Re: Database Design - Multiple locations

2008-04-22 Thread Mr. Frank Cobia

Thanks. I will look into this and see if I can make it work.

Frank

On Apr 21, 2008, at 3:45 PM, Chuck Hill wrote:


On Apr 20, 2008, at 6:15 PM, Mr. Frank Cobia wrote:

I have a database design question. I am posting it to the  
WebObjects list because I am building a WebObjects app and want a  
solution that works well for EOF.


I have a table of data that has to have a location associated with  
it. Unfortunately the level of the location is not constant. The  
data can be associated with a Country, Region, State, Metro Area,  
County or City. Each of those levels is its own table with  
relationships pointing to the related locations. i.e. a State knows  
which Region it belongs to and which Metro Areas belong to it.


I have been unable to come up with a design that seems elegant. I  
have thought of having 6 separate relationships to each of the  
location levels, but it seems to duplicate data and the data could  
get out of sync if a state is moved to a different region or a City  
is moved to a different county.


Has anyone had a situation like this and come up with a good  
solution?



This seems like a good place for inheritance with an abstract EO of  
Location with concrete subclasses of Country, Region, State, Metro  
Area, County or City.  For some operations, having them in separate  
tables might cause more fetches than you want.  Using single table  
inheritance would address that.


Chuck

--

Practical WebObjects - for developers who want to increase their  
overall knowledge of WebObjects or who are trying to solve specific  
problems.

http://www.global-village.net/products/practical_webobjects








___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


CLOSE_WAIT processes/ instances HANG ???

2008-04-22 Thread Shravan Kumar.M

Hello Group,

We have been battered by CLOSE_WAIT processes and instances hang.

We are running our application on Solaris 5.8 boxes with WebObjects 
5.2.4, Java 1.4.2_17.


We have collected some process dumps and consulted Sun, they analyzed 
and blame our code, alas!!!


We see CLOSE_WAIT state of a process for both java and httpd processes.

We see degradation in our application performance as the day progresses 
and our app becomes unresponsive various times, we restart the app 
various times during day.


Could any one please advise us a solution to find the reason for 
instance hang or fix CLOSE_WAIT processes generation.


Let me know if you need any other information.

We appreciate a lot your valuable responses. Please respond asap.

--
Thank You
Shravan Kumar. M
effigent India Pvt., Ltd.,
--

 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]

Re: Database Design - Multiple locations

2008-04-22 Thread Mr. Frank Cobia
I had thought of something along theses lines, but I was worried about  
searches. Most of the time I will not be accessing the locations  
through the tree, but doing a search such as where location = Georgia  
which will also need to pull up anything with a location in a country,  
metro region or city that is inside Georgia. It seems like this type  
of structure does not lend itself well to searches.


One thing I had thought of doing is having a sort of code where the  
code is made up of the code of its parent. For example the City of  
atlanta might have a code of us.georgia.metroatlanta.fulton This way  
if I want to find an object with location of Metro Atlanta I could do:  
where location.code = 'us.georgia.metroatlanta'


What I was worried about is there are times when the tree structure  
changes (usually because of human error classifying the county or  
city) then the code might get out of sync. Is that fear irrational?


Thanks,
Frank


On Apr 22, 2008, at 9:42 AM, Robert Walker wrote:

Optionally, you could use a reflexive relationship for this. I've  
used them in the past for this type of situation and it seemed to  
work out well. This would allow you to have an arbitrarily deep tree  
structure using a single database table and single EO.


I can't think of any particular disadvantages to this type of  
relationship, and the structure is very simple. You create a one-to- 
many relationship that references a different row in the same table.  
The only caveat I'm aware of is to watch out for recursive  
associations.


I've built similar database structures for bills of material  
(although they are more complex than this, and involve a many-to- 
many reflexive relationship). Basically you just have to make sure  
that any root node is never used anywhere within that particular  
branch of the tree. This can usually be accomplished though the UI,  
but it's best to recursively validate the branch of the tree on save.


-
Location
-
id
name
type (Country, Region, State, Metro Area, County, or City)
parentID
-
parent (to-one)
children (to-many)
-

United States (Country)
-- South East (Region)
 Alabama (State)
 Arkansas (State)
 Georgia (State)
-- Atlanta (Metro Area)
 Fulton (County)
-- Atlanta (City)
 
-- North West (Region)
 Washington (State)
 Oregon (State)
 Montana (State)

Watch out for infinite recursion problems as illustrated below:

United States (Country)
-- South East (Region)
 Alabama (State)
-- United States (Country) --- Causes infinite recursion
 South East (Region)
-- Alabama (State)
 United States (Country) -- As is obvious here  
since you never find leaf nodes


On Apr 22, 2008, at 8:33 AM, Mr. Frank Cobia wrote:


Thanks. I will look into this and see if I can make it work.

Frank

On Apr 21, 2008, at 3:45 PM, Chuck Hill wrote:


On Apr 20, 2008, at 6:15 PM, Mr. Frank Cobia wrote:

I have a database design question. I am posting it to the  
WebObjects list because I am building a WebObjects app and want a  
solution that works well for EOF.


I have a table of data that has to have a location associated  
with it. Unfortunately the level of the location is not constant.  
The data can be associated with a Country, Region, State, Metro  
Area, County or City. Each of those levels is its own table with  
relationships pointing to the related locations. i.e. a State  
knows which Region it belongs to and which Metro Areas belong to  
it.


I have been unable to come up with a design that seems elegant. I  
have thought of having 6 separate relationships to each of the  
location levels, but it seems to duplicate data and the data  
could get out of sync if a state is moved to a different region  
or a City is moved to a different county.


Has anyone had a situation like this and come up with a good  
solution?



This seems like a good place for inheritance with an abstract EO  
of Location with concrete subclasses of Country, Region, State,  
Metro Area, County or City.  For some operations, having them in  
separate tables might cause more fetches than you want.  Using  
single table inheritance would address that.


Chuck

--

Practical WebObjects - for developers who want to increase their  
overall knowledge of WebObjects or who are trying to solve  
specific problems.

http://www.global-village.net/products/practical_webobjects








___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:

Re: Database Design - Multiple locations

2008-04-22 Thread Mr. Frank Cobia

Correction, the search below with the code would be

where location.code like us.georgia.metroatlanta%

On Apr 22, 2008, at 10:35 AM, Mr. Frank Cobia wrote:

I had thought of something along theses lines, but I was worried  
about searches. Most of the time I will not be accessing the  
locations through the tree, but doing a search such as where  
location = Georgia which will also need to pull up anything with a  
location in a country, metro region or city that is inside Georgia.  
It seems like this type of structure does not lend itself well to  
searches.


One thing I had thought of doing is having a sort of code where the  
code is made up of the code of its parent. For example the City of  
atlanta might have a code of us.georgia.metroatlanta.fulton This  
way if I want to find an object with location of Metro Atlanta I  
could do: where location.code = 'us.georgia.metroatlanta'


What I was worried about is there are times when the tree structure  
changes (usually because of human error classifying the county or  
city) then the code might get out of sync. Is that fear irrational?


Thanks,
Frank


On Apr 22, 2008, at 9:42 AM, Robert Walker wrote:

Optionally, you could use a reflexive relationship for this. I've  
used them in the past for this type of situation and it seemed to  
work out well. This would allow you to have an arbitrarily deep  
tree structure using a single database table and single EO.


I can't think of any particular disadvantages to this type of  
relationship, and the structure is very simple. You create a one-to- 
many relationship that references a different row in the same  
table. The only caveat I'm aware of is to watch out for recursive  
associations.


I've built similar database structures for bills of material  
(although they are more complex than this, and involve a many-to- 
many reflexive relationship). Basically you just have to make sure  
that any root node is never used anywhere within that particular  
branch of the tree. This can usually be accomplished though the UI,  
but it's best to recursively validate the branch of the tree on save.


-
Location
-
id
name
type (Country, Region, State, Metro Area, County, or City)
parentID
-
parent (to-one)
children (to-many)
-

United States (Country)
-- South East (Region)
 Alabama (State)
 Arkansas (State)
 Georgia (State)
-- Atlanta (Metro Area)
 Fulton (County)
-- Atlanta (City)
 
-- North West (Region)
 Washington (State)
 Oregon (State)
 Montana (State)

Watch out for infinite recursion problems as illustrated below:

United States (Country)
-- South East (Region)
 Alabama (State)
-- United States (Country) --- Causes infinite recursion
 South East (Region)
-- Alabama (State)
 United States (Country) -- As is obvious here  
since you never find leaf nodes


On Apr 22, 2008, at 8:33 AM, Mr. Frank Cobia wrote:


Thanks. I will look into this and see if I can make it work.

