[HELP]: blocks and domainmodel

2008-03-20 Thread Patrick Heiden
Hello together!

I'm trying to find a 'best' way to integrate my domainmodel/businesslayer into 
cocoon 22 block-semantics. Imagine a rich domainmodel with several 
package-specific services presented to the client as facade (decoupling 
persistence and the like). Development of the domainmodel is kept outside for 
testing-purposes and so on. How should I integrate a 'finished' domainmodel 
into cocoon-22 blocks?

Several blocks use different parts of the model, of course, and the usage is 
triggered through a main-controller-block. Following the discussions within 
this mailing-list so far, I am able to imagine, that I could spread my 
domainmodel to different blocks, but this is not what I want to achieve.
Should there be a 'domain'-block, wich contains all domain-beans/objects and 
from wich other blocks can get access to the domain?

Maybe my problem is, that there is no archetype for such an integrated 
dev-environment (or my thoughts about using the given ones is somewhat 
incomplete). I would like a structure like the following:

myProject
+ domain
   + ...
+ dao
   + ...
+ web
   + cocoon-blocks

Or is it the best way to build the domain totally seperate and add the 
domain.jar as dependency to my blocks?

Any hints possible based on my short description?

Best Greetings
Patrick
-- 
Psst! Geheimtipp: Online Games kostenlos spielen bei den GMX Free Games! 
http://games.entertainment.gmx.net/de/entertainment/games/free

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Andre Juffer

Patrick,

some time ago I raised a similar question:

http://www.mail-archive.com/users@cocoon.apache.org/msg40943.html

The discussion there may be helpful to you as well.

Andre.

Patrick Heiden wrote:

Hello together!

I'm trying to find a 'best' way to integrate my domainmodel/businesslayer into 
cocoon 22 block-semantics. Imagine a rich domainmodel with several 
package-specific services presented to the client as facade (decoupling 
persistence and the like). Development of the domainmodel is kept outside for 
testing-purposes and so on. How should I integrate a 'finished' domainmodel 
into cocoon-22 blocks?

Several blocks use different parts of the model, of course, and the usage is 
triggered through a main-controller-block. Following the discussions within 
this mailing-list so far, I am able to imagine, that I could spread my 
domainmodel to different blocks, but this is not what I want to achieve.
Should there be a 'domain'-block, wich contains all domain-beans/objects and 
from wich other blocks can get access to the domain?

Maybe my problem is, that there is no archetype for such an integrated 
dev-environment (or my thoughts about using the given ones is somewhat 
incomplete). I would like a structure like the following:

myProject
+ domain
   + ...
+ dao
   + ...
+ web
   + cocoon-blocks

Or is it the best way to build the domain totally seperate and add the 
domain.jar as dependency to my blocks?

Any hints possible based on my short description?

Best Greetings
Patrick



--
Andre H. Juffer  | Phone: +358-8-553 1161
The Biocenter and| Fax: +358-8-553-1141
the Dep. of Biochemistry | Email: [EMAIL PROTECTED]
University of Oulu, Finland  | WWW: www.biochem.oulu.fi/Biocomputing/
NordProt | WWW: www.nordprot.org

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Patrick Heiden
Hello Andre!

It definitively is helpfull, thanks a lot!!

Greetings,
Patrick
 Original-Nachricht 
 Datum: Thu, 20 Mar 2008 13:24:29 +0200
 Von: Andre Juffer [EMAIL PROTECTED]
 An: users@cocoon.apache.org
 Betreff: Re: [HELP]: blocks and domainmodel

 Patrick,
 
 some time ago I raised a similar question:
 
 http://www.mail-archive.com/users@cocoon.apache.org/msg40943.html
 
 The discussion there may be helpful to you as well.
 
 Andre.
 
 Patrick Heiden wrote:
  Hello together!
  
  I'm trying to find a 'best' way to integrate my
 domainmodel/businesslayer into cocoon 22 block-semantics. Imagine a rich 
 domainmodel with several
 package-specific services presented to the client as facade (decoupling
 persistence and the like). Development of the domainmodel is kept outside for
 testing-purposes and so on. How should I integrate a 'finished' domainmodel
 into cocoon-22 blocks?
  
