How can I tell Struts 2 to throw/log exceptions for invalid OGNL Expressions

[mraible]

I have the following class:

public class TestAction {

    public static String concat(String str1, String str2) {
        return str1 + " and " + str2;
    }
}

In a JSP, I can write the following and it works fine:

<s:property value="@org.appfuse.web.TestAction@concat('1', '2')"/>

However, if I change it to an invalid class name, it fails silently:

<s:property value="@org.appfuse.web.InvalidClass@concat('1', '2')"/>

I've tried turning up the logging for org.apache.struts2, com.opensymphony.xwork2 and ognl - but it still fails silently. Any ideas how to get better error messages about invalid OGNL expressions?

Thanks,

Matt

[mraible]

mraible wrote:

I have the following class:

public class TestAction {

    public static String concat(String str1, String str2) {
        return str1 + " and " + str2;
    }
}

In a JSP, I can write the following and it works fine:

<s:property value="@org.appfuse.web.TestAction@concat('1', '2')"/>

However, if I change it to an invalid class name, it fails silently:

<s:property value="@org.appfuse.web.InvalidClass@concat('1', '2')"/>

I've tried turning up the logging for org.apache.struts2, com.opensymphony.xwork2 and ognl - but it still fails silently. Any ideas how to get better error messages about invalid OGNL expressions?

I'll take that as a "no", it is not possible to get better error messages. That's too bad since it seems I get better type-safety and errors from scriptlets instead of OGNL.

<%@ page import="org.foo.bar.baz.SettingsServlet,org.foo.bar.baz.enums.EnumCodec"  %>

<c:set var="q"><%=EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates)%></c:set>

<s:property value="url(@vs@SETTINGS_SERVLET).q('${q}')"/>

VS:

<s:property value="url(@vs@SETTINGS_SERVLET).q(@[e-mail](@org.foo.bar.baz.SettingsServlet$ACTIONS@networkUpdates))"/>

The framework I'm looking to migrate to Struts 2 has the ability for it's EL to read from imported Statics - a pretty cool feature IMO.

${url(pageBean.MAPPING.SETTINGS_SERVLET).q(EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates))}

Matt

[mraible]

Is it possible to use altSyntax to solve this problem? In other words, require %{} around all expressions? Looking through the docs, it doesn't seem like altSyntax is supported in Struts 2.

https://issues.apache.org/struts/browse/WW-1819

Thanks,

Matt

mraible wrote:

mraible wrote:

I have the following class:

public class TestAction {

    public static String concat(String str1, String str2) {
        return str1 + " and " + str2;
    }
}

In a JSP, I can write the following and it works fine:

<s:property value="@org.appfuse.web.TestAction@concat('1', '2')"/>

However, if I change it to an invalid class name, it fails silently:

<s:property value="@org.appfuse.web.InvalidClass@concat('1', '2')"/>

I've tried turning up the logging for org.apache.struts2, com.opensymphony.xwork2 and ognl - but it still fails silently. Any ideas how to get better error messages about invalid OGNL expressions?

I'll take that as a "no", it is not possible to get better error messages. That's too bad since it seems I get better type-safety and errors from scriptlets instead of OGNL.

<%@ page import="org.foo.bar.baz.SettingsServlet,org.foo.bar.baz.enums.EnumCodec"  %>

<c:set var="q"><%=EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates)%></c:set>

<s:property value="url(@vs@SETTINGS_SERVLET).q('${q}')"/>

VS:

<s:property value="url(@vs@SETTINGS_SERVLET).q(@[e-mail](@org.foo.bar.baz.SettingsServlet$ACTIONS@networkUpdates))"/>

The framework I'm looking to migrate to Struts 2 has the ability for it's EL to read from imported Statics - a pretty cool feature IMO.

${url(pageBean.MAPPING.SETTINGS_SERVLET).q(EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates))}

Matt

[mraible]

I'm now digging into XWork to see if I can add this functionality. I have the following in a JSP and I'd like to receive an error from it:

<s:property value="methodDoesntExist('foo')"/>

OgnlValueStack lines 260-265:

            Object value = OgnlUtil.getValue(expr, context, root, asType);
            if (value != null) {
                return value;
            } else {
                return findInContext(expr);
            }

In this example, the following method is called with "methodDoesntExist('foo')"" as the value for name:

    private Object findInContext(String name) {
        return getContext().get(name);
    }

So if nothing is found in the map, null is returned (I'm assuming). It seems like the if/else statement might be better with the following:

            Object value = OgnlUtil.getValue(expr, context, root, asType);
            if (value != null) {
                return value;
            } else {
                value = findInContext(expr);
                if (value == null) {
                    logLookupFailure(expr, ??);
                }
            }

However, I don't know what to put in the ?? location. Thoughts on this?

