-1 for the proposal as is.
We certainly understand and in favor of your proposal for web.xml cleanup 
and API ease of use. However, we have requirements to support both 2.4 and 
2.5.
Is there a way we can manage this to provide backward compatibility to 
take care of any possible API conflicts/breakages ?

Thanks,
Han



From:   Paul Lindner <[email protected]>
To:     [email protected]
Date:   05/12/2010 04:38 PM
Subject:        Re: upgrade to servlet-api 2.5
Sent by:        [email protected]



You should be fine.  This is mostly for the benefit of people developing 
off
the code base and less about the people consuming the artifacts.

I can revert back to 2.4 if there's a compelling argument.

On Wed, May 12, 2010 at 1:24 PM, Mark D Weitzel <[email protected]> 
wrote:

> Paul,
>
> Would you be kind enough to let me check with a few of our product teams
> inside of IBM? I *think* this may give us some heartache b/c we'll need 
to
> deploy on the current version. So I definitely don't want to vote yes or
> no without having all the facts.
>
> Thanks,
>
> -Mark W.
>
>
>
> From:
> John Hjelmstad <[email protected]>
> To:
> [email protected]
> Date:
> 05/12/2010 03:53 PM
> Subject:
> Re: upgrade to servlet-api 2.5
>
>
>
> +1 from here.
>
> On Wed, May 12, 2010 at 12:12 PM, Paul Lindner
> <[email protected]>wrote:
>
> > I'm going ahead with this change
> >
> > * web.xml has already had servlet-2.5 markup for a long time...
> > * I forgot that I raised this issue long ago and no one complained.
> >
> >
> > On Wed, May 12, 2010 at 11:20 AM, Paul Lindner <[email protected]
> > >wrote:
> >
> > > I'd like to move us to servlet-api 2.5
> > >
> > > pros:
> > > * Allows easier web.xml configuration for various url patterns.
> > > * Allows for use of getContextPath() to ease configuration hassles
> > >
> > > cons:
> > > * People can't deploy source builds directly to tomcat 5.5 or older
> > > containers.
> > >
> > > To make this easy I'll wrap calls to getContextPath() into it's own
> class
> > > that can be customized with hard coded or configured value.
> > >
> > >
> >
>
>
>

Reply via email to