d there's a bunch of
stuff there that needs to be made available to the action(if used for
Portlet implementation). The user is one of those things. But this
should only be done if the action is actually used as a portlet. So, the
ActionContext stuff shouldn't be bloated unnecessarily.
a really powerful thing, which will be great for many things
(delegated searches come to mind). Add it :-)
/Rickard
--
Rickard Öberg
Senselogic
---
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http:/
e (instead of doing many
applet/server calls).
/Rickard
--
Rickard Öberg
Senselogic
---
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
___
Webwork-user
ly more than 5
> messages a day to _both_ lists (and often cross posted!) :)
>
> My $0.02 - IMHO we see how it goes, and we can add another list later.
I'd agree with the above. Keep it one for now, for both user and dev.
/Rickard
--
Rickard Öberg
gt; become the defacto standard in the future.
That is quite probable, sure. But I think there's enough room for more
implementations than just AspectJ.
/Rickard
--
Rickard Öberg
---
This sf.net email is sponsored by:ThinkGe
ious data appears again (the action
> doesn't executes).
With only the limited info you gave above it is impossible to answer
this. Your best bet is to add a lot of debug output, or use a real
debugger to find out what's going on. WebWork does not cache actions, so
tha
vailable for purchase/demo. I'll probably also
describe how we're using WebWork, as it seems we will be using it alot,
both for applet/server comms and actual page rendering.
/Rickard
--
Rickard Öberg
---
This sf.net email is sponsored
mably you won't be opensourcing
> this code?
Nope, this is actually my first purely closed source project :-) Feels
strange really. We might do a developer source license though, or at
least an open API for plugins.
/Rickard
--
Rickard Öberg
-
it was possible to implement AOP using plain
Java.
That's about it I guess.
/Rickard
--
Rickard Öberg
---
This sf.net email is sponsored by:ThinkGeek
Gadgets, caffeine, t-shirts, fun stuff.
nagement system). We'd like
to test it in our own stuff first, to see if it can be used in a large
system effectively. We think so, but need to try to actually know so :-)
/Rickard
--
Rickard Öberg
---
ifferent ways, etc.)
>
> Is there a rule? Perhaps I missed it in the documentation.
'text'="text" string. gives a constant value
text=evaluate the text property
This holds for all attributes except id, since that is an attribute that
is gener
ies. With JDK1.4 the overhead for that is so
small that it's rather ok to apply it just about everywhere. And boy oh
boy does it rock :-) It's so cool to have an object implement 10
interfaces but not have a single class implement more than 1... it's
what code reus
now (no, I'm not using AspectJ) and I
never ever want to go back to plain old OOP again :-))) I just love the
flexibility I get from being able to "tack on" new interfaces and call
interceptors to objects without redesigning my
Toby Hede wrote:
> The rules will still apply.
>
> I might be way off here, but it is my understanding that implementing the
> SessionAware interface simply allows the ActionContext to pass in the
> HTTPSession Map for access.
That is precisely correct.
/Rickard
--
one through the CD this becomes trivial to implement :-)
Me like it.
/Rickard
ps. and instead of EJB we're using AOP which makes the app really
ultracool, but that's another story..
--
Rickard Öberg
---
This sf.net email is spon
define your web.xml, cause .xml requests wont
> handled by webwork default web.xml.
No, the URL would be something like "foo.xml.action".
/Rickard
--
Rickard Öberg
---
This sf.net email is sponsored by:ThinkGeek
ring this subject up again. I
> think WW could benefit from the exposure, common
> infrastructure, developers, etc.
>
> So, I'll start off and say +1. What do you say?
I'm all for it. +1.
/Rickard
---
This sf.net email
e:
>
> title="Style">
> title="Style">
>
> Is there any way of changing the theme name centrally (i.e.
> webwork.properties file) instead of having to change all of my JSP pages as
> well
ld work fine by just using the method
name as is. I.e. if your method is "foo" then use "foo" as property name.
/Rickard
--
Rickard Öberg
___
Have big pipes? SourceForge.net is looking for download mirrors. W
Vaishakhi Ajmera wrote:
> hello,
> I wanted to know if my view can call only those methods from my actions
> that start with a 'get' or not.
Have you tried?
/Rickard
___
Have big pipes? SourceForge.net is l
WW valuestack.
>
> It's too bad JSTL doesn't provide a way to do this internally where you
> could implement your own version of PageContent.findAttribute similar to
> what you did for the Velocity implementation.
Yeah, but the filter thingy should do the trick though. Not
oads
request.getAttribute() and delegates to the stack before doing the
normal stuff. The semantics of that may get a little iffy, but at least
it would provide super-trivial access to the WW valuestack.
/Rickard
--
Rickard Öberg
__
my views.prop's is generated
> that way.
In that case it doesn't really matter. Also, the above action package
prefix will be ignored since XDoclet generates complete class names.
