You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@struts.apache.org by Ted Husted <hu...@apache.org> on 2004/09/01 00:31:27 UTC
Re: [VOTE] Move minimum to 2.3 (was Re: Changing how CommonsMultipartRequestHandler handles text parameters?)
On Mon, 30 Aug 2004 10:14:15 -0700, Craig McClanahan wrote:
>�I also suspect, given our track record :-), that re-engineering
>�Struts from scratch based on the latest platform APIs wouldn't take
>�more time than a gradual refactoring from the current code.
I sometimes wonder if not getting a clean start is why things keep taking so long. :) The codebase was not designed to be easily tested. People are skittish about making significant changes, and so we tread softly and slowly. To mangle an old chestnut: I've been test-first, and I've been test-last. Test-first is better (and faster!).
But I probably won't be involved with new development myself, and them that do the work can make the decisions :)
Anyway, it does sound like the consensus is to create STRUTS_1_2_BRANCH at the 1.2.2 tag, dub the HEAD 1.3.0, and document the minimum for the HEAD as Servlet 2.3.
-Ted.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@struts.apache.org
For additional commands, e-mail: dev-help@struts.apache.org
Re:Move minimum to 2.3 (was Re: Changing how CommonsMultipartRequestHandler
handles text parameters?)
Posted by Vic <vi...@portalvu.com>.
Craig McClanahan wrote:
<snip>
focused on the "revolution" approach rather than "evolution".
> That's not to say that we (as a project) can't do both in parallel.
Ted wrote: " Hopefully, initiatives like Struts Chain and Portlet
support will help us make the HTTP layer an "integration" layer, rather
than a "system" layer, as it is today. "
</snip>
Just to state the obvious, in CoR Context (a map really) the signature
does not change based on request/response or anything realy.
The Struts chain code based on Commons CoR code in CVS can support 2.3
and 2.4, Portlet, Faces, sitemesh, Soa, whatever.
Even the signature is defined.
So 2.3 vs 2.4 vs parallel is minor point, if you just leverage the Craig
code in there already. I assume both evolution and revolution would
leverage CoR (and not fork the limited resources).
If somone smart can tune the "base" classes, then a 2.3 and 2.4 can just
be packages under, just like portlet and faces.
.V
--
Please post on Rich Internet Applications User Interface (RiA/SoA)
<http://www.portalvu.com>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@struts.apache.org
For additional commands, e-mail: dev-help@struts.apache.org
Re: [VOTE] Move minimum to 2.3 (was Re: Changing how CommonsMultipartRequestHandler handles text parameters?)
Posted by Ted Husted <hu...@apache.org>.
On Wed, 01 Sep 2004 01:26:01 -0700, Craig McClanahan wrote:
>�Given when (in the evolution of Java best practices) Struts was
>�developed, I've been delighted at how long it has remained viable.
Absolutely. It well done well in the first instance, and has worked well ever since.
I'm still being told that the vast majority of new projects are based on Struts.
Many of us have retrofitted some fine test suites to Struts. But products like Struts TestCase are much more complex than they should have to be.
Mainly (as we know), because of direct use of HTTP APIs in the framework. Hopefully, initiatives like Struts Chain and Portlet support will help us make the HTTP layer an "integration" layer, rather than a "system" layer, as it is today.
>�Given where technology has proceeded since then, I'm personally
>�going to be focused on the "revolution" approach rather than
>�"evolution". That's not to say that we (as a project) can't do both
>�in parallel.
Absolutely. If someone were to start a revision of Struts based on Servlet 2.4, that would seem like an excellent time to start fresh with a test-first design.
Meanwhile, others (like me) can keep the original codebase patched.
>�To the extent that this (you not being involved in the new
>�development) turns out to be true, I'll be sad for not getting to
>�continue to appreciate the immense number of high quality
>�contributions you've made to Struts (in code, documentation,
>�process, and community building) -- and glad to see that you
>�haven't given up on open source projects at Apache :-). �Good luck
>�with your future endeavors!
Since it's a .NET team, we're not using Struts in my day-job now, but we do use iBATIS, so that's where the bulk volunteer hours are going. iBATIS has applied to the incubator, and the vote should be closing soon. As soon as the CLAs clear, the Apache MyFaces team should hit-the-ground-running too.
>�But we won't have any problem keeping your committer access readily
>�usable either :-).
I can still apply patches to the Struts I remember, but new Struts development isn't scratching any of my itches right now. But, if anyone else is interested, they should certainly pickup the torch and run with it. Don't wait on me :)
-Ted
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@struts.apache.org
For additional commands, e-mail: dev-help@struts.apache.org
Re: [VOTE] Move minimum to 2.3 (was Re: Changing how CommonsMultipartRequestHandler handles text parameters?)
Posted by Craig McClanahan <cr...@gmail.com>.
On Tue, 31 Aug 2004 18:31:27 -0400, Ted Husted <hu...@apache.org> wrote:
> On Mon, 30 Aug 2004 10:14:15 -0700, Craig McClanahan wrote:
> >�I also suspect, given our track record :-), that re-engineering
> >�Struts from scratch based on the latest platform APIs wouldn't take
> >�more time than a gradual refactoring from the current code.
>
> I sometimes wonder if not getting a clean start is why things keep taking so long. :)
This is definitely a factor.
> The codebase was not designed to be easily tested. People are skittish about making significant changes, and so we tread softly and slowly. To mangle an old chestnut: I've been test-first, and I've been test-last. Test-first is better (and faster!).
Given when (in the evolution of Java best practices) Struts was
developed, I've been delighted at how long it has remained viable.
Given where technology has proceeded since then, I'm personally going
to be focused on the "revolution" approach rather than "evolution".
That's not to say that we (as a project) can't do both in parallel.
>
> But I probably won't be involved with new development myself, and them that do the work can make the decisions :)
>
To the extent that this (you not being involved in the new
development) turns out to be true, I'll be sad for not getting to
continue to appreciate the immense number of high quality
contributions you've made to Struts (in code, documentation, process,
and community building) -- and glad to see that you haven't given up
on open source projects at Apache :-). Good luck with your future
endeavors!
But we won't have any problem keeping your committer access readily
usable either :-).
> Anyway, it does sound like the consensus is to create STRUTS_1_2_BRANCH at the 1.2.2 tag, dub the HEAD 1.3.0, and document the minimum for the HEAD as Servlet 2.3.
+1
>
> -Ted.
Craig
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@struts.apache.org
For additional commands, e-mail: dev-help@struts.apache.org