Thanks,

Matt

Is it possible to use altSyntax to solve this problem? In other words, require %{} around all expressions? Looking through the docs, it doesn't seem like altSyntax is supported in Struts 2.

https://issues.apache.org/struts/browse/WW-1819

Thanks,

Matt

mraible wrote:

mraible wrote:

I have the following class:

public class TestAction {

    public static String concat(String str1, String str2) {
        return str1 + " and " + str2;
    }
}

In a JSP, I can write the following and it works fine:

<s:property value="@org.appfuse.web.TestAction@concat('1', '2')"/>

However, if I change it to an invalid class name, it fails silently:

<s:property value="@org.appfuse.web.InvalidClass@concat('1', '2')"/>

I've tried turning up the logging for org.apache.struts2, com.opensymphony.xwork2 and ognl - but it still fails silently. Any ideas how to get better error messages about invalid OGNL expressions?

I'll take that as a "no", it is not possible to get better error messages. That's too bad since it seems I get better type-safety and errors from scriptlets instead of OGNL.

<%@ page import="org.foo.bar.baz.SettingsServlet,org.foo.bar.baz.enums.EnumCodec"  %>

<c:set var="q"><%=EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates)%></c:set>

<s:property value="url(@vs@SETTINGS_SERVLET).q('${q}')"/>

VS:

<s:property value="url(@vs@SETTINGS_SERVLET).q(@[e-mail](@org.foo.bar.baz.SettingsServlet$ACTIONS@networkUpdates))"/>

The framework I'm looking to migrate to Struts 2 has the ability for it's EL to read from imported Statics - a pretty cool feature IMO.

${url(pageBean.MAPPING.SETTINGS_SERVLET).q(EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates))}

Matt

[Ted Husted]

Have you tried doing this with devMode turned on?

<struts>
    <constant name="struts.devMode" value="true" />

For example, if I add to the form on MailReader Login.jsp

     <s:textfield name="doesntexist" />

and submit the form with devmode=true, the following is logged

Sep 6, 2007 11:14:52 AM
com.opensymphony.xwork2.validator.ActionValidatorManagerFactory
<clinit>
INFO: Detected AnnotationActionValidatorManager, initializing it...
Sep 6, 2007 11:15:01 AM
com.opensymphony.xwork2.interceptor.ParametersInterceptor
setParameters
SEVERE: ParametersInterceptor - [setParameters]: Unexpected Exception
caught setting 'button.save' on 'class mailreader2.Login: Error
setting expression 'button.save' with value
'[Ljava.lang.String;@e33e18'
Sep 6, 2007 11:15:01 AM
com.opensymphony.xwork2.interceptor.ParametersInterceptor
setParameters
SEVERE: ParametersInterceptor - [setParameters]: Unexpected Exception
caught setting 'doesntexist' on 'class mailreader2.Login: Error
setting expression 'doesntexist' with value
'[Ljava.lang.String;@15f1f9c'

HTH, Ted.
http://husted.com/ted/blog/


On 8/28/07, mraible <matt@...> wrote:

>
> I'm now digging into XWork to see if I can add this functionality. I have the
> following in a JSP and I'd like to receive an error from it:
>
> <s:property value="methodDoesntExist('foo')"/>
>
> OgnlValueStack lines 260-265:
>
>             Object value = OgnlUtil.getValue(expr, context, root, asType);
>             if (value != null) {
>                 return value;
>             } else {
>                 return findInContext(expr);
>             }
>
> In this example, the following method is called with
> "methodDoesntExist('foo')"" as the value for name:
>
>     private Object findInContext(String name) {
>         return getContext().get(name);
>     }
>
> So if nothing is found in the map, null is returned (I'm assuming). It seems
> like the if/else statement might be better with the following:
>
>             Object value = OgnlUtil.getValue(expr, context, root, asType);
>             if (value != null) {
>                 return value;
>             } else {
>                 value = findInContext(expr);
>                 if (value == null) {
>                     logLookupFailure(expr, ??);
>                 }
>             }
>
> However, I don't know what to put in the ?? location. Thoughts on this?
>
> Thanks,
>
> Matt
>
> mraible wrote:
> >
> > Is it possible to use altSyntax to solve this problem? In other words,
> > require %{} around all expressions? Looking through the docs, it doesn't
> > seem like altSyntax is supported in Struts 2.
> >
> > https://issues.apache.org/struts/browse/WW-1819
> >
> > Thanks,
> >
> > Matt
> >
> >
> > mraible wrote:
> >>
> >>
> >> mraible wrote:
> >>>
> >>> I have the following class:
> >>>
> >>> public class TestAction {
> >>>
> >>>     public static String concat(String str1, String str2) {
> >>>         return str1 + " and " + str2;
> >>>     }
> >>> }
> >>>
> >>> In a JSP, I can write the following and it works fine:
> >>>
> >>> <s:property value="@org.appfuse.web.TestAction@concat('1', '2')"/>
> >>>
> >>> However, if I change it to an invalid class name, it fails silently:
> >>>
> >>> <s:property value="@org.appfuse.web.InvalidClass@concat('1', '2')"/>
> >>>
> >>> I've tried turning up the logging for org.apache.struts2,
> >>> com.opensymphony.xwork2 and ognl - but it still fails silently. Any
> >>> ideas how to get better error messages about invalid OGNL expressions?
> >>>
> >>
> >> I'll take that as a "no", it is not possible to get better error
> >> messages. That's too bad since it seems I get better type-safety and
> >> errors from scriptlets instead of OGNL.
> >>
> >> <%@ page
> >> import="org.foo.bar.baz.SettingsServlet,org.foo.bar.baz.enums.EnumCodec"
> >> %>
> >>
> >> <c:set
> >> var="q"><%=EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates)%></c:set>
> >>
> >> <s:property value="url(@vs@SETTINGS_SERVLET).q('${q}')"/>
> >>
> >> VS:
> >>
> >> <s:property
> >> value="url(@vs@SETTINGS_SERVLET).q(@org.foo.bar.baz.enums.EnumCodec@...(@org.foo.bar.baz.SettingsServlet$ACTIONS@networkUpdates))"/>
> >>
> >> The framework I'm looking to migrate to Struts 2 has the ability for it's
> >> EL to read from imported Statics - a pretty cool feature IMO.
> >>
> >> ${url(pageBean.MAPPING.SETTINGS_SERVLET).q(EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates))}
> >>
> >> Matt
> >>
> >>
> >
> >
>
> --
> View this message in context: http://www.nabble.com/How-can-I-tell-Struts-2-to-throw-log-exceptions-for-invalid-OGNL-Expressions-tf4100102.html#a12373437
> Sent from the Struts - User mailing list archive at Nabble.com.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@...
> For additional commands, e-mail: user-help@...
>
>

--
HTH, Ted <http://www.husted.com/ted/blog/>

---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@...
For additional commands, e-mail: user-help@...

[mraible]

struts.devMode=true will catch setting invalid properties, but not reading from them. I want to be warned when a property/method doesn't exist. Or in the instance of calling stating methods with FQCN, I want to be made aware when I fat finger the package name. If I could use JSP's import and then read from the package, that'd be even nicer. It's somewhat verbose to call FQ static methods. I suppose I could move the methods into my action, so it's not too bad.

Matt

Ted Husted wrote:

Have you tried doing this with devMode turned on?

<struts>
    <constant name="struts.devMode" value="true" />

For example, if I add to the form on MailReader Login.jsp

     <s:textfield name="doesntexist" />

and submit the form with devmode=true, the following is logged

Sep 6, 2007 11:14:52 AM
com.opensymphony.xwork2.validator.ActionValidatorManagerFactory
<clinit>
INFO: Detected AnnotationActionValidatorManager, initializing it...
Sep 6, 2007 11:15:01 AM
com.opensymphony.xwork2.interceptor.ParametersInterceptor
setParameters
SEVERE: ParametersInterceptor - [setParameters]: Unexpected Exception
caught setting 'button.save' on 'class mailreader2.Login: Error
setting expression 'button.save' with value
'[Ljava.lang.String;@e33e18'
Sep 6, 2007 11:15:01 AM
com.opensymphony.xwork2.interceptor.ParametersInterceptor
setParameters
SEVERE: ParametersInterceptor - [setParameters]: Unexpected Exception
caught setting 'doesntexist' on 'class mailreader2.Login: Error
setting expression 'doesntexist' with value
'[Ljava.lang.String;@15f1f9c'

