Re: [Springframework-developer] Re: [Springframework-user] applicationContext.xml for servlet 2.2

[Darren D. <da...@da...>]

On Monday 11 August 2003 19:43, j=C3=BCrgen h=C3=B6ller [werk3AT] wrote:
> Hi Darren,
>
> You're actually the first one to ask for reviving ContextLoaderServlet ;-)

I guess a lot of in-house developers won't know anything of Spring yet.  Th=
ose=20
that do will be battling against corporate sluggishness like I do so I'm su=
re=20
they'll appreciate it ;)


> Of course, the ContextLoader delegate already allowed being called in a
> single line from within a custom startup servlet. So we're now re-adding
> ContextLoaderServlet for pure convenience on Servlet 2.2, to avoid the ne=
ed
> for a custom startup servlet with a ContextLoader.initContext call.

I'm not sure from what you say here whether the attached file I sent came=20
through OK, but it was a very cut down version of the original servlet whic=
h=20
simply does just what you said and makes the single line call to=20
ContextLoader.initContext from the init() method.  I also amended the servi=
ce=20
method (just in case it ever mistakenly got called) and changed javadoc to=
=20
highlight the deprecation.  My feeling was just that it was very definitely=
=20
framework code rather than application code and being so simple, it made=20
sense to add it.


> Frankly, I haven't been aware that the listeners have been our only Servl=
et
> 2.3 dependency. But the main reason for removing the loader servlet was
> consistency: We also got WebAppRootListener and Log4jConfigListener, and
> splitting each of them into a listener, a servlet, and a delegate class t=
oo
> seemed like overkill. Do you see any need for this?

If all of the services can be encapsulated in one servlet, I'd suggest doin=
g=20
that rather than maintaining separate servlets and their associated web.xml=
=20
entries.  I'm happy to keep it up to date for you since I'll be using it=20
anyway.  It may not be ideal from a design point of view, but this is only =
a=20
2.2 compromise for us laggers.  A further updated version with duplicate=20
log4j and webAppRoot init code is attached here.  Note that I've NOT tested=
=20
this yet (it compiles ok), but I'll drop it into a WebSphere 4.0x server=20
tomorrow morning and try to break it.  If you can hold off committing=20
anything until then, I can send you the proper results and any further=20
changes required.

All the best,
=2D-=20

Darren Davison
Public Key: http://www.davison.uk.net/key.jsp