transfer listeners as some providers weren't registering them  
correctly.

Jason suggested on IRC that it might be better to retain the slide  
webdav
provider as-is and move the new code to wagon-webdav-jackrabbit.

Any opinions on this before I go ahead and do that?

I'm thinking of still renaming the slide wagon to wagon-webdav- 
slide in
such a case, so that on upgrading the version they are forced to  
choose, but
still have access to the slide one if needed.

The only issue I can see is it clashing with the other one built in  
to
2.0.9 for the plexus 'dav' component, but I haven't tried to see if  
it wins
out correctly if used.

BTW, I've also brought the SCM provider in to trunk from the  
sandbox - it's
apparent from JIRA that a number of people are using it, so I think  
we
should support it for the limited use cases it currently supports.

Cheers,
Brett

--
Brett Porter
[EMAIL PROTECTED]
http://blogs.exist.com/bporter/


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder,  Apache Maven
jason at sonatype dot com
----------------------------------------------------------

Selfish deeds are the shortest path to self destruction.

-- The Seven Samuari, Akira Kirosawa 




---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]