Frank

On Apr 21, 2008, at 3:45 PM, Chuck Hill wrote:


On Apr 20, 2008, at 6:15 PM, Mr. Frank Cobia wrote:

I have a database design question. I am posting it to the  
WebObjects list because I am building a WebObjects app and want  
a solution that works well for EOF.


I have a table of data that has to have a location associated  
with it. Unfortunately the level of the location is not  
constant. The data can be associated with a Country, Region,  
State, Metro Area, County or City. Each of those levels is its  
own table with relationships pointing to the related locations.  
i.e. a State knows which Region it belongs to and which Metro  
Areas belong to it.


I have been unable to come up with a design that seems elegant.  
I have thought of having 6 separate relationships to each of the  
location levels, but it seems to duplicate data and the data  
could get out of sync if a state is moved to a different region  
or a City is moved to a different county.


Has anyone had a situation like this and come up with a good  
solution?



This seems like a good place for inheritance with an abstract EO  
of Location with concrete subclasses of Country, Region, State,  
Metro Area, County or City.  For some operations, having them in  
separate tables might cause more fetches than you want.  Using  
single table inheritance would address that.


Chuck

--

Practical WebObjects - for developers who want to increase their  
overall knowledge of WebObjects or who are trying to solve  
specific problems.

http://www.global-village.net/products/practical_webobjects








___
Do not post 

WOWebServiceRegistrar.setSecurityDelegate failure

2008-04-22 Thread Jon Nolan

I'm in the process of upgrading all my WebObjects projects from 
Xcode/Java1.4/WebObjects5.2.3 to Eclipse/WOLips/Java1.5/WebObjects5.4.1.  I've 
possibly bitten off way too big a bite all at once.  ;-)

I'm muddling through without too much trouble except when it comes to 
webservices.  I make substantive use of WOSecurityDelegate.  It is absolutely 
necessary in a number of places.  I just can't seem to get a delegate 
registered though.  Here's the code:

org.apache.xml.security.Init.init();
WOWebServiceRegistrar.setSecurityDelegate(new 
MyWOSecurityDelegateClass());

The registration fails with:

java.lang.NullPointerException
at 
com.webobjects.appserver._private.WOServerSessionHandler.clinit(WOServerSessionHandler.java:25)
at java.lang.Class.forName0(Native Method)
at java.lang.Class.forName(Class.java:242)
at org.apache.axis.utils.ClassUtils$2.run(ClassUtils.java:177)
at java.security.AccessController.doPrivileged(Native Method)
at org.apache.axis.utils.ClassUtils.loadClass(ClassUtils.java:160)
at org.apache.axis.utils.ClassUtils.forName(ClassUtils.java:100)
at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getJavaClass(WSDDDeployableItem.java:353)
at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.makeNewInstance(WSDDDeployableItem.java:295)
at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getNewInstance(WSDDDeployableItem.java:274)
at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getInstance(WSDDDeployableItem.java:260)
at 
org.apache.axis.deployment.wsdd.WSDDChain.makeNewInstance(WSDDChain.java:125)
at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getNewInstance(WSDDDeployableItem.java:274)
at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getInstance(WSDDDeployableItem.java:260)
at 
org.apache.axis.deployment.wsdd.WSDDDeployment.getGlobalRequest(WSDDDeployment.java:473)
at 
com.webobjects.webservices.support.WOXMLProvider.getGlobalRequest(WOXMLProvider.java:168)
at org.apache.axis.AxisEngine.getGlobalRequest(AxisEngine.java:365)
at 
com.webobjects.appserver._private.WOWebService.setSecurityDelegate(WOWebService.java:205)
at 
com.webobjects.appserver.WOWebServiceRegistrar.setSecurityDelegate(WOWebServiceRegistrar.java:358)

My configuration is:

Eclipse 3.3.2 
WOLips 3.3.5047

OS X 10.5.2
Java 1.5.0_13

I have standard Apple distribution jars for:

axis.jar (1.4 I think)
commons-discovery-0.2.jar
commons-logging-1.0.4.jar
jaxrpc.jar
saaj.jar
wsdl4j-1.5.1.jar
xmlsec-1.3.0.jar

What else?  Yes, the WebObjects webservices frameworks are installed.  Yes, the 
error is the same running within Eclipse or at the command line.

Any help is very appreciated.

--
Jon Nolan

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: ERXMigrationTable.newColumn does not set default value

2008-04-22 Thread David Avendasora

Hey Mike,

I'm trying to add this to the ERXMicrosoftPlugin. I've added a  
MicrosoftExpression Inner Class, and copied over much of what was in  
the FrontbaseExpression IC with some modifications, but it looks like  
where it's looking for the default value is from the UserInfo  
dictionary in the model as opposed to the default value from the  
Migration class.


Is the userInfo dictionary the place to set defaults, and if so, what  
does the default do in the Migration class?


Thanks!

Dave

On Apr 10, 2008, at 8:50 AM, Mike Schrag wrote:

I lied anyway ... I DID actually add it to PGSQL, but I was looking  
in the same wrong place that you were.  FB PlugIn has two inner  
classes (the schemasync and the expression).  That method is  
actually in expression, not in sync, so when I looked in PG, I  
didn't see it.  However, it IS in PGExpession.  So I suspect your  
problems with defaults not working are due to 5.4.  I'll have to  
check this for the next WO version, but I SUSPECT it works  
properly ...


ms

On Apr 10, 2008, at 8:38 AM, Mike Schrag wrote:

The statment is built by a call of  
EOSchemaSynchronizationFactory._columnCreationClauseForAttribute  
method which calls - in my case
I take it (based on the class name you're referencing here) that  
you're using WO 5.4.x?  The problems I've discussed previously with  
respect to SQL generation in 5.4.x make this not worth trying to  
fix on your 5.4.x build until the next WO release ...


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/mschrag%40mdimension.com

This email sent to [EMAIL PROTECTED]



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/webobjects%40avendasora.com

This email sent to [EMAIL PROTECTED]




___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: Database Design - Multiple locations

2008-04-22 Thread James Cicenia

About that irrational fear...

Sometimes code is your best friend.

I like your URL approach.

-j-

On Apr 22, 2008, at 9:35 AM, Mr. Frank Cobia wrote:

I had thought of something along theses lines, but I was worried  
about searches. Most of the time I will not be accessing the  
locations through the tree, but doing a search such as where  
location = Georgia which will also need to pull up anything with a  
location in a country, metro region or city that is inside Georgia.  
It seems like this type of structure does not lend itself well to  
searches.


One thing I had thought of doing is having a sort of code where the  
code is made up of the code of its parent. For example the City of  
atlanta might have a code of us.georgia.metroatlanta.fulton This  
way if I want to find an object with location of Metro Atlanta I  
could do: where location.code = 'us.georgia.metroatlanta'


What I was worried about is there are times when the tree structure  
changes (usually because of human error classifying the county or  
city) then the code might get out of sync. Is that fear irrational?


Thanks,
Frank


On Apr 22, 2008, at 9:42 AM, Robert Walker wrote:

Optionally, you could use a reflexive relationship for this. I've  
used them in the past for this type of situation and it seemed to  
work out well. This would allow you to have an arbitrarily deep  
tree structure using a single database table and single EO.


I can't think of any particular disadvantages to this type of  
relationship, and the structure is very simple. You create a one-to- 
many relationship that references a different row in the same  
table. The only caveat I'm aware of is to watch out for recursive  
associations.


I've built similar database structures for bills of material  
(although they are more complex than this, and involve a many-to- 
many reflexive relationship). Basically you just have to make sure  
that any root node is never used anywhere within that particular  
branch of the tree. This can usually be accomplished though the UI,  
but it's best to recursively validate the branch of the tree on save.


-
Location
-
id
name
type (Country, Region, State, Metro Area, County, or City)
parentID
-
parent (to-one)
children (to-many)
-

United States (Country)
-- South East (Region)
 Alabama (State)
 Arkansas (State)
 Georgia (State)
-- Atlanta (Metro Area)
 Fulton (County)
-- Atlanta (City)
 
-- North West (Region)
 Washington (State)
 Oregon (State)
 Montana (State)

Watch out for infinite recursion problems as illustrated below:

United States (Country)
-- South East (Region)
 Alabama (State)
-- United States (Country) --- Causes infinite recursion
 South East (Region)
-- Alabama (State)
 United States (Country) -- As is obvious here  
since you never find leaf nodes


On Apr 22, 2008, at 8:33 AM, Mr. Frank Cobia wrote:


Thanks. I will look into this and see if I can make it work.

Frank

On Apr 21, 2008, at 3:45 PM, Chuck Hill wrote:


On Apr 20, 2008, at 6:15 PM, Mr. Frank Cobia wrote:

I have a database design question. I am posting it to the  
WebObjects list because I am building a WebObjects app and want  
a solution that works well for EOF.