  Several blocks use different parts of the model, of course, and the
 usage is triggered through a main-controller-block. Following the discussions
 within this mailing-list so far, I am able to imagine, that I could spread
 my domainmodel to different blocks, but this is not what I want to achieve.
  Should there be a 'domain'-block, wich contains all domain-beans/objects
 and from wich other blocks can get access to the domain?
  
  Maybe my problem is, that there is no archetype for such an integrated
 dev-environment (or my thoughts about using the given ones is somewhat
 incomplete). I would like a structure like the following:
  
  myProject
  + domain
 + ...
  + dao
 + ...
  + web
 + cocoon-blocks
  
  Or is it the best way to build the domain totally seperate and add the
 domain.jar as dependency to my blocks?
  
  Any hints possible based on my short description?
  
  Best Greetings
  Patrick
 
 
 -- 
 Andre H. Juffer  | Phone: +358-8-553 1161
 The Biocenter and| Fax: +358-8-553-1141
  the Dep. of Biochemistry | Email: [EMAIL PROTECTED]
 University of Oulu, Finland  | WWW: www.biochem.oulu.fi/Biocomputing/
 NordProt | WWW: www.nordprot.org
 
 -
 To unsubscribe, e-mail: [EMAIL PROTECTED]
 For additional commands, e-mail: [EMAIL PROTECTED]

-- 
GMX startet ShortView.de. Hier findest Du Leute mit Deinen Interessen!
Jetzt dabei sein: http://www.shortview.de/[EMAIL PROTECTED]

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Patrick Heiden
Hello again!

So following your advised discussion, I am able to create the whole domainmodel 
as standalone .jar and add this as dependency to webapp-block, right? Of course 
I also have to put the domain-appCtx.xml into the META-INF/cocoon/spring 
directory. If I would do so, are all other blocks automatically aware of my 
domain.jar, or do I have to put the dependency to every blocks pom.xml wich is 
going to use it (besides testing!)?

Thank you,
Patrick
 Original-Nachricht 
 Datum: Thu, 20 Mar 2008 13:24:29 +0200
 Von: Andre Juffer [EMAIL PROTECTED]
 An: users@cocoon.apache.org
 Betreff: Re: [HELP]: blocks and domainmodel

 Patrick,
 
 some time ago I raised a similar question:
 
 http://www.mail-archive.com/users@cocoon.apache.org/msg40943.html
 
 The discussion there may be helpful to you as well.
 
 Andre.
 
 Patrick Heiden wrote:
  Hello together!
  
  I'm trying to find a 'best' way to integrate my
 domainmodel/businesslayer into cocoon 22 block-semantics. Imagine a rich 
 domainmodel with several
 package-specific services presented to the client as facade (decoupling
 persistence and the like). Development of the domainmodel is kept outside for
 testing-purposes and so on. How should I integrate a 'finished' domainmodel
 into cocoon-22 blocks?
  
  Several blocks use different parts of the model, of course, and the
 usage is triggered through a main-controller-block. Following the discussions
 within this mailing-list so far, I am able to imagine, that I could spread
 my domainmodel to different blocks, but this is not what I want to achieve.
  Should there be a 'domain'-block, wich contains all domain-beans/objects
 and from wich other blocks can get access to the domain?
  
  Maybe my problem is, that there is no archetype for such an integrated
 dev-environment (or my thoughts about using the given ones is somewhat
 incomplete). I would like a structure like the following:
  
  myProject
  + domain
 + ...
  + dao
 + ...
  + web
 + cocoon-blocks
  
  Or is it the best way to build the domain totally seperate and add the
 domain.jar as dependency to my blocks?
  
  Any hints possible based on my short description?
  