/Rickard
--
Rickard Öberg
___
Have big
Philipp Meier wrote:
> I would rather suggest to make the configuration XML-agnostic.
Well, it already is, since it's pluggable. I was talking about a more
XML-ish default.
/Rickard
--
Rickard Öberg
___
Webwork-user mailing lis
od
idea. Not sure though, since XML would make it much much more verbose.
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
; need some XSLT java extension for URL-escaping and so on. I don't know
> if I'll have the time but I don't want to give up the XSLT thingy. So
> is their anybody except me using it? Huh?
It's tough at the top you know ;-)
/Rickard
--
Rickard Öberg
___
my
> actions.xml and webwork:url tags to make this happen I would be
> grateful.
>
> What do people see as the best way to approach this sort of problem ?
> Are there better options I have missed ?
I usually do 2 (as suggested in the docs,
s = request.getAttribute("myProperty");
%>
or simply:
<%
String s = (String)ValueStack.getStack(request).findValue("myProperty");
%>
without the tag stuff.
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
le have been asking for, but we
haven't been able to figure out a good way to describe these chains and
how parameters should be transferred.
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
the box with the
current ClientServletDispatcher, but that needs to be tested.
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
Grobe, Gary wrote:
> Do you mean where I currently have the following in my properties file ...
> forms.InitFieldForm.done.success=overviews.jsp
>
> that I should use ...
> forms.InitFieldForm.done.success=redirect.action?url=overviews.jsp
Have you tried it?
/Rickard
--
his has to do with the more complicated problem of chaining actions,
which is something that needs to be dealt with soon, agreed.
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
uncommon). If you *need* the same form to go
to different commands on submit then that can be accomplished using
JavaScript to modify the value of the hidden "command" field.
/Rickard
--
Rickard Öberg
___
Webwork-use
sec)
> Average time: 1,331 (0 min:1 sec)
> Mean time: 1,362 (0 min:1 sec)
> Fastest time: 837 (0 min:0 sec)
> Slowest time: 1,886 (0 min:1 sec)
Whoa, that was not very nice. And if you remove component tag uses the
times drop?
/Rickard
--
Rickard Öberg
__
That is one way of doing it. You can do it the way I showed too.
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
27;t understand how it's possible this degradation in
> speed.
>
> I'm missing something?
It would be good if you could do a profile of where the time is actually
spent. I.e. is this really a WebWork problem?
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
G.L. Grobe wrote:
> Now that I'm back on this prob, can someone clarify the below solution for
> me as I don't understand it.
Not sure I can be more clear... what part of it don't you understand?
/Rickard
>
> - Original Message -
> From: "Ric
al since you're not really
considering using scriptlets... are you...??
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
that can result in an error state, then it is
advisable that you use the Model-2 approach and not Model-1. I.e. do not
use the action tag, but instead do an include of the action. That way
you can delegate to an error page if an exception occurs.
/Rickard
--
Rickard Öberg
an.PropertyEditor's.
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
art of the webwork taglib.
I think the easiest way is to create a servlet filter for .vm's that
simply add whatever you want to the request.
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
ow i would like to be able to do this, i
> haven't really thinked it over but this approach would give my designers
> the flexibility they need and it would be easy to use for them.
Why don't you want to use the template way? Your proposal seems to
introduce a lot of extra work
cause it to be on the
stack without executing doExecute().
/Rickard
--
Rickard Öberg
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
system to include it,
but since I'm on a modem and just about to leave for J1 I can't make a
new binary. Matt should be able to do it though.
Sorry for the screwup. There's always something like this when you think
"ok, *now* we're done" :-)
/Rickard
--
Rickar
s a side effect, this would increase
> the html result more than necessary (need additional divs and CSS class
> definition) .
Somehow in the bigger scheme of things I don't see that as a huge
obstacle :-) But maybe that's just me...
/Rickard
--
Rickard Öberg
Author of
m, and without having to
change anything in WebWork.
Good enough?
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The Middleware Company - We Build Experts!
___
Webwork-user mailing list
[E
m.. the labels use specific CSS classes that you can use. Is that what
you're after?
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The Middleware Company - We Build Experts!
___
Webwork
Taavi Tiirik wrote:
> I am afraid this is cvs faq... but can you please tell me how to get
> webwork 1.0 from cvs. I do not want to get any modifications
> that are added after release of 1.0.
I just added a CVS tag "release_1_0" which you can specify when checking
o
There's an index.html file that you should point your browser to. If
there's not then you might want to download the latest version, since
1.0 was released just now.
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect,
me class there's not an
explosion of classes, which is a common risk with using the command pattern.
That's the gist of it anyway.
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The Middleware Company - We Build Exper
Matt Baldree wrote:
> Well, WW 1.0 is out the door. Many thanks go out to everyone involved
> especially Rickard and Maurice. The only thing left now is for Rickard to
> announce WW to the world.
You bet. So, what's the laundry list? I can think of:
* TSS
* JavaLobby
* EJB-
Matt Baldree wrote:
> Rickard, are you going to release?