I have a table of data that has to have a location associated  
with it. Unfortunately the level of the location is not  
constant. The data can be associated with a Country, Region,  
State, Metro Area, County or City. Each of those levels is its  
own table with relationships pointing to the related locations.  
i.e. a State knows which Region it belongs to and which Metro  
Areas belong to it.


I have been unable to come up with a design that seems elegant.  
I have thought of having 6 separate relationships to each of the  
location levels, but it seems to duplicate data and the data  
could get out of sync if a state is moved to a different region  
or a City is moved to a different county.


Has anyone had a situation like this and come up with a good  
solution?



This seems like a good place for inheritance with an abstract EO  
of Location with concrete subclasses of Country, Region, State,  
Metro Area, County or City.  For some operations, having them in  
separate tables might cause more fetches than you want.  Using  
single table inheritance would address that.


Chuck

--

Practical WebObjects - for developers who want to increase their  
overall knowledge of WebObjects or who are trying to solve  
specific problems.

http://www.global-village.net/products/practical_webobjects








___
Do not post admin 

Fake Entity Modeler validation errors preventing veogen running

2008-04-22 Thread Simon McLean

Hi All -

I've updated to WOLips 3.3.5047 but now can't veogenerate because of  
model validation errors relating to key paths in model based fetch  
specifications:


/Users/Eclipse2/ClickTravelSystem/Resources/TravelSystem.eogen:
	Error: The fetch specification TrainLegSearch has a qualifier that  
refers to an invalid key 'segment.voucher.ticketQueue'.
	Error: The fetch specification TrainLegSearch has a qualifier that  
refers to an invalid key 'flightNumber'.


However, the key paths refer to attributes contained within subclasses  
so are actually valid.


We had the same problem with Component Editor validation but can work  
around it by marking them with


//VALID

Is there something similar in Entity Modeler ?

Thanks, Simon



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: Inheritance and overriding

2008-04-22 Thread Florijan Stamenkovic


On Apr 21, 2008, at 17:14, Chuck Hill wrote:

I have never used flattened relationships.  For performance, you  
can cache in the EOs as long as you are careful to clear or rebuild  
the cache when it is no longer valid.


I have an idea using JBND (which takes care of listening along  
keypaths, when used to bind EO values) for this. That way the  
retrieved values from the keypaths will be cached by GUI components,  
until either there is change detected along the keypath tree, or I do  
something major like saving changes to the server...


Still, it will be tricky, to say the least...

Thanks,
Flor
___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: WOWebServiceRegistrar.setSecurityDelegate failure

2008-04-22 Thread Mike Schrag
There are known issues with WebServices on 5.4.1.  I don't know if  
this one in particular is fixed in 5.4.2 or not, but you should  
definitely get a baseline on 5.3 FIRST before doing the full move to  
5.4.1.


ms

On Apr 22, 2008, at 12:09 PM, Jon Nolan wrote:

I'm in the process of upgrading all my WebObjects projects from  
Xcode/Java1.4/WebObjects5.2.3 to Eclipse/WOLips/Java1.5/ 
WebObjects5.4.1.  I've possibly bitten off way too big a bite all at  
once.  ;-)


I'm muddling through without too much trouble except when it comes  
to webservices.  I make substantive use of WOSecurityDelegate.  It  
is absolutely necessary in a number of places.  I just can't seem to  
get a delegate registered though.  Here's the code:


org.apache.xml.security.Init.init();
	WOWebServiceRegistrar.setSecurityDelegate(new  
MyWOSecurityDelegateClass());


The registration fails with:

java.lang.NullPointerException
	at  
com 
.webobjects 
.appserver 
._private 
.WOServerSessionHandler.clinit(WOServerSessionHandler.java:25)

at java.lang.Class.forName0(Native Method)
at java.lang.Class.forName(Class.java:242)
at org.apache.axis.utils.ClassUtils$2.run(ClassUtils.java:177)
at java.security.AccessController.doPrivileged(Native Method)
at org.apache.axis.utils.ClassUtils.loadClass(ClassUtils.java:160)
at org.apache.axis.utils.ClassUtils.forName(ClassUtils.java:100)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getJavaClass(WSDDDeployableItem.java:353)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.makeNewInstance(WSDDDeployableItem.java:295)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getNewInstance(WSDDDeployableItem.java:274)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getInstance(WSDDDeployableItem.java:260)
	at  
org 
.apache 
.axis.deployment.wsdd.WSDDChain.makeNewInstance(WSDDChain.java:125)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getNewInstance(WSDDDeployableItem.java:274)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getInstance(WSDDDeployableItem.java:260)
	at  
org 
.apache 
.axis 
.deployment.wsdd.WSDDDeployment.getGlobalRequest(WSDDDeployment.java: 
473)
	at  
com 
.webobjects 
.webservices 
.support.WOXMLProvider.getGlobalRequest(WOXMLProvider.java:168)

at org.apache.axis.AxisEngine.getGlobalRequest(AxisEngine.java:365)
	at  
com 
.webobjects 
.appserver 
._private.WOWebService.setSecurityDelegate(WOWebService.java:205)
	at  
com 
.webobjects 
.appserver 
.WOWebServiceRegistrar 
.setSecurityDelegate(WOWebServiceRegistrar.java:358)


My configuration is:

Eclipse 3.3.2 WOLips 3.3.5047
OS X 10.5.2
Java 1.5.0_13

I have standard Apple distribution jars for:

axis.jar (1.4 I think)
commons-discovery-0.2.jar
commons-logging-1.0.4.jar
jaxrpc.jar
saaj.jar
wsdl4j-1.5.1.jar
xmlsec-1.3.0.jar

What else?  Yes, the WebObjects webservices frameworks are  
installed.  Yes, the error is the same running within Eclipse or at  
the command line.


Any help is very appreciated.

--
Jon Nolan

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/mschrag%40mdimension.com

This email sent to [EMAIL PROTECTED]



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: ERXMigrationTable.newColumn does not set default value

2008-04-22 Thread Mike Schrag
Defaults are not covered by the spec at all ... Wonder migrations set  
the default value into the er.extensions.eoattribute.default  
attribute in the attribute's userInfo during the migration process, so  
that's the one you should use.  FrontBasePlugIn has an example of this.


ms

On Apr 22, 2008, at 11:45 AM, David Avendasora wrote:


Hey Mike,

I'm trying to add this to the ERXMicrosoftPlugin. I've added a  
MicrosoftExpression Inner Class, and copied over much of what was in  
the FrontbaseExpression IC with some modifications, but it looks  
like where it's looking for the default value is from the UserInfo  
dictionary in the model as opposed to the default value from the  
Migration class.


Is the userInfo dictionary the place to set defaults, and if so,  
what does the default do in the Migration class?


Thanks!

Dave

On Apr 10, 2008, at 8:50 AM, Mike Schrag wrote:

I lied anyway ... I DID actually add it to PGSQL, but I was looking  
in the same wrong place that you were.  FB PlugIn has two inner  
classes (the schemasync and the expression).  That method is  
actually in expression, not in sync, so when I looked in PG, I  
didn't see it.  However, it IS in PGExpession.  So I suspect your  
problems with defaults not working are due to 5.4.  I'll have to  
check this for the next WO version, but I SUSPECT it works  
properly ...


ms

On Apr 10, 2008, at 8:38 AM, Mike Schrag wrote:

The statment is built by a call of  
EOSchemaSynchronizationFactory._columnCreationClauseForAttribute  
method which calls - in my case
I take it (based on the class name you're referencing here) that  
you're using WO 5.4.x?  The problems I've discussed previously  
with respect to SQL generation in 5.4.x make this not worth trying  
to fix on your 5.4.x build until the next WO release ...


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/mschrag%40mdimension.com

This email sent to [EMAIL PROTECTED]



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/webobjects%40avendasora.com

This email sent to [EMAIL PROTECTED]




___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/mschrag%40mdimension.com

This email sent to [EMAIL PROTECTED]



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: Fake Entity Modeler validation errors preventing veogen running

2008-04-22 Thread Mike Schrag
I've updated to WOLips 3.3.5047 but now can't veogenerate because of  
model validation errors relating to key paths in model based fetch  
specifications:


/Users/Eclipse2/ClickTravelSystem/Resources/TravelSystem.eogen:
	Error: The fetch specification TrainLegSearch has a qualifier that  
refers to an invalid key 'segment.voucher.ticketQueue'.
	Error: The fetch specification TrainLegSearch has a qualifier that  
refers to an invalid key 'flightNumber'.


However, the key paths refer to attributes contained within  
subclasses so are actually valid.


