hibernate-devel — development

[Hibernate] [ hibernate-Patches-645169 ] codegen enhancements

[SourceForge.net <no...@so...>]

Patches item #645169, was opened at 2002-11-28 10:24
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428710&aid=645169&group_id=40712

Category: Codegen
Group: None
Status: Closed
Resolution: Out of Date
Priority: 3
Submitted By: Max R. Andersen (maxcsaucdk)
Assigned to: Nobody/Anonymous (nobody)
Summary: codegen enhancements

Initial Comment:
The following patch provides two things (sorry - 
had to do both to make it work in my reallife 
project :)




1. Codegen now generates getter/setters for 
version and timestamps (it did not do it before!) 
(a 2 line change)




2. Support for codegenname in the definition of 
class/subclasses.




3. Removed an unused and redundant paremter 
in the Render interface.




#2 makes it possible to do the following:




<class name="Person" 
codegenname="AutoPerson">


...


...




Where the codegen attribute (if provided) will be 
used by the codegenerator instead of the name 
attribute.




This makes it possible to "extend" the 
codegenerated code - and still have the option 
of "regenerating" code without messing with your 
own "extensions"




so you could have:




class Person extends AutoPerson {


 


 public Object getNonPersistedProperty() {


     return "whatever I want":


  }




}




This improves hibernates round-trip abilities! :)


----------------------------------------------------------------------

>Comment By: Max R. Andersen (maxcsaucdk)
Date: 2003-01-06 09:30

Message:
Logged In: YES 
user_id=18119

Yes - it is obsolete.


----------------------------------------------------------------------

Comment By: Gavin King (oneovthafew)
Date: 2003-01-06 07:33

Message:
Logged In: YES 
user_id=384580

This patch is now obsolete, I assume?

----------------------------------------------------------------------

Comment By: Gavin King (oneovthafew)
Date: 2002-12-06 00:47

Message:
Logged In: YES 
user_id=384580

Cool. I will integrate this for 1.2.1. Trying to make
as-minimal-as-possible changes before 1.2final.

Thanks :)

----------------------------------------------------------------------

Comment By: Max R. Andersen (maxcsaucdk)
Date: 2002-12-05 10:28

Message:
Logged In: YES 
user_id=18119

and now the patch :)


----------------------------------------------------------------------

Comment By: Max R. Andersen (maxcsaucdk)
Date: 2002-12-05 10:27

Message:
Logged In: YES 
user_id=18119

Ok - now a patch that fixes ALL problems - ah, 
almost :)




This time the codegen get's its information from


<metaattribute name="Codegen.Classname" 
value="something"/>




And now it supports/requires a fully qualified name.




Currently it is only the codegenerator that is aware of 
metaattribute tag.




It works for normal classes, and should work for 
subclass'es also.




Is this good enough for inclusion now ? :)


----------------------------------------------------------------------

Comment By: Max R. Andersen (maxcsaucdk)
Date: 2002-11-28 14:06

Message:
Logged In: YES 
user_id=18119

btw. the current implementation only support that the 
codegenname is in the same package as the "real" 
one. 




Supporting another packagename requires some 
more cumbersome work :(


----------------------------------------------------------------------

Comment By: Max R. Andersen (maxcsaucdk)
Date: 2002-11-28 14:05

Message:
Logged In: YES 
user_id=18119

Fine for me - maybe call it metaattribute/tag/info to 
emphasize it has nothing to do with 
attributes/properties ? 




It would make "future" codegen enhancments easier 
- and maybe even make it availiable thorugh the 
metadata api ?!!? :)


----------------------------------------------------------------------

Comment By: Gavin King (oneovthafew)
Date: 2002-11-28 13:42

Message:
Logged In: YES 
user_id=384580

I have committed the version/timestamp stuff.

As to the second part, rather than adding a new attribute to
<class>, can we instead add a new metadata element like:

<attribute name="CodeGenerator.className"
value="eg.AbstractFoo"/>

What do you think?

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428710&aid=645169&group_id=40712

Re: [Hibernate] Suggestion: Encapsulate SQLException in HibernateException

[Jozsa K. <dy...@on...>]

On Sun, Jan 05, 2003 at 06:39:32PM -1000, Mark Woon wrote:
> Gavin King wrote:
> 
> >Has this thread died?
> >
> >I still havn't got a really good feeling for what other people
> >think about this issue.... 
> >
> 
> +1, would much prefer SQLException be nested within HibernateException.

+1 here too (if that counts :). I'd prefer this way to the current way of
handling exceptions. Cleaner, easier, makes sense for me.