Sure, I can do that almost now. Unfortunately I'm in the middle of a
massive CVS update on another project and since I'm currently a modem
looser it takes awhile... as soon as it's done I'll commit my very las
deal with myself (but need to).
Are there any occasions where it is *not* necessary to escape the
output? I.e. should we always do it?
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The Middleware Company - We Build Experts!
_
can do is to (in your page) bind it to a request attribute:
and then access that in your decorator:
...
/Rickard
ps. Yes, WW+SiteMesh kicks some serious a** :-)))
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The Middleware Company
ons on how to best design our app. We're
> calling the Business Layer through a Session Facade, which returns us
> Value Objects.
Well, the above should work just fine. I'm using it with great success,
although (as noted) I use the command pattern instead of session
facade/value ob
ommonview.jsp
>
> action2=Action2
> action2.success=action2.jsp
> action2.commonview=commonview.jsp
>
> I am sure there is a trivial way of doing it without having to define common
> view for every action.
Add this:
commonview=commonview.jsp
/Rickard
_
d().getContextClassLoader() as the "switch" (but
of course at some point the underlying implementation of currentThread()
using something like a "threadlocal").
regards,
Rickard
ps. A little confession: those with good memory will remember my fierce
opposition to the "jav
of the static method access by introducing the "java:comp/env"
namespace in JNDI (which effectively is the same as a static method).
It worked there, so why not here?
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The Middleware Company -
ake sure we don't oversee the
> arguments you had before.
Hm.. I wonder what they were... if I can't remember, maybe they weren't
that good 8-)
Unfortunately the mailing list archives aren't searchable, and I don't
have the old emails anymore. If anyone can dig
faces they implement).
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The Middleware Company - We Build Experts!
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
invoked the action. Hence, one can now simply set "referer.action"
as result of actions who modify the state of a page and does not have
result data that should be shown (other than showing the initial page,
but with updated contents).
/Rickard
--
Rickard Öberg
Author of "Maste
etUrl(), execute().
Probably easiest.
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The Middleware Company - We Build Experts!
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
ge refreshes, it's not only refreshed
> itself, but it's still execing the URL also.
>
> Anyone know how to handle this?
Can you do a redirect to the view? I.e. set "redirect.action?url=..." as
view. That's what I do when I want to allow users to do refresh in
br
Taavi Tiirik wrote:
>>It is possible to do this without using the session, and also by only
>>using the standard redirect action as the result of your Login code
>>(which you can also do as an action).
>>
>>
>>I've done this in TSS and it works very we
y, then the following gets
> dumped.
Why do you think this has to do with the SD? The trace implies that it's
a JSP bean tag use that crashes. My guess would be that you forgot to
add '' around the param name, so it tries to resolve it as a property
name which yields n
t
> begin with
> '/' when forwarding to another page.
It is possible to do this without using the session, and also by only using the
standard redirect action as the result of your Login code (which you can also do as an
action).
I've done this in TSS and it works very
issue since a user can tweak the URI
to get to data which should usually not be possible. Not good.
/Rickard
--
Rickard Öberg
Chief Architect, TheServerSide.com
The Middleware Company
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
thin
> the Action API?
I do exactly this by using the standard Redirect action, in combination
with a servlet filter that does the login check and shows the login
form. Works very well.
/Rickard
--
Rickard Öberg
Chief Architect, TheServerS
Matt Baldree wrote:
> The latest code in CVS is up and running and all tests should work. Please
> feel free to check it out and test it. I believe the only thing left before
> 1.0 release is a couple of sections (HVAC and Action) in the docs and
> Rickard to push the button. If
actions, since that should not be needed
with modern VM's.
Can you create a minimal test-case? Can you avoid other possible causes (e.g.
incorrectly used JDBC connections etc.)?
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The
ve no use for this or
> find that reinventing the wheel is much better, that's ok :)
It sounds very interesting. Go ahead and add it to WW CVS under
webwork.formproc package.
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief
setA(b);
action1.execute();
Action2 action2 = (Action2)ActionFactory.getAction(Action2.class);
action2.setFoo(bar);
action2.execute();
...
That's the code manifestation of the above statement :-) You'll probably
want to do different things depending on the result of the execute()s
thou
to retirve
>>> parameters manually and calling doValidate from my doXxx command
>>> implementation.
>
> I have recently done integration with formproc. Amazingly it's just one
> class and requires no changes to WW. I'll commit tonight if you guys
> like t
ow have "isCommand()" it should be possible to do validation
selectively.
So, no, not that I can think of.
/Rickard
___
Webwork-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webwork-user
aving each action delegate explicitly
to the next action. I would prefer the second option if possible.
/Rickard
--
Rickard Öberg
Author of "Mastering RMI"
Chief Architect, TheServerSide.com
The Middleware Company - We Build Experts!
__
75 matches
Mail list logo