We had the same problem with Component Editor validation but can  
work around it by marking them with


//VALID

Is there something similar in Entity Modeler ?

It's a known bug in EM and there is no way to mark them valid ...

ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: ERXMigrationTable.newColumn does not set default value

2008-04-22 Thread David Avendasora
Okay, I'll get the ability to read the  
er.extensions.eoattribute.default along with the simpler Default  
from the userInfo dictionary setup in the ERXMicrosoftPlugIn like it  
is in the FrontbasePlugIn, and then hand it off you you and Chuck for  
review.


Out of curiosity, what does the defaultValue parameter as defined in  
the Wonder API for ERXMigrationsTable.newXxxxColumn() do?


For example: 
http://webobjects.mdimension.com/wonder/api/er/extensions/migration/ERXMigrationTable.html#newStringColumn(java.lang.String,%20int,%20boolean,%20java.lang.String)

Thanks!

Dave

On Apr 22, 2008, at 12:52 PM, Mike Schrag wrote:

Defaults are not covered by the spec at all ... Wonder migrations  
set the default value into the er.extensions.eoattribute.default  
attribute in the attribute's userInfo during the migration process,  
so that's the one you should use.  FrontBasePlugIn has an example of  
this.


ms

On Apr 22, 2008, at 11:45 AM, David Avendasora wrote:


Hey Mike,

I'm trying to add this to the ERXMicrosoftPlugin. I've added a  
MicrosoftExpression Inner Class, and copied over much of what was  
in the FrontbaseExpression IC with some modifications, but it looks  
like where it's looking for the default value is from the UserInfo  
dictionary in the model as opposed to the default value from the  
Migration class.


Is the userInfo dictionary the place to set defaults, and if so,  
what does the default do in the Migration class?


Thanks!

Dave

On Apr 10, 2008, at 8:50 AM, Mike Schrag wrote:

I lied anyway ... I DID actually add it to PGSQL, but I was  
looking in the same wrong place that you were.  FB PlugIn has two  
inner classes (the schemasync and the expression).  That method is  
actually in expression, not in sync, so when I looked in PG, I  
didn't see it.  However, it IS in PGExpession.  So I suspect your  
problems with defaults not working are due to 5.4.  I'll have to  
check this for the next WO version, but I SUSPECT it works  
properly ...


ms

On Apr 10, 2008, at 8:38 AM, Mike Schrag wrote:

The statment is built by a call of  
EOSchemaSynchronizationFactory._columnCreationClauseForAttribute  
method which calls - in my case
I take it (based on the class name you're referencing here) that  
you're using WO 5.4.x?  The problems I've discussed previously  
with respect to SQL generation in 5.4.x make this not worth  
trying to fix on your 5.4.x build until the next WO release ...


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/mschrag%40mdimension.com

This email sent to [EMAIL PROTECTED]



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/webobjects%40avendasora.com

This email sent to [EMAIL PROTECTED]




___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/mschrag%40mdimension.com

This email sent to [EMAIL PROTECTED]



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/webobjects%40avendasora.com

This email sent to [EMAIL PROTECTED]




___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: ERXMigrationTable.newColumn does not set default value

2008-04-22 Thread Mike Schrag
Okay, I'll get the ability to read the  
er.extensions.eoattribute.default along with the simpler Default  
from the userInfo dictionary setup in the ERXMicrosoftPlugIn like it  
is in the FrontbasePlugIn, and then hand it off you you and Chuck  
for review.
Default is a contrivance added in the FB plugin only, and is not  
used by anything, so I wouldn't bother.  I actually would have used  
Default except that FB decided to define that thing as having an  
unclear attribute formatting, which I why I defined a new one.


Out of curiosity, what does the defaultValue parameter as defined in  
the Wonder API for ERXMigrationsTable.newXxxxColumn() do?


For example: 
http://webobjects.mdimension.com/wonder/api/er/extensions/migration/ERXMigrationTable.html#newStringColumn(java.lang.String,%20int,%20boolean,%20java.lang.String)
It gets passed as a default value for the database column (you can see  
the sql it generates).  This exists only to accommodate migrations  
from allows null true to allows null false to deal with existing data  
and should not be relied on for ANYTHING else.


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


iBox Question

2008-04-22 Thread James Cicenia

Hello -

I am trying to use iBox and create a dialog in my browser. I am using  
the embedded noshow div to hide that part of the form.
However, I can't for the life of me figure why it is sending me null  
back for any webobject bound in that noshow div area.


Has anyone used iBox?

Thanks
James Cicenia

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: Fake Entity Modeler validation errors preventing veogen running

2008-04-22 Thread Simon McLean
I've updated to WOLips 3.3.5047 but now can't veogenerate because  
of model validation errors relating to key paths in model based  
fetch specifications:


/Users/Eclipse2/ClickTravelSystem/Resources/TravelSystem.eogen:
	Error: The fetch specification TrainLegSearch has a qualifier that  
refers to an invalid key 'segment.voucher.ticketQueue'.
	Error: The fetch specification TrainLegSearch has a qualifier that  
refers to an invalid key 'flightNumber'.


However, the key paths refer to attributes contained within  
subclasses so are actually valid.


We had the same problem with Component Editor validation but can  
work around it by marking them with


//VALID

Is there something similar in Entity Modeler ?

It's a known bug in EM and there is no way to mark them valid ...


I can't see a Jira open for it. Do you want me to create one ?

I've successfully rolled my Eclipse back to a previous state so I'm  
working again now, but the errors are still sitting in my warning  
pane. Tried a Clean all but that's not got rid of them. I seem to  
remember someone posting a way of deleting them manually - is that  
possible ? Or any other suggestions on how I get rid of them ?


Thanks, Simon



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: ERXMigrationTable.newColumn does not set default value

2008-04-22 Thread Mike Schrag
Microsoft is so helpful in that after the initial creation of a  
table it doesn't allow you to add a new column and mark it as NOT  
NULL without defining a default value. You can _not_ create the  
column, then populate it after. You have to do it all in one  
function. I suppose the work around it to create it as NULL, then  
_change_ it to NOT NULL with a second function. It would be nice to  
handle it so it is invisible to the developer that SQL Server is a  
PITA though...
Even with no rows in the table?  That's an odd restriction ...  
However, after table creation presumably means in another  
migration which couldn't guarantee the table is empty anyway, so this  
seems like a reasonable restriction (as your migration should probably  
ALWAYS declare a default for new not-null columns to protect against  
this).


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: WOWebServiceRegistrar.setSecurityDelegate failure

2008-04-22 Thread Mr. Pierre Frisch
Actually 5.3.x has most of the same issues as 5.4.1. Grab 5.4.2 from  
the iPhone SDK beta if you can. WebObjects 5.3.x and 5.4.x use  
different versions of Axis. 5.3.x uses Axis 1.1. 5.4.x uses Axis 1.4.


I am a bit puzzled by the statement:  
org.apache.xml.security.Init.init(); Where does this come from? It is  
not part of regular Axis 1.4.


Cheers

Pierre
--
Pierre Frisch
[EMAIL PROTECTED]


On Apr 22, 2008, at 9:49, Mike Schrag wrote:

There are known issues with WebServices on 5.4.1.  I don't know if  
this one in particular is fixed in 5.4.2 or not, but you should  
definitely get a baseline on 5.3 FIRST before doing the full move to  
5.4.1.


ms

On Apr 22, 2008, at 12:09 PM, Jon Nolan wrote:

I'm in the process of upgrading all my WebObjects projects from  
Xcode/Java1.4/WebObjects5.2.3 to Eclipse/WOLips/Java1.5/ 
WebObjects5.4.1.  I've possibly bitten off way too big a bite all  
at once.  ;-)


I'm muddling through without too much trouble except when it comes  
to webservices.  I make substantive use of WOSecurityDelegate.  It  
is absolutely necessary in a number of places.  I just can't seem  
to get a delegate registered though.  Here's the code:


org.apache.xml.security.Init.init();
	WOWebServiceRegistrar.setSecurityDelegate(new  
MyWOSecurityDelegateClass());