  Best Greetings
  Patrick
 
 
 -- 
 Andre H. Juffer  | Phone: +358-8-553 1161
 The Biocenter and| Fax: +358-8-553-1141
  the Dep. of Biochemistry | Email: [EMAIL PROTECTED]
 University of Oulu, Finland  | WWW: www.biochem.oulu.fi/Biocomputing/
 NordProt | WWW: www.nordprot.org
 
 -
 To unsubscribe, e-mail: [EMAIL PROTECTED]
 For additional commands, e-mail: [EMAIL PROTECTED]

-- 
Psst! Geheimtipp: Online Games kostenlos spielen bei den GMX Free Games! 
http://games.entertainment.gmx.net/de/entertainment/games/free

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Grzegorz Kossakowski

Patrick Heiden pisze:

Hello again!

So following your advised discussion, I am able to create the whole domainmodel 
as standalone
.jar and add this as dependency to webapp-block, right? Of course I also have 
to put the
domain-appCtx.xml into the META-INF/cocoon/spring directory. If I would do so, 
are all other
blocks automatically aware of my domain.jar, or do I have to put the dependency 
to every blocks
pom.xml wich is going to use it (besides testing!)?


Hello Patrick,

As it has been said earlier it will be visible because there is only one Global context and now 
isolation. Anyway, it's just an _implementation detail_ and *not* a design goal. That means you 
should always reflect dependencies of your code, logic in POM files.


Not doing so really breaks design and it's a great opportunity for serious problems to occur. The 
great example is current situation in SSF described by COCOON-2176[1] issue. While adding dependency 
on CocoonSourceResolver as a temporary solution we introduced dependency on cocoon-core but we 
forgot to state this fact in POM. Everything was working fine because we were testing SSF mostly in 
a Cocoon environment where cocoon-core dependency was present due to other artifacts. This lead to 
the situation that we eventually forgot about this obscure dependency on cocoon-core and 
rediscovered it just before final release of SSF.


Not stating dependencies in POM files can lead to the hidden circular dependencies which are even 
more nastier. In short: always, always state all of your dependencies in all of your POMS, period.


[1] https://issues.apache.org/jira/browse/COCOON-2176


--
Grzegorz Kossakowski

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Andre Juffer

Patrick Heiden wrote:

Hello again!

So following your advised discussion, I am able to create the whole domainmodel 
as standalone .jar and add this as dependency to webapp-block, right? Of course 
I also have to put the domain-appCtx.xml into the META-INF/cocoon/spring 
directory. If I would do so, are all other blocks automatically aware of my 
domain.jar, or do I have to put the dependency to every blocks pom.xml wich is 
going to use it (besides testing!)?


It is actually possible to import beans etc from other files.

See 
http://static.springframework.org/spring/docs/2.5.x/reference/beans.html#beans-basics

Section 3.2.2.1. Composing XML-based configuration metadata

You would write something like:

import resource=classpath://services.xml/

The service.xml would then sit in your standalone jar which can 
therefore also be tested outside of cocoon and can also be used in other 
environments than cocoon (as if there is anything else worth 
considering...). The services.xml would define beans, transactions, etc 
in the regular way. I have typically a few of includes.


Other beans you have made yourself that are typical to cocoon would be 
defined in xml files placed in the designated subdirectories 
(META-INF/cocoon/spring.)


You still need to include a dependency in the pom of each block that 
uses your domain jar, because you would need the jar for deployment. The 
issue that remains to be looked into is that a number of blocks may 
require the same setup for transactions. The definitions for this 
probably would need to be defined globally.


Andre.




Thank you,
Patrick
 Original-Nachricht 

Datum: Thu, 20 Mar 2008 13:24:29 +0200
Von: Andre Juffer [EMAIL PROTECTED]
An: users@cocoon.apache.org
Betreff: Re: [HELP]: blocks and domainmodel



Patrick,

some time ago I raised a similar question:

http://www.mail-archive.com/users@cocoon.apache.org/msg40943.html

The discussion there may be helpful to you as well.

Andre.

Patrick Heiden wrote:

Hello together!

I'm trying to find a 'best' way to integrate my

domainmodel/businesslayer into cocoon 22 block-semantics. Imagine a rich 
domainmodel with several
package-specific services presented to the client as facade (decoupling
persistence and the like). Development of the domainmodel is kept outside for
testing-purposes and so on. How should I integrate a 'finished' domainmodel
into cocoon-22 blocks?

