james-server-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Bernd Fondermann" <bernd.fonderm...@googlemail.com>
Subject Re: Status of trunk [WAS: Re: [jira] Commented: (JAMES-876) cannot start as spring-deployment when built from trunk as described in HOW-TO.TXT]
Date Mon, 06 Oct 2008 18:27:51 GMT
On Sun, Oct 5, 2008 at 10:59, Robert Burrell Donkin
<robertburrelldonkin@gmail.com> wrote:
> On Sat, Oct 4, 2008 at 11:40 AM, Stefano Bagnara <apache@bago.org> wrote:
>> Robert Burrell Donkin ha scritto:
>>> On Fri, Oct 3, 2008 at 8:15 AM, Stefano Bagnara <apache@bago.org> wrote:
>>>> Robert Burrell Donkin ha scritto:
>>>>> On Wed, Oct 1, 2008 at 1:39 PM, Stefano Bagnara <apache@bago.org>
>>>>> <snip>
>>>>>> Just to be clear I also guess that this specific issue has been
>>>>>> introduced by Robert and not by me, and I simply decided to answer
>>>>>> because I don't think Robert made a mistake. We are in a big refactoring
>>>>>> and we decided to work in trunk: this is expected until someone will
>>>>>> write integration tests for the deployment.
>>>>> i've taken a look and i think the problem is that i didn't copy the
>>>>> latest james configuration file into spring
>>>> Very likely. In past I reported the issue of the duplicate
>>>> configuration, but we had not a solution and Bernd also suggested that
>>>> the duplication was there by purpose.
>>> could anyone elaborate on this?
>>> - robert
>> http://markmail.org/message/iesxwotnashlfytb
> thanks
> <blockquote cite='http://markmail.org/message/iesxwotnashlfytb'>
> First of all, this is intentional. (Whether this is good or bad or
> should be changed is another thing.) I refrained from copying them
> over from phoenix-deployment at build time, because changes in them
> might trigger more work within the spring-deployment module and there
> where times when they had to be adopted. This is also mentioned in the
> </blockquote>
> i think it may be time to revise this

the spring-deployment was developed as an add-on to the regular
phoenix-deployment. so I wanted to avoid that it could break easily
because nobody was testing it. well, this has proven to be very
realistic. it broke, because nobody was testing it.

if we change our strategy to fully support both deployments I agree
that we have certainly have to revise this.
and it should work as soon as we have the proposed smoke tests.


To unsubscribe, e-mail: server-dev-unsubscribe@james.apache.org
For additional commands, e-mail: server-dev-help@james.apache.org

View raw message