The registration fails with:

java.lang.NullPointerException
	at  
com 
.webobjects 
.appserver 
._private 
.WOServerSessionHandler.clinit(WOServerSessionHandler.java:25)

at java.lang.Class.forName0(Native Method)
at java.lang.Class.forName(Class.java:242)
at org.apache.axis.utils.ClassUtils$2.run(ClassUtils.java:177)
at java.security.AccessController.doPrivileged(Native Method)
at org.apache.axis.utils.ClassUtils.loadClass(ClassUtils.java:160)
at org.apache.axis.utils.ClassUtils.forName(ClassUtils.java:100)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getJavaClass(WSDDDeployableItem.java:353)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.makeNewInstance(WSDDDeployableItem.java:295)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getNewInstance(WSDDDeployableItem.java:274)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getInstance(WSDDDeployableItem.java:260)
	at  
org 
.apache 
.axis.deployment.wsdd.WSDDChain.makeNewInstance(WSDDChain.java:125)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getNewInstance(WSDDDeployableItem.java:274)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployableItem.getInstance(WSDDDeployableItem.java:260)
	at  
org 
.apache 
.axis 
.deployment 
.wsdd.WSDDDeployment.getGlobalRequest(WSDDDeployment.java:473)
	at  
com 
.webobjects 
.webservices 
.support.WOXMLProvider.getGlobalRequest(WOXMLProvider.java:168)

at org.apache.axis.AxisEngine.getGlobalRequest(AxisEngine.java:365)
	at  
com 
.webobjects 
.appserver 
._private.WOWebService.setSecurityDelegate(WOWebService.java:205)
	at  
com 
.webobjects 
.appserver 
.WOWebServiceRegistrar 
.setSecurityDelegate(WOWebServiceRegistrar.java:358)


My configuration is:

Eclipse 3.3.2 WOLips 3.3.5047
OS X 10.5.2
Java 1.5.0_13

I have standard Apple distribution jars for:

axis.jar (1.4 I think)
commons-discovery-0.2.jar
commons-logging-1.0.4.jar
jaxrpc.jar
saaj.jar
wsdl4j-1.5.1.jar
xmlsec-1.3.0.jar

What else?  Yes, the WebObjects webservices frameworks are  
installed.  Yes, the error is the same running within Eclipse or at  
the command line.


Any help is very appreciated.

--
Jon Nolan

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/mschrag%40mdimension.com

This email sent to [EMAIL PROTECTED]



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/pierre%40apple.com

This email sent to [EMAIL PROTECTED]




smime.p7s
Description: S/MIME cryptographic signature
 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]

Re: ERXMigrationTable.newColumn does not set default value

2008-04-22 Thread David Avendasora

The problem is that even though I use:

table.newStringColumn(Some_Column, 255, false, Default)

The SQL generated is be: ALTER TABLE dbo.My_Table ADD Some_Column  
varchar(255) NOT NULL


Which generates an error stating that either it has to be NULL, or a  
default be passed. As you can see, I did define a default value  
(Default) in the Migration, but the plugin isn't picking it up. I'm  
working on getting the ERXMicrosoftPlugIn to do it, but I can't even  
find the code in the FrontbasePlugIn that would add the DEFAULT clause  
so I'm not sure how to get the Microsoft plugin to do it. That is what  
is confusing me.


Dave


On Apr 22, 2008, at 2:46 PM, Mike Schrag wrote:

Even with no rows in the table?  That's an odd restriction ...  
However, after table creation presumably means in another  
migration which couldn't guarantee the table is empty anyway, so  
this seems like a reasonable restriction (as your migration should  
probably ALWAYS declare a default for new not-null columns to  
protect against this).


___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: ERXMigrationTable.newColumn does not set default value

2008-04-22 Thread David Avendasora


On Apr 22, 2008, at 1:23 PM, Mike Schrag wrote:

Okay, I'll get the ability to read the  
er.extensions.eoattribute.default along with the simpler  
Default from the userInfo dictionary setup in the  
ERXMicrosoftPlugIn like it is in the FrontbasePlugIn, and then hand  
it off you you and Chuck for review.
Default is a contrivance added in the FB plugin only, and is not  
used by anything, so I wouldn't bother.  I actually would have used  
Default except that FB decided to define that thing as having an  
unclear attribute formatting, which I why I defined a new one.


Okay. No problem.



Out of curiosity, what does the defaultValue parameter as defined  
in the Wonder API for ERXMigrationsTable.newXxxxColumn() do?


For example: 
http://webobjects.mdimension.com/wonder/api/er/extensions/migration/ERXMigrationTable.html#newStringColumn(java.lang.String,%20int,%20boolean,%20java.lang.String)
It gets passed as a default value for the database column (you can  
see the sql it generates).  This exists only to accommodate  
migrations from allows null true to allows null false to deal with  
existing data and should not be relied on for ANYTHING else.


Hmm...

Microsoft is so helpful in that after the initial creation of a table  
it doesn't allow you to add a new column and mark it as NOT NULL  
without defining a default value. You can _not_ create the column,  
then populate it after. You have to do it all in one function. I  
suppose the work around it to create it as NULL, then _change_ it to  
NOT NULL with a second function. It would be nice to handle it so it  
is invisible to the developer that SQL Server is a PITA though...


Dave
___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: CLOSE_WAIT processes/ instances HANG ???

2008-04-22 Thread Alexander Spohr

I think you are using an old Apache-plugin (mod_WebObjects.so).
Grab, compile and install a new one.

atze



Am 22.04.2008 um 15:47 schrieb Shravan Kumar.M:


We have been battered by CLOSE_WAIT processes and instances hang.
We see CLOSE_WAIT state of a process for both java and httpd  
processes.


___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: Removing extra blank lines from generated html?

2008-04-22 Thread Mark Morris
I think if you look at your component, you'll see that the blank lines  
are in there, not generated by WebObjects.  I've put enough  
conditionals in the middle of a line that I definitely would have been  
bit by now if that was the case! ;-)


Regards,
Mark

On Apr 18, 2008, at 4:35 PM, John Huss wrote:

Why do you care about the blank line?  Or do you want to remove all  
extra whitespace of any kind?


John

On Fri, Apr 18, 2008 at 4:29 PM, Alan Zebchuk [EMAIL PROTECTED] 
 wrote:
Does anyone have a solution for removing the extra blank lines from  
the html generated by a component? i.e. when there's a conditional  
where the condition is false, the generated html has blank lines  
where the conditional would have been.


Thanks,

Alan


___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/markm%40onpointsoftware.com

This email sent to [EMAIL PROTECTED]


 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]

Re: WOWebServiceRegistrar.setSecurityDelegate failure

2008-04-22 Thread Jon Nolan

Thanks guys.

Pierre, if 5.4.x versions all use Axis 1.4 how does moving to 5.4.2 help?  Regarding the call to org.apache.xml.security.Init, I'm sorry, it's extraneous to this conversation and I didn't mean to post it.  I use it for some webservice encryption needs. 

Any advice on a somewhat guaranteed stable/working WebObjects/WOLips/Java/Axis environment for non-trivial webservices is very, very appreciated.  


Thanks,
Jon Nolan

Mr. Pierre Frisch wrote:
Actually 5.3.x has most of the same issues as 5.4.1. Grab 5.4.2 from the 
iPhone SDK beta if you can. WebObjects 5.3.x and 5.4.x use different 
versions of Axis. 5.3.x uses Axis 1.1. 5.4.x uses Axis 1.4.


I am a bit puzzled by the statement: 
org.apache.xml.security.Init.init(); Where does this come from? It is 
not part of regular Axis 1.4.


Cheers

Pierre
--
Pierre Frisch
[EMAIL PROTECTED]


On Apr 22, 2008, at 9:49, Mike Schrag wrote:

There are known issues with WebServices on 5.4.1.  I don't know if 
this one in particular is fixed in 5.4.2 or not, but you should 
definitely get a baseline on 5.3 FIRST before doing the full move to 
5.4.1.


ms

On Apr 22, 2008, at 12:09 PM, Jon Nolan wrote:

I'm in the process of upgrading all my WebObjects projects from 
Xcode/Java1.4/WebObjects5.2.3 to 
Eclipse/WOLips/Java1.5/WebObjects5.4.1.  I've possibly bitten off way 
too big a bite all at once.  ;-)


I'm muddling through without too much trouble except when it comes to 
webservices.  I make substantive use of WOSecurityDelegate.  It is 
absolutely necessary in a number of places.  I just can't seem to get 
a delegate registered though.  Here's the code:


org.apache.xml.security.Init.init();
WOWebServiceRegistrar.setSecurityDelegate(new 
MyWOSecurityDelegateClass());