Several blocks use different parts of the model, of course, and the

usage is triggered through a main-controller-block. Following the discussions
within this mailing-list so far, I am able to imagine, that I could spread
my domainmodel to different blocks, but this is not what I want to achieve.

Should there be a 'domain'-block, wich contains all domain-beans/objects

and from wich other blocks can get access to the domain?

Maybe my problem is, that there is no archetype for such an integrated

dev-environment (or my thoughts about using the given ones is somewhat
incomplete). I would like a structure like the following:

myProject
+ domain
   + ...
+ dao
   + ...
+ web
   + cocoon-blocks

Or is it the best way to build the domain totally seperate and add the

domain.jar as dependency to my blocks?

Any hints possible based on my short description?

Best Greetings
Patrick


--
Andre H. Juffer  | Phone: +358-8-553 1161
The Biocenter and| Fax: +358-8-553-1141
 the Dep. of Biochemistry | Email: [EMAIL PROTECTED]
University of Oulu, Finland  | WWW: www.biochem.oulu.fi/Biocomputing/
NordProt | WWW: www.nordprot.org

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





--
Andre H. Juffer  | Phone: +358-8-553 1161
The Biocenter and| Fax: +358-8-553-1141
the Dep. of Biochemistry | Email: [EMAIL PROTECTED]
University of Oulu, Finland  | WWW: www.biochem.oulu.fi/Biocomputing/
NordProt | WWW: www.nordprot.org

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Patrick Heiden
Hi Grzegorz,

I've just tested the approach of seperating domain/cocoon-blocks and it works 
fine. I will follow your suggestion and put a dependency in every block wich 
will use the domain-model (partly). Thank you for that!

One other issue: I am wondering a little bit, why a call to 
webbAppCtx.containsBeanDefinition(def) returns null. I am able to use the 
bean through cocoon.getComponent(def). I guess this is one of the 
implementation-details again ;)

Best Greetings,
Patrick
 Original-Nachricht 
 Datum: Thu, 20 Mar 2008 13:12:34 +0100
 Von: Grzegorz Kossakowski [EMAIL PROTECTED]
 An: users@cocoon.apache.org
 Betreff: Re: [HELP]: blocks and domainmodel

 Patrick Heiden pisze:
  Hello again!
  
  So following your advised discussion, I am able to create the whole
 domainmodel as standalone
  .jar and add this as dependency to webapp-block, right? Of course I also
 have to put the
  domain-appCtx.xml into the META-INF/cocoon/spring directory. If I would
 do so, are all other
  blocks automatically aware of my domain.jar, or do I have to put the
 dependency to every blocks
  pom.xml wich is going to use it (besides testing!)?
 
 Hello Patrick,
 
 As it has been said earlier it will be visible because there is only one
 Global context and now 
 isolation. Anyway, it's just an _implementation detail_ and *not* a design
 goal. That means you 
 should always reflect dependencies of your code, logic in POM files.
 
 Not doing so really breaks design and it's a great opportunity for serious
 problems to occur. The 
 great example is current situation in SSF described by COCOON-2176[1]
 issue. While adding dependency 
 on CocoonSourceResolver as a temporary solution we introduced dependency
 on cocoon-core but we 
 forgot to state this fact in POM. Everything was working fine because we
 were testing SSF mostly in 
 a Cocoon environment where cocoon-core dependency was present due to other
 artifacts. This lead to 
 the situation that we eventually forgot about this obscure dependency on
 cocoon-core and 
 rediscovered it just before final release of SSF.
 
 Not stating dependencies in POM files can lead to the hidden circular
 dependencies which are even 
 more nastier. In short: always, always state all of your dependencies in
 all of your POMS, period.
 
 [1] https://issues.apache.org/jira/browse/COCOON-2176
 
 
 -- 
 Grzegorz Kossakowski
 
 -
 To unsubscribe, e-mail: [EMAIL PROTECTED]
 For additional commands, e-mail: [EMAIL PROTECTED]

-- 
GMX startet ShortView.de. Hier findest Du Leute mit Deinen Interessen!
Jetzt dabei sein: http://www.shortview.de/[EMAIL PROTECTED]

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Patrick Heiden
Hi Andre!

snip/

 It is actually possible to import beans etc from other files.
 