Christopher
-- 
.Digital.Yearning.for.Networked.Assassination.and.Xenocide

Re: [Hibernate] Suggestion: Encapsulate SQLException in HibernateException

[Christian B. <chr...@bl...>]

On 06 Jan (14:09), Gavin King wrote:

> Has this thread died?
> 
> I still havn't got a really good feeling for what other people
> think about this issue.... 

I think every distinct layer in a system should throw only one Exception
type, possibly wrapping another Exception (this is called Exception
Chaining, standard with JDK 1.4 now). 

So, wrapping an SQLException in a HibernateException is OK. 

> > I was thinking about wrapping SQLException in 
> > HibernateException, so that we don't need to worry about 
> > catch HibernateException and SQLException. Since now the work 
> > is focused on some changes for Hibernate 2, this could be the 
> > right time to do this kind of change. This would simplify the 
> > error handling in the applications.

-- 
Christian Bauer
tu...@in...

Re: [Hibernate] FW: [hibernate - Open Discussion] first crack at maven'ization

[Christian B. <chr...@bl...>]

On 06 Jan (13:25), Gavin King wrote:

> Check this out :)

I evaluated project tools like Centipede and Maven half a year ago for
Hibernate.

You see, Maven is primarily used in Apache Group style projects. The
amazing thing is: The goal of Maven is to unify the development process
so a project can provide quality information/documentation. 

Aside from some technical issues, one thing I don't like about Apache
Group Jakarta projects is the technical documentation! I hate these
"Still nothing here" pages you'll find often in Struts, OJB, or the
evil, completely undocumented Commons stuff.

I don't see where the benefit of using Maven is and certainly don't want
to follow the Apache Group road. Is there something wrong with the
current build process?

-- 
Christian Bauer
tu...@in...

[Hibernate] [ hibernate-Patches-645169 ] codegen enhancements

[SourceForge.net <no...@so...>]

Patches item #645169, was opened at 2002-11-28 20:24
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428710&aid=645169&group_id=40712

Category: Codegen
Group: None
>Status: Closed
>Resolution: Out of Date
Priority: 3
Submitted By: Max R. Andersen (maxcsaucdk)
Assigned to: Nobody/Anonymous (nobody)
Summary: codegen enhancements

Initial Comment:
The following patch provides two things (sorry - 
had to do both to make it work in my reallife 
project :)




1. Codegen now generates getter/setters for 
version and timestamps (it did not do it before!) 
(a 2 line change)




2. Support for codegenname in the definition of 
class/subclasses.




3. Removed an unused and redundant paremter 
in the Render interface.




#2 makes it possible to do the following:




<class name="Person" 
codegenname="AutoPerson">


...


...




Where the codegen attribute (if provided) will be 
used by the codegenerator instead of the name 
attribute.




This makes it possible to "extend" the 
codegenerated code - and still have the option 
of "regenerating" code without messing with your 
own "extensions"




so you could have:




class Person extends AutoPerson {


 


 public Object getNonPersistedProperty() {


     return "whatever I want":


  }




}




This improves hibernates round-trip abilities! :)


----------------------------------------------------------------------

>Comment By: Gavin King (oneovthafew)
Date: 2003-01-06 17:33

Message:
Logged In: YES 
user_id=384580

This patch is now obsolete, I assume?

----------------------------------------------------------------------

Comment By: Gavin King (oneovthafew)
Date: 2002-12-06 10:47

Message:
Logged In: YES 
user_id=384580

Cool. I will integrate this for 1.2.1. Trying to make
as-minimal-as-possible changes before 1.2final.

Thanks :)

----------------------------------------------------------------------

Comment By: Max R. Andersen (maxcsaucdk)
Date: 2002-12-05 20:28

Message:
Logged In: YES 
user_id=18119

and now the patch :)


----------------------------------------------------------------------

Comment By: Max R. Andersen (maxcsaucdk)
Date: 2002-12-05 20:27

Message:
Logged In: YES 
user_id=18119

Ok - now a patch that fixes ALL problems - ah, 
almost :)




This time the codegen get's its information from


<metaattribute name="Codegen.Classname" 
value="something"/>




And now it supports/requires a fully qualified name.




Currently it is only the codegenerator that is aware of 
metaattribute tag.




It works for normal classes, and should work for 
subclass'es also.




Is this good enough for inclusion now ? :)


----------------------------------------------------------------------

Comment By: Max R. Andersen (maxcsaucdk)
Date: 2002-11-29 00:06

Message:
Logged In: YES 
user_id=18119

btw. the current implementation only support that the 
codegenname is in the same package as the "real" 
one. 




Supporting another packagename requires some 
more cumbersome work :(


----------------------------------------------------------------------

Comment By: Max R. Andersen (maxcsaucdk)
Date: 2002-11-29 00:05

Message:
Logged In: YES 
user_id=18119

Fine for me - maybe call it metaattribute/tag/info to 
emphasize it has nothing to do with 
attributes/properties ? 