The registration fails with:

java.lang.NullPointerException
at 
com.webobjects.appserver._private.WOServerSessionHandler.clinit(WOServerSessionHandler.java:25) 


at java.lang.Class.forName0(Native Method)
at java.lang.Class.forName(Class.java:242)
at org.apache.axis.utils.ClassUtils$2.run(ClassUtils.java:177)
at java.security.AccessController.doPrivileged(Native Method)
at org.apache.axis.utils.ClassUtils.loadClass(ClassUtils.java:160)
at org.apache.axis.utils.ClassUtils.forName(ClassUtils.java:100)
at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getJavaClass(WSDDDeployableItem.java:353) 

at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.makeNewInstance(WSDDDeployableItem.java:295) 

at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getNewInstance(WSDDDeployableItem.java:274) 

at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getInstance(WSDDDeployableItem.java:260) 

at 
org.apache.axis.deployment.wsdd.WSDDChain.makeNewInstance(WSDDChain.java:125) 

at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getNewInstance(WSDDDeployableItem.java:274) 

at 
org.apache.axis.deployment.wsdd.WSDDDeployableItem.getInstance(WSDDDeployableItem.java:260) 

at 
org.apache.axis.deployment.wsdd.WSDDDeployment.getGlobalRequest(WSDDDeployment.java:473) 

at 
com.webobjects.webservices.support.WOXMLProvider.getGlobalRequest(WOXMLProvider.java:168) 


at org.apache.axis.AxisEngine.getGlobalRequest(AxisEngine.java:365)
at 
com.webobjects.appserver._private.WOWebService.setSecurityDelegate(WOWebService.java:205) 

at 
com.webobjects.appserver.WOWebServiceRegistrar.setSecurityDelegate(WOWebServiceRegistrar.java:358) 



My configuration is:

Eclipse 3.3.2 WOLips 3.3.5047
OS X 10.5.2
Java 1.5.0_13

I have standard Apple distribution jars for:

axis.jar (1.4 I think)
commons-discovery-0.2.jar
commons-logging-1.0.4.jar
jaxrpc.jar
saaj.jar
wsdl4j-1.5.1.jar
xmlsec-1.3.0.jar

What else?  Yes, the WebObjects webservices frameworks are 
installed.  Yes, the error is the same running within Eclipse or at 
the command line.


Any help is very appreciated.

--
Jon Nolan

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/mschrag%40mdimension.com 



This email sent to [EMAIL PROTECTED]



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/pierre%40apple.com

This email sent to [EMAIL PROTECTED]



___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: WOWebServiceRegistrar.setSecurityDelegate failure

2008-04-22 Thread Mike Schrag
Pierre, if 5.4.x versions all use Axis 1.4 how does moving to 5.4.2  
help?  Regarding the call to org.apache.xml.security.Init, I'm  
sorry, it's extraneous to this conversation and I didn't mean to  
post it.  I use it for some webservice encryption needs.
Any advice on a somewhat guaranteed stable/working WebObjects/WOLips/ 
Java/Axis environment for non-trivial webservices is very, very  
appreciated.
I have been using 5.3 with web services for several years without  
problems, though this very much depends on what is on the other side  
of the service.  I know Anjo had several issues with 5.4.0/5.4.1 and  
web services, though Pierre has fixed all of these in 5.4.2 as far as  
I know.  I have an app that I can run now with 5.4.2 or 5.3 without  
any issues, though I don't know about your particular case.  I  
recommended trying 5.3 first because it has been out for quite a while  
and most of the issues are known, or at the very least, as Pierre  
recommends, download 5.4.2 from the iPhone SDK and try that.  I  
wouldn't spend any time trying to diagnose this in any environment  
other than 5.3 or 5.4.2 because I think it's probably a waste of time  
since work was specifically done on this area.


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: WOWebServiceRegistrar.setSecurityDelegate failure

2008-04-22 Thread Anjo Krank
5.4.1 wouldn't even start without serious prodding, there where pretty  
serious bugs in registering a service. In 5.4.2, all the issues I have  
left are in axis itself and not really under Pierre's control to fix  
(which are related to special client and server libs). In particular  
the bug reported lately (axis 1.4 creates refs for primitive types).


Cheers, Anjo

Am 22.04.2008 um 22:15 schrieb Mike Schrag:

Pierre, if 5.4.x versions all use Axis 1.4 how does moving to 5.4.2  
help?  Regarding the call to org.apache.xml.security.Init, I'm  
sorry, it's extraneous to this conversation and I didn't mean to  
post it.  I use it for some webservice encryption needs.
Any advice on a somewhat guaranteed stable/working WebObjects/ 
WOLips/Java/Axis environment for non-trivial webservices is very,  
very appreciated.
I have been using 5.3 with web services for several years without  
problems, though this very much depends on what is on the other side  
of the service.  I know Anjo had several issues with 5.4.0/5.4.1 and  
web services, though Pierre has fixed all of these in 5.4.2 as far  
as I know.  I have an app that I can run now with 5.4.2 or 5.3  
without any issues, though I don't know about your particular case.   
I recommended trying 5.3 first because it has been out for quite a  
while and most of the issues are known, or at the very least, as  
Pierre recommends, download 5.4.2 from the iPhone SDK and try that.   
I wouldn't spend any time trying to diagnose this in any environment  
other than 5.3 or 5.4.2 because I think it's probably a waste of  
time since work was specifically done on this area.


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/krank%40logicunited.com

This email sent to [EMAIL PROTECTED]


___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


WOEventDisplay (WOEvent logging password)

2008-04-22 Thread Gavin Eadie
I've been trying to get the /wa/WOEventDisplay direct action to work.   
The description in the Javadoc for WOEvent says you need a property  
EOEventLoggingPassword set to provide a password that will let you  
access the event displays.


In fact, this does NOT work as described -- you need to:

 
EOEventCenter.setPassword(System.getProperty(EOEventLoggingPassword));


(in the application constructor is a good place).

I'll log this in BugReporter today (5.4.2 has the bug) unless someone  
jumps out and tells me I'm missing something obvious.

_

Moral PS:  When I went hunting for some reference to this I found a  
message exchange between myself and Alex Johnson from January 2003:




At 10:26 AM -0600 1/7/03, Alex Johnson wrote:
I noticed that there's a setPassword method in the EOEventCenter.   
You could use this and see if it changes (a)  
EOEventCenter.password() and (b) the system property.  Maybe the  
key is not actually EOEventLoggingPassword in spite of the  
documentation.


I'm now using EOEventCenter.setPassword(xxx) because I can't get  
it to work via properties.  I checked the key too, and  
EOEventLoggingPassword seems to be right (I decompiled the  
WebObjects code to check!).


My code is now:


EOEventCenter 
.setPassword(System.getProperty(EOEventLoggingPassword))


  which is what WebObjects should be doing itself ... Oh  
Well ... Gav

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Extract the component name from a binding

2008-04-22 Thread Helmut Schottmüller

Hi,

surely another stupid question but I am stuck and can't think  
straight... Maybe it's time to go to bed.

Assume I have three components:

comp1, comp2, comp3

The name of the component is set via an AjaxUpdateLink to a binding  
selectedComponent.
Depending on the value of selectedComponent, one of the components is  
shown in the update container:


wo:if condition = ~selectedComponent.equals(\comp1\) ? true :  
false

  wo:comp1/wo:comp1
/wo:if
wo:if condition = ~selectedComponent.equals(\comp2\) ? true :  
false

  wo:comp2/wo:comp2
/wo:if
wo:if condition = ~selectedComponent.equals(\comp3\) ? true :  
false

  wo:comp3/wo:comp3
/wo:if


Is it possible to have one single binding to call the selected  
component in an AjaxUpdateContainer instead of having a conditional  
for each of the components?


Regards,

Helmut

smime.p7s
Description: S/MIME cryptographic signature
 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]

Re: Extract the component name from a binding

2008-04-22 Thread Mike Schrag

comp1, comp2, comp3

The name of the component is set via an AjaxUpdateLink to a binding  
selectedComponent.
Depending on the value of selectedComponent, one of the components  
is shown in the update container:


wo:if condition = ~selectedComponent.equals(\comp1\) ? true :  
false

Comments:

1) use 'comp1' instead of comp1 -- no need for escaping then
2) ? true : false is not necessary -- equals returns a boolean
3) use ERXEqualConditional instead of OGNL
4) use WOSwitchComponent instead of all of this

ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: Extract the component name from a binding

2008-04-22 Thread Helmut Schottmüller

Hi Mike,

thanks, especially for 4)
Sometimes it's hard to identify the suitable component...

regards,
Helmut


Am 22.04.2008 um 23:02 schrieb Mike Schrag:


comp1, comp2, comp3

The name of the component is set via an AjaxUpdateLink to a binding  
selectedComponent.
Depending on the value of selectedComponent, one of the components  
is shown in the update container:


wo:if condition = ~selectedComponent.equals(\comp1\) ? true :  
false

Comments:

1) use 'comp1' instead of comp1 -- no need for escaping then
2) ? true : false is not necessary -- equals returns a boolean
3) use ERXEqualConditional instead of OGNL
4) use WOSwitchComponent instead of all of this

ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/helmut.schottmueller%40mac.com

This email sent to [EMAIL PROTECTED]




smime.p7s
Description: S/MIME cryptographic signature
 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]

Re: ERXMigrationTable.newColumn does not set default value

2008-04-22 Thread David Avendasora


On Apr 22, 2008, at 3:05 PM, David Avendasora wrote:

Which generates an error stating that either it has to be NULL, or a  
default be passed. As you can see, I did define a default value  
(Default) in the Migration, but the plugin isn't picking it up.  
I'm working on getting the ERXMicrosoftPlugIn to do it, but I can't  
even find the code in the FrontbasePlugIn that would add the DEFAULT  
clause so I'm not sure how to get the Microsoft plugin to do it.  
That is what is confusing me.