 See 
 http://static.springframework.org/spring/docs/2.5.x/reference/beans.html#beans-basics
 Section 3.2.2.1. Composing XML-based configuration metadata
 
 You would write something like:
 
 import resource=classpath://services.xml/

Do you mean that this should be put inside 
META-INF/cocoon/spring/someContext.xml? That is what would make sense to me.

 
 The service.xml would then sit in your standalone jar which can 
 therefore also be tested outside of cocoon and can also be used in other 
 environments than cocoon (as if there is anything else worth 
 considering...). The services.xml would define beans, transactions, etc 
 in the regular way. I have typically a few of includes.
 
 Other beans you have made yourself that are typical to cocoon would be 
 defined in xml files placed in the designated subdirectories 
 (META-INF/cocoon/spring.)
 
 You still need to include a dependency in the pom of each block that 
 uses your domain jar, because you would need the jar for deployment. The 
 issue that remains to be looked into is that a number of blocks may 
 require the same setup for transactions. The definitions for this 
 probably would need to be defined globally.

This is a bit condusing to me, because I would define my transactions inside my 
domain-model-appCtx and let cocoon-controller (e.g. flowscript) transparently 
use them. So there already are 'global' defined transactions.

Greetings,
Patrick
-- 
GMX startet ShortView.de. Hier findest Du Leute mit Deinen Interessen!
Jetzt dabei sein: http://www.shortview.de/[EMAIL PROTECTED]

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Andre Juffer

Patrick Heiden wrote:

Hi Andre!

snip/


It is actually possible to import beans etc from other files.

See 
http://static.springframework.org/spring/docs/2.5.x/reference/beans.html#beans-basics

Section 3.2.2.1. Composing XML-based configuration metadata

You would write something like:

import resource=classpath://services.xml/


Do you mean that this should be put inside 
META-INF/cocoon/spring/someContext.xml? That is what would make sense to me.


Yes, exactly.



The service.xml would then sit in your standalone jar which can 
therefore also be tested outside of cocoon and can also be used in other 
environments than cocoon (as if there is anything else worth 
considering...). The services.xml would define beans, transactions, etc 
in the regular way. I have typically a few of includes.


Other beans you have made yourself that are typical to cocoon would be 
defined in xml files placed in the designated subdirectories 
(META-INF/cocoon/spring.)


You still need to include a dependency in the pom of each block that 
uses your domain jar, because you would need the jar for deployment. The 
issue that remains to be looked into is that a number of blocks may 
require the same setup for transactions. The definitions for this 
probably would need to be defined globally.


This is a bit condusing to me, because I would define my transactions inside my 
domain-model-appCtx and let cocoon-controller (e.g. flowscript) transparently 
use them. So there already are 'global' defined transactions.


Sorry, I was apparently not clear. I use facades that sit on top of the 
domain layer. My flow scripts just deal with these facades. These 
facades are made transaction aware. You do exactly the same. Currently, 
I have just one spring conf file (in the jar) that defines the 
transaction management for the domain. I am just not entirely clear 
whether the transactions are truly global or just local within a block. 
If they are local than I would need to include the same transaction 
definitions in a number of blocks. If they are truly global, than just 
one block would need to include the transactions, I presume.




Greetings,
Patrick



--
Andre H. Juffer  | Phone: +358-8-553 1161
The Biocenter and| Fax: +358-8-553-1141
the Dep. of Biochemistry | Email: [EMAIL PROTECTED]
University of Oulu, Finland  | WWW: www.biochem.oulu.fi/Biocomputing/
NordProt | WWW: www.nordprot.org

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Patrick Heiden
Hello!

snip/

 Sorry, I was apparently not clear. I use facades that sit on top of the 
 domain layer. My flow scripts just deal with these facades. These 
 facades are made transaction aware. You do exactly the same. Currently, 
 I have just one spring conf file (in the jar) that defines the 
 transaction management for the domain. I am just not entirely clear 
 whether the transactions are truly global or just local within a block. 
 If they are local than I would need to include the same transaction 
 definitions in a number of blocks. If they are truly global, than just 
 one block would need to include the transactions, I presume.