HTH, Ted.
http://husted.com/ted/blog/


On 8/28/07, mraible <[e-mail]> wrote:
>
> I'm now digging into XWork to see if I can add this functionality. I have the
> following in a JSP and I'd like to receive an error from it:
>
> <s:property value="methodDoesntExist('foo')"/>
>
> OgnlValueStack lines 260-265:
>
>             Object value = OgnlUtil.getValue(expr, context, root, asType);
>             if (value != null) {
>                 return value;
>             } else {
>                 return findInContext(expr);
>             }
>
> In this example, the following method is called with
> "methodDoesntExist('foo')"" as the value for name:
>
>     private Object findInContext(String name) {
>         return getContext().get(name);
>     }
>
> So if nothing is found in the map, null is returned (I'm assuming). It seems
> like the if/else statement might be better with the following:
>
>             Object value = OgnlUtil.getValue(expr, context, root, asType);
>             if (value != null) {
>                 return value;
>             } else {
>                 value = findInContext(expr);
>                 if (value == null) {
>                     logLookupFailure(expr, ??);
>                 }
>             }
>
> However, I don't know what to put in the ?? location. Thoughts on this?
>
> Thanks,
>
> Matt
>
> mraible wrote:
> >
> > Is it possible to use altSyntax to solve this problem? In other words,
> > require %{} around all expressions? Looking through the docs, it doesn't
> > seem like altSyntax is supported in Struts 2.
> >
> > https://issues.apache.org/struts/browse/WW-1819
> >
> > Thanks,
> >
> > Matt
> >
> >
> > mraible wrote:
> >>
> >>
> >> mraible wrote:
> >>>
> >>> I have the following class:
> >>>
> >>> public class TestAction {
> >>>
> >>>     public static String concat(String str1, String str2) {
> >>>         return str1 + " and " + str2;
> >>>     }
> >>> }
> >>>
> >>> In a JSP, I can write the following and it works fine:
> >>>
> >>> <s:property value="@org.appfuse.web.TestAction@concat('1', '2')"/>
> >>>
> >>> However, if I change it to an invalid class name, it fails silently:
> >>>
> >>> <s:property value="@org.appfuse.web.InvalidClass@concat('1', '2')"/>
> >>>
> >>> I've tried turning up the logging for org.apache.struts2,
> >>> com.opensymphony.xwork2 and ognl - but it still fails silently. Any
> >>> ideas how to get better error messages about invalid OGNL expressions?
> >>>
> >>
> >> I'll take that as a "no", it is not possible to get better error
> >> messages. That's too bad since it seems I get better type-safety and
> >> errors from scriptlets instead of OGNL.
> >>
> >> <%@ page
> >> import="org.foo.bar.baz.SettingsServlet,org.foo.bar.baz.enums.EnumCodec"
> >> %>
> >>
> >> <c:set
> >> var="q"><%=EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates)%></c:set>
> >>
> >> <s:property value="url(@vs@SETTINGS_SERVLET).q('${q}')"/>
> >>
> >> VS:
> >>
> >> <s:property
> >> value="url(@vs@SETTINGS_SERVLET).q(@[e-mail](@org.foo.bar.baz.SettingsServlet$ACTIONS@networkUpdates))"/>
> >>
> >> The framework I'm looking to migrate to Struts 2 has the ability for it's
> >> EL to read from imported Statics - a pretty cool feature IMO.
> >>
> >> ${url(pageBean.MAPPING.SETTINGS_SERVLET).q(EnumCodec.URL.encode(SettingsServlet.ACTIONS.networkUpdates))}
> >>
> >> Matt
> >>
> >>
> >
> >
>
> --
> View this message in context: http://www.nabble.com/How-can-I-tell-Struts-2-to-throw-log-exceptions-for-invalid-OGNL-Expressions-tf4100102.html#a12373437
> Sent from the Struts - User mailing list archive at Nabble.com.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [e-mail]
> For additional commands, e-mail: [e-mail]
>
>


--
HTH, Ted <http://www.husted.com/ted/blog/>

---------------------------------------------------------------------
To unsubscribe, e-mail: [e-mail]
For additional commands, e-mail: [e-mail]

[Ted Husted]

Good point. Feel free to open a S2 JIRA ticket that cross-references
XW-557 (since I'm not sure if its something we could do here or not)
-- and feel even more free to submit a patch!