Okay, so it looks to me like the EOSynchronizationFactory method  
_columnCreationClauseForAttribute(EOAttribute attribute) is the one I  
need to override and add the DEFAULT clause to the end of. I notice  
that this method is not overridden in the FrontbasePlugIn though. Is  
this the right place?


Previously Mike said:

It gets passed as a default value for the database column (you can  
see the sql it generates).  This exists only to accommodate  
migrations from allows null true to allows null false to deal with  
existing data and should not be relied on for ANYTHING else.



And this is pretty-much exactly what I want to do, but the  
ERXMicrosoftPlugIn doesn't seem to add the DEFAULT clause to the ALTER  
TABLE statement.


Dave
___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: WOWebServiceRegistrar.setSecurityDelegate failure

2008-04-22 Thread Mr. Pierre Frisch
Without going into too much details when WebServices for WebObjects  
were created with Axis 1.0 type information in WDSL document was  
optional, WebOjects did not include it. With Axis 1.1 and event more  
for Axis 1.4 it became required for custom types. This was fixed in WO  
5.4.x unfortunately it took us multiple iterations to cover all the  
cases (and there are many). WebObjects 5.3.x and Axis 1.1 work if you  
do not use any custom types i.e. you do not have your own serializer/ 
deserializer. As Anjo said there are some major difference in the  
registration between 5.4.0/5.4.1 and 5.4.2. You should use the later.  
I have written a quick test case and at least it does not throw the  
exception you are seeing.


As I said WO 5.4.2 is available with the iPhone SDK.

Cheers

Pierre
--
Pierre Frisch
[EMAIL PROTECTED]


On Apr 22, 2008, at 13:24, Anjo Krank wrote:

5.4.1 wouldn't even start without serious prodding, there where  
pretty serious bugs in registering a service. In 5.4.2, all the  
issues I have left are in axis itself and not really under Pierre's  
control to fix (which are related to special client and server  
libs). In particular the bug reported lately (axis 1.4 creates refs  
for primitive types).


Cheers, Anjo

Am 22.04.2008 um 22:15 schrieb Mike Schrag:

Pierre, if 5.4.x versions all use Axis 1.4 how does moving to  
5.4.2 help?  Regarding the call to org.apache.xml.security.Init,  
I'm sorry, it's extraneous to this conversation and I didn't mean  
to post it.  I use it for some webservice encryption needs.
Any advice on a somewhat guaranteed stable/working WebObjects/ 
WOLips/Java/Axis environment for non-trivial webservices is very,  
very appreciated.
I have been using 5.3 with web services for several years without  
problems, though this very much depends on what is on the other  
side of the service.  I know Anjo had several issues with  
5.4.0/5.4.1 and web services, though Pierre has fixed all of these  
in 5.4.2 as far as I know.  I have an app that I can run now with  
5.4.2 or 5.3 without any issues, though I don't know about your  
particular case.  I recommended trying 5.3 first because it has  
been out for quite a while and most of the issues are known, or at  
the very least, as Pierre recommends, download 5.4.2 from the  
iPhone SDK and try that.  I wouldn't spend any time trying to  
diagnose this in any environment other than 5.3 or 5.4.2 because I  
think it's probably a waste of time since work was specifically  
done on this area.


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/krank%40logicunited.com

This email sent to [EMAIL PROTECTED]


___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/pierre%40apple.com

This email sent to [EMAIL PROTECTED]




smime.p7s
Description: S/MIME cryptographic signature
 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]

Re: WOComponent children

2008-04-22 Thread Mike Schrag

On Jan 16, 2008, at 6:09 PM, [EMAIL PROTECTED] wrote:

In parting, I hope you do get that DOM manipulation working and we  
can all benefit from it.


Yours,
-- Aaron
Just to tie a ribbon on this thread (which I've lugged around in my  
inbox for months :) ), on-demand loading of Javascript has finally  
been committed to Ajax.framework.  The nastiest part is actually  
standing on the shoulders of a patch that will be in Prototype 2.0  
that we brought forward into Ajax.framework, but it appears to work in  
all of the browsers I tested on (including Safari 2.0 which was one of  
the ones that previously had issues with the dom route -- the patch is  
disgusting and clever, though).


ms ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]

Re: CLOSE_WAIT processes/ instances HANG ???

2008-04-22 Thread shravan
Hello Alexandar,

Not really I think!

I installed Apache 1.3.36 recently and then compiled mod_WebObjects.so in
the same system later.

Any other idea?

Thank You
Shravan
--

On Wed, April 23, 2008 1:22 am, Alexander Spohr wrote:
 I think you are using an old Apache-plugin (mod_WebObjects.so).
 Grab, compile and install a new one.

   atze



 Am 22.04.2008 um 15:47 schrieb Shravan Kumar.M:

 We have been battered by CLOSE_WAIT processes and instances hang.
 We see CLOSE_WAIT state of a process for both java and httpd
 processes.




 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


The type NSMutableArray is not generic; it cannot be parameterized with arguments

2008-04-22 Thread Archibald Singleton

HI All,

OK, I'm stuck.

I'm trying to make some Apple provided WO examples to work. The  
examples came with WO 5.4 but I'm running WO 5.3.(I could switch to  
using 5.4 just to run the examples but I'd like to understand what's  
going on)


I keep getting the above error message even though I've included  
Wonder JavaWOExtensions as both a project and a frameworks in the  
examples project (using the Java Build Path project properties.
Since JavaWOExtensions provides an implementation of NSArray with  
generics I would expect this the the example code would pick that  
implementation. Apparently it doesn't.


So I guess my question is why?!?

Does Wonder have to do some behind the scene patching to make the  
JavaWOExtensions classes supersedes those in the regular WO  
frameworks? And if so would someone points me to where it does this in  
the code?


Any help will be greatly appreciated.

= tmk =

PS: I know that WO 5.4 now provides an implementation of NSArray and  
friends with generics. 
___

Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: The type NSMutableArray is not generic; it cannot be parameterized with arguments

2008-04-22 Thread Jeff Schmitz
I'm pretty sure it's purely a versioning problem.  NSMutableArray  
didn't become generic until WO5.4.  If you want to know about Java  
Generics I'd suggest googling on the subject.  There is LOTS of info  
on it.


Jeff

On Apr 22, 2008, at 8:39 PM, Archibald Singleton wrote:


HI All,

OK, I'm stuck.

I'm trying to make some Apple provided WO examples to work. The  
examples came with WO 5.4 but I'm running WO 5.3.(I could switch to  
using 5.4 just to run the examples but I'd like to understand what's  
going on)


I keep getting the above error message even though I've included  
Wonder JavaWOExtensions as both a project and a frameworks in the  
examples project (using the Java Build Path project properties.
Since JavaWOExtensions provides an implementation of NSArray with  
generics I would expect this the the example code would pick that  
implementation. Apparently it doesn't.


So I guess my question is why?!?

Does Wonder have to do some behind the scene patching to make the  
JavaWOExtensions classes supersedes those in the regular WO  
frameworks? And if so would someone points me to where it does this  
in the code?


Any help will be greatly appreciated.

= tmk =

PS: I know that WO 5.4 now provides an implementation of NSArray and  
friends with generics.___

Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/jeffandmonica%40mac.com

This email sent to [EMAIL PROTECTED]


___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: The type NSMutableArray is not generic; it cannot be parameterized with arguments

2008-04-22 Thread Mike Schrag
I keep getting the above error message even though I've included  
Wonder JavaWOExtensions as both a project and a frameworks in the  
examples project (using the Java Build Path project properties.
Since JavaWOExtensions provides an implementation of NSArray with  
generics I would expect this the the example code would pick that  
implementation. Apparently it doesn't.
Generics for 5.3 are in Wonder's ERExtensions framework, not  
JavaWOExtensions.  It's not enough to just include JavaWOExtensions  
from Wonder if you want that.