Following what Grzegorz mentioned earlier, I would say, that transaction 
definitions within domain-app-ctx are global for all blocks as long as one put 
a dependency to the domain.jar within the blocks pom.xml. Then it is just up to 
the blocks using domain.jar to include 'your' statement to 
META-INF/cocoon/spring/appCtx.xml to tell that there is another appCtx in the 
classpath. Actually there is only one global webAppCtx, including all 
sub-contexts and there should be no problem with transactions. Or do I miss 
something? I am not sure in wich cases there should be other transaction-rules 
within some blocks, because I only want to use transactional behaviour out of 
my domainmodel and never want to let cocoon do any DB-stuff itself.

Greetings,
Patrick
-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser

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



RE: [HELP]: blocks and domainmodel

2008-03-20 Thread Andre Juffer
Hi Patrick,

Yeah, I guess that this would settle the matter then. One transaction
definition file to transact them all...

Cheers,
Andre


-Original Message-
From: Patrick Heiden [mailto:[EMAIL PROTECTED] 
Sent: Thursday, March 20, 2008 16:54
To: users@cocoon.apache.org
Subject: Re: [HELP]: blocks and domainmodel

Hello!

snip/

 Sorry, I was apparently not clear. I use facades that sit on top of the 
 domain layer. My flow scripts just deal with these facades. These 
 facades are made transaction aware. You do exactly the same. Currently, 
 I have just one spring conf file (in the jar) that defines the 
 transaction management for the domain. I am just not entirely clear 
 whether the transactions are truly global or just local within a block. 
 If they are local than I would need to include the same transaction 
 definitions in a number of blocks. If they are truly global, than just 
 one block would need to include the transactions, I presume.

Following what Grzegorz mentioned earlier, I would say, that transaction
definitions within domain-app-ctx are global for all blocks as long as one
put a dependency to the domain.jar within the blocks pom.xml. Then it is
just up to the blocks using domain.jar to include 'your' statement to
META-INF/cocoon/spring/appCtx.xml to tell that there is another appCtx in
the classpath. Actually there is only one global webAppCtx, including all
sub-contexts and there should be no problem with transactions. Or do I miss
something? I am not sure in wich cases there should be other
transaction-rules within some blocks, because I only want to use
transactional behaviour out of my domainmodel and never want to let cocoon
do any DB-stuff itself.

Greetings,
Patrick
-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser

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



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



Re: RE: [HELP]: blocks and domainmodel

2008-03-20 Thread Patrick Heiden
Hi Andre!
 
 Yeah, I guess that this would settle the matter then. One transaction
 definition file to transact them all...

Maybe one last gap in addition to your thought about block-specific 
transaction-management:

If one must somehow define special transaction purposes, e.g. different 
method-name-patterns, then usage of different transaction-managers could be 
fine. But I would try to avoid such design and spend a bit more effort upfront.

Cheers ;)
Patrick
-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser

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



Re: [HELP]: blocks and domainmodel

2008-03-20 Thread Andre Juffer

Patrick Heiden wrote:

Hi Andre!
 

Yeah, I guess that this would settle the matter then. One transaction
definition file to transact them all...


Maybe one last gap in addition to your thought about block-specific 
transaction-management:

If one must somehow define special transaction purposes, e.g. different 
method-name-patterns, then usage of different transaction-managers could be 
fine. But I would try to avoid such design and spend a bit more effort upfront.


Agreed. My transactions are also very straightforward. I do not use 
different method-name pattern, except to differentiate between methods 
that just get something from the database without actually modifying it 
and methods that result in a modification of one or more tables in the 
database. That's it. That's why I have just one transaction.xml.


Thanks,
Andre




Cheers ;)
Patrick



--
Andre H. Juffer  | Email: [EMAIL PROTECTED]
The Biocenter and| WWW: www.biochem.oulu.fi/Biocomputing/
the Dep. of Biochemistry | Fax: +358-8-553-1141
University of Oulu, Finland  | Phone: +358-8-553 1161
Triacle Biocomputing | WWW: www.triacle-bc.com

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