As usual, we are short of active volunteers, and we can always use a
hand with the code!

-Ted.

On 9/6/07, mraible <matt@...> wrote:

>
> struts.devMode=true will catch setting invalid properties, but not reading
> from them. I want to be warned when a property/method doesn't exist. Or in
> the instance of calling stating methods with FQCN, I want to be made aware
> when I fat finger the package name. If I could use JSP's import and then
> read from the package, that'd be even nicer. It's somewhat verbose to call
> FQ static methods. I suppose I could move the methods into my action, so
> it's not too bad.
>
> Matt

---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@...
For additional commands, e-mail: user-help@...

[mraible]

Following up months later as this has come up again now that 2.1 is close to release. The current version does log messages for invalid properties, but it doesn't blow up for invalid properties which is more of what I'm looking for.

When I blogged about this[1], one of the responses was the following:

When you are using freemarker as your result technology in struts2 you get great feedback on problems like this. This is not indicative of the whole framework -- just the result-type that you are using (probably jsp?)

I tried FreeMarker this morning and discovered that the problem exists there too. Grrr. I've experienced Struts 2 being eliminated as a web framework candidate on a couple projects because of this. Kindof annoying.

Matt

[1] http://raibledesigns.com/rd/entry/does_struts_2_suck

Ted Husted wrote:

Good point. Feel free to open a S2 JIRA ticket that cross-references
XW-557 (since I'm not sure if its something we could do here or not)
-- and feel even more free to submit a patch!

As usual, we are short of active volunteers, and we can always use a
hand with the code!

-Ted.

On 9/6/07, mraible <[e-mail]> wrote:
>
> struts.devMode=true will catch setting invalid properties, but not reading
> from them. I want to be warned when a property/method doesn't exist. Or in
> the instance of calling stating methods with FQCN, I want to be made aware
> when I fat finger the package name. If I could use JSP's import and then
> read from the package, that'd be even nicer. It's somewhat verbose to call
> FQ static methods. I suppose I could move the methods into my action, so
> it's not too bad.
>
> Matt

---------------------------------------------------------------------
To unsubscribe, e-mail: [e-mail]
For additional commands, e-mail: [e-mail]

[Jeromy Evans - Blue Sky Minds]

mraible wrote:
> Following up months later as this has come up again now that 2.1 is close to
> release. The current version does log messages for invalid properties, but
> it doesn't blow up for invalid properties which is more of what I'm looking
> for.
>
>  
Hi Matt,
I'm not convinced that you really want what you say you want.

Sure, if you use:
<s:property name="prop"/>
You want to see a message in the log stating that the property didn't
exist.  Fair enough.

However, if you use:
<s:property name="prop" default="a"/>
Then I presume you don't want a message in the log if property didn't
exist.  But OGNL doesn't know about the default attribute.

Similarly in FTL if you used:
${parameters.nameValue}?default("a")
You wouldn't want a warning from the framework if nameValue doesn't
exist because freemarker will handle that if you don't provide a default.

Quite clearly, when setting a property it's the framework's
responsibility to report if the property didn't exist.
But whose responsibility is it to report that an attempt to read a
property was made but the property didn't exist. I'm sure many
algorithms depending on silent failure so they can provide a default.  
My opinion is that it's best handled in the view and FTL has the best
implementation of this (particularly in 2.4.x where there's improved
handling for nulls and blanks).

> When I blogged about this[1], one of the responses was the following:
>
>
>
>  
>> When you are using freemarker as your result technology in struts2 you get
>> great feedback on problems like this. This is not indicative of the whole
>> framework -- just the result-type that you are using (probably jsp?)
>>
>>    
>
> I tried FreeMarker this morning and discovered that the problem exists there
> too. Grrr. I've experienced Struts 2 being eliminated as a web framework
> candidate on a couple projects because of this. Kindof annoying.
>
>  

So what do you really want?  My guess here is that you invoked a struts
tag and the tag swallowed the error and provided a default to the
freemarker template.