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: The type NSMutableArray is not generic; it cannot be parameterized with arguments

2008-04-22 Thread Archibald Singleton

On 23 Apr 2008, at 04:12, Mike Schrag wrote:

I keep getting the above error message even though I've included  
Wonder JavaWOExtensions as both a project and a frameworks in the  
examples project (using the Java Build Path project properties.
Since JavaWOExtensions provides an implementation of NSArray with  
generics I would expect this the the example code would pick that  
implementation. Apparently it doesn't.
Generics for 5.3 are in Wonder's ERExtensions framework, not  
JavaWOExtensions.  It's not enough to just include JavaWOExtensions  
from Wonder if you want that.


OK, Thanks Mike!

= tmk =


___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: WOFileUpload

2008-04-22 Thread sirius black
Chuck,
I'm curious how you would have Apache handle this.  I know
of LimitRequestBody, but I haven't used it much, is this the route you were
thinking?

Thanks.



On Fri, Apr 18, 2008 at 5:23 PM, Chuck Hill [EMAIL PROTECTED]
wrote:


 On Apr 18, 2008, at 5:10 PM, sirius black wrote:

  I'm trying to verify the size of a file upload and limit its upload
  before actually sending any data across the wire.  I know I can check the
  content-length header early on in the game in dispatchRequest, but even if I
  just throw a RuntimeException at that point, I can't stop WebObjects from
  reading in the data anyways, before sending back a response.  I'm using
  WOFileUpload with streamToFilePath.
 
  Fwiw, I've looked into javascript validation to try and check the
  content-length client side first, but haven't had success with that.
 
  Thoughts?
 


 Your best bet is probably to get Apache to do this.

 Chuck


 --

 Practical WebObjects - for developers who want to increase their overall
 knowledge of WebObjects or who are trying to solve specific problems.
 http://www.global-village.net/products/practical_webobjects






 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]

Re: WOFileUpload

2008-04-22 Thread Chuck Hill
Offhand, I am not certain.  LimitRequestBody sounds correct.  It has  
been a while..


Chuck

On Apr 22, 2008, at 7:29 PM, sirius black wrote:


Chuck,

I'm curious how you would have Apache handle this.  I know of  
LimitRequestBody, but I haven't used it much, is this the route you  
were thinking?


Thanks.



On Fri, Apr 18, 2008 at 5:23 PM, Chuck Hill [EMAIL PROTECTED] 
village.net wrote:


On Apr 18, 2008, at 5:10 PM, sirius black wrote:

I'm trying to verify the size of a file upload and limit its upload  
before actually sending any data across the wire.  I know I can  
check the content-length header early on in the game in  
dispatchRequest, but even if I just throw a RuntimeException at that  
point, I can't stop WebObjects from reading in the data anyways,  
before sending back a response.  I'm using WOFileUpload with  
streamToFilePath.


Fwiw, I've looked into javascript validation to try and check the  
content-length client side first, but haven't had success with that.


Thoughts?


Your best bet is probably to get Apache to do this.

Chuck


--

Practical WebObjects - for developers who want to increase their  
overall knowledge of WebObjects or who are trying to solve specific  
problems.

http://www.global-village.net/products/practical_webobjects








--

Practical WebObjects - for developers who want to increase their  
overall knowledge of WebObjects or who are trying to solve specific  
problems.

http://www.global-village.net/products/practical_webobjects





___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: WOComponent children

2008-04-22 Thread Lachlan Deck

On 23/04/2008, at 9:42 AM, Mike Schrag wrote:


On Jan 16, 2008, at 6:09 PM, [EMAIL PROTECTED] wrote:

In parting, I hope you do get that DOM manipulation working and we  
can all benefit from it.


Yours,
-- Aaron
Just to tie a ribbon on this thread (which I've lugged around in my  
inbox for months :) ), on-demand loading of Javascript has finally  
been committed to Ajax.framework.  The nastiest part is actually  
standing on the shoulders of a patch that will be in Prototype 2.0  
that we brought forward into Ajax.framework, but it appears to work  
in all of the browsers I tested on (including Safari 2.0 which was  
one of the ones that previously had issues with the dom route -- the  
patch is disgusting and clever, though).


Well done!

with regards,
--

Lachlan Deck
___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]


Re: WOWebServiceRegistrar.setSecurityDelegate failure

2008-04-22 Thread Gaastra Dennis - WO Lists

Hi Pierre,

Is there a reason WO 5.4.2 is only available with the iPhone SDK? Is  
it a 5.4.2 beta - not a real 5.4.2? When will 5.4.2 be available for  
general downloading + and Mac OS X Server update? Thanks for your  
answers.


With Kind Regards,

Dennis Gaastra, M.B.A.[sfu.ca], B.Sc.[ubc.ca]
Chief Technology Officer,



On 22-Apr-08, at 3:19 PM, Mr. Pierre Frisch wrote:

Without going into too much details when WebServices for WebObjects  
were created with Axis 1.0 type information in WDSL document was  
optional, WebOjects did not include it. With Axis 1.1 and event more  
for Axis 1.4 it became required for custom types. This was fixed in  
WO 5.4.x unfortunately it took us multiple iterations to cover all  
the cases (and there are many). WebObjects 5.3.x and Axis 1.1 work  
if you do not use any custom types i.e. you do not have your own  
serializer/deserializer. As Anjo said there are some major  
difference in the registration between 5.4.0/5.4.1 and 5.4.2. You  
should use the later. I have written a quick test case and at least  
it does not throw the exception you are seeing.


As I said WO 5.4.2 is available with the iPhone SDK.

Cheers

Pierre
--
Pierre Frisch
[EMAIL PROTECTED]


On Apr 22, 2008, at 13:24, Anjo Krank wrote:

5.4.1 wouldn't even start without serious prodding, there where  
pretty serious bugs in registering a service. In 5.4.2, all the  
issues I have left are in axis itself and not really under Pierre's  
control to fix (which are related to special client and server  
libs). In particular the bug reported lately (axis 1.4 creates refs  
for primitive types).


Cheers, Anjo

Am 22.04.2008 um 22:15 schrieb Mike Schrag:

Pierre, if 5.4.x versions all use Axis 1.4 how does moving to  
5.4.2 help?  Regarding the call to org.apache.xml.security.Init,  
I'm sorry, it's extraneous to this conversation and I didn't mean  
to post it.  I use it for some webservice encryption needs.
Any advice on a somewhat guaranteed stable/working WebObjects/ 
WOLips/Java/Axis environment for non-trivial webservices is very,  
very appreciated.
I have been using 5.3 with web services for several years without  
problems, though this very much depends on what is on the other  
side of the service.  I know Anjo had several issues with  
5.4.0/5.4.1 and web services, though Pierre has fixed all of these  
in 5.4.2 as far as I know.  I have an app that I can run now with  
5.4.2 or 5.3 without any issues, though I don't know about your  
particular case.  I recommended trying 5.3 first because it has  
been out for quite a while and most of the issues are known, or at  
the very least, as Pierre recommends, download 5.4.2 from the  
iPhone SDK and try that.  I wouldn't spend any time trying to  
diagnose this in any environment other than 5.3 or 5.4.2 because I  
think it's probably a waste of time since work was specifically  
done on this area.


ms

___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/krank%40logicunited.com

This email sent to [EMAIL PROTECTED]


___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/pierre%40apple.com

This email sent to [EMAIL PROTECTED]


___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/webobjects_lists%40webappz.com

This email sent to [EMAIL PROTECTED]


 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]

Connection Pooling + WOAdaptor ?

2008-04-22 Thread Shravan Kumar.M

Hello Group,

We launch our WO apps using scripts and the core launch command is as 
follows:


*nohup /MyApps/JuciyAdmin/Juicy -WOPort 8001 -Config JuicyPAdmin 
-IsRMLicInstalled true -Xms64m -Xmx512m -WOSMTPHost 192.168.1.10 
-WOOutputPath /MyApps/logs/ | tee **/MyApps/logs/JuicyAppLog**.8001.log *


We configure WOAdaptor with the above application as follows, 
WOConfig.xml file:


?xml version=1.0 encoding=ASCII?
!DOCTYPE WebObjectsAdaptorConfiguration SYSTEM woadaptor.dtd
adaptor
application name=juicyadmin retries=25 scheduler=ROUNDROBIN 
protocol=HTTP poolsize=6767 sendTimeout=100 recvTimeout=300 
cnctTimeout=100
   instance id=1 port=8001 host=192.168.1.10 cnctTimeout=100 
sendTimeout=100 recvTimeout=300 /

/application

We use Apache 1.3.36, WebObjects 5.2.4 with Java 1.4.2_17 on a Solaris 
5.8 box.


Given this,

*Do we really dependant on WOTaskd process in any way?

How can I determine our WOAdaptor is configured for Connection pooling ?*

Please advise. Thanks in advance.

--
Thank You
Shravan Kumar. M
effigent India Pvt., Ltd.,
--

 ___
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list  (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to [EMAIL PROTECTED]