---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@...
For additional commands, e-mail: user-help@...

[Jeromy Evans - Blue Sky Minds]

Jeromy Evans wrote:

> mraible wrote:
>
>>
>> I tried FreeMarker this morning and discovered that the problem
>> exists there
>> too. Grrr. I've experienced Struts 2 being eliminated as a web framework
>> candidate on a couple projects because of this. Kindof annoying.
>>  
>
> So what do you really want?  My guess here is that you invoked a
> struts tag and the tag swallowed the error and provided a default to
> the freemarker template.
>
>

On further thought I think this comes down to an interesting issue with
coupling between the view and the model.
In struts2 the model is a ValueStack that consists of both a stack and
map (context).  OGNL is used to address properties or methods in that
model and it's a trivial for OGNL to log when it fails to match a
property or method.

In the view, Struts2 provides tags for jsp, ftl or velocity that
evaluate OGNL expressions to access the model.  More often that not the
tags fail silently if the expression was invalid or failed to evaluate
anything.  Herein lies the problem.

If I were using FTL directly to access the value stack, as a developer I
could decide how to handle missing properties, blank properties or
default values with FTL's excellent error handling.  However, when I use
FTL to invoke a Struts2 tag I lose that benefit as the tag
implementations *transfer* properties into the FTL context. I think its
a flaw that Struts2 tags evaluate OGNL expressions themselves prior to
rendering the template..

A tag overhaul is well overdue (but major effort).  I've expressed other
concerns about the Struts2 tags (XWork Component and UIBean) on several
occasions already.



---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@...
For additional commands, e-mail: user-help@...

[mraible]

Jeromy Evans - Blue Sky Minds wrote:

Jeromy Evans wrote:
> mraible wrote:
>
>>
>> I tried FreeMarker this morning and discovered that the problem
>> exists there
>> too. Grrr. I've experienced Struts 2 being eliminated as a web framework
>> candidate on a couple projects because of this. Kindof annoying.
>>  
>
> So what do you really want?  My guess here is that you invoked a
> struts tag and the tag swallowed the error and provided a default to
> the freemarker template.
>
>

On further thought I think this comes down to an interesting issue with
coupling between the view and the model.
In struts2 the model is a ValueStack that consists of both a stack and
map (context).  OGNL is used to address properties or methods in that
model and it's a trivial for OGNL to log when it fails to match a
property or method.

In the view, Struts2 provides tags for jsp, ftl or velocity that
evaluate OGNL expressions to access the model.  More often that not the
tags fail silently if the expression was invalid or failed to evaluate
anything.  Herein lies the problem.

If I were using FTL directly to access the value stack, as a developer I
could decide how to handle missing properties, blank properties or
default values with FTL's excellent error handling.  However, when I use
FTL to invoke a Struts2 tag I lose that benefit as the tag
implementations *transfer* properties into the FTL context. I think its
a flaw that Struts2 tags evaluate OGNL expressions themselves prior to
rendering the template..

A tag overhaul is well overdue (but major effort).  I've expressed other
concerns about the Struts2 tags (XWork Component and UIBean) on several
occasions already.

I realize it might be a major effort, but this seems to be a pretty big deal to me. Not for users that've already chosen Struts 2 as a web framework, but for folks evaluating Java Web Frameworks. The fact that Struts 2 is the *only* one that doesn't blow up on invalid properties makes it look pretty bad for evaluations. It's kindof like the old blank page you'd sometimes get with Tiles - you really have no idea what you did wrong and it can take quite some time to track it down if you don't know what's causing it.

Matt

[Jeromy Evans - Blue Sky Minds]

mraible wrote:

> I realize it might be a major effort, but this seems to be a pretty big deal
> to me. Not for users that've already chosen Struts 2 as a web framework, but
> for folks evaluating Java Web Frameworks. The fact that Struts 2 is the
> *only* one that doesn't blow up on invalid properties makes it look pretty
> bad for evaluations. It's kindof like the old blank page you'd sometimes get
> with Tiles - you really have no idea what you did wrong and it can take
> quite some time to track it down if you don't know what's causing it.
>
> Matt
>  

Strongly agreed.  This issue does have a productivity cost.  I'll try to
do more about it myself.
I've committed an improvement for message resources
(https://issues.apache.org/struts/browse/WW-2592).



---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@...
For additional commands, e-mail: user-help@...