It would make "future" codegen enhancments easier 
- and maybe even make it availiable thorugh the 
metadata api ?!!? :)


----------------------------------------------------------------------

Comment By: Gavin King (oneovthafew)
Date: 2002-11-28 23:42

Message:
Logged In: YES 
user_id=384580

I have committed the version/timestamp stuff.

As to the second part, rather than adding a new attribute to
<class>, can we instead add a new metadata element like:

<attribute name="CodeGenerator.className"
value="eg.AbstractFoo"/>

What do you think?

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428710&aid=645169&group_id=40712

Re: [Hibernate] Suggestion: Encapsulate SQLException in HibernateException

[Mark W. <mor...@SM...>]

Gavin King wrote:

>Has this thread died?
>
>I still havn't got a really good feeling for what other people
>think about this issue.... 
>

+1, would much prefer SQLException be nested within HibernateException.

-Mark

RE: [Hibernate] Suggestion: Encapsulate SQLException in HibernateException

[Gavin K. <Gav...@ex...>]

This is exactly the kind of input I was looking for :) And your point
is quite correct; even if we *did* rationalize the Session interface
to throw only HibernateExceptions, in general, we *still* wouldn't be=20
able to do this:

Session s =3D sf.openSession()
try {
    // do some work
}
catch (HibernateException he) {
   // yada yada yada
}
finally {
   s.close();
}


Why? because "yada yada yada" needs to happed no matter *what*
kind of exception occurred (even RuntimeExceptions).

However, there is another argument for this change, that no-one has
really brought up. Nested exceptions let us add some contextual
information to the wrapped SQLException. So the developer could
see what Hibernate was doing when the SQLException occurred.
At present you see a stack trace which is somewhat helpful but
probably less helpful than what we could do with nesting.



> This is just my two cents here, and I'm still relatively new to those=20
> parts, but here goes:
>=20
> I've been doing development on database backed persistence stuff, and=20
> after a brief unfullfilling fling with Castor, I ditched that=20
> and went=20
> with Hibernate and lets just say that this thing is a whole=20
> lot better.
>=20
> But I digress.  Regarding SQLExceptions, I wrap everything up in the=20
> persistence layer with it's own exception, so theres just one=20
> exception=20
> to throw in the application when accessing the data, but in the=20
> persitence class file and with Hibernate, I found myself having to=20
> catch 4-5 different types of exceptions when doing a hibernate query,=20
> and I eventually just gave up on that and just put them in a vanilla=20
> Exception try-catch block and sort it out afterwards.
>=20
> I guess my point here is that in my case, I'm already catching it all=20
> with Exception so it doesn't make much difference to me=20
> because of all=20
> the other exceptions I have to catch anyway compels me to use this. =20
> But if the exceptions could be consolidated to just one or two types,=20
> then i would see a benefit to catching them individually.
>=20
> I don't know how others do it, they may have a better=20
> solution or know=20
> better than I do.  But you wanted feelings, man!  :)
>=20
> -tim
>=20
> On Sunday, January 5, 2003, at 10:09 PM, Gavin King wrote:
>=20
> > Has this thread died?
> >
> > I still havn't got a really good feeling for what other=20
> people think=20
> > about this issue....
> >
> >> -----Original Message-----
> >> From: rmi...@po... [mailto:rmi...@po...]
> >> Sent: Saturday, 4 January 2003 1:46 AM
> >> To: hib...@li...
> >> Subject: [Hibernate] Suggestion: Encapsulate SQLException in=20
> >> HibernateException
> >>
> >>
> >> Hi,
> >>
> >> I was thinking about wrapping SQLException in=20
> HibernateException, so=20
> >> that we don't need to worry about catch HibernateException and=20
> >> SQLException. Since now the work is focused on some changes for=20
> >> Hibernate 2, this could be the right time to do this kind=20
> of change.=20
> >> This would simplify the error handling in the applications.
> >>
> >> Robson
> >>
> >>
> >>
> >> -------------------------------------------------------
> >> This sf.net email is sponsored by:ThinkGeek
> >> Welcome to geek heaven.
> >> http://thinkgeek.com/sf=20
> >> _______________________________________________
> >> hibernate-devel mailing list hib...@li...
> >> https://lists.sourceforge.net/lists/listinfo/hibernate-devel
> >>
> >
> >
> > ********** CAUTION - Disclaimer **********
> > This message may contain privileged and confidential=20
> information. If=20
> > you are not the intended recipient of this message (or=20
> responsible for=20
> > delivery of the message to such person) you are hereby=20
> notified that=20
> > any use, dissemination, distribution or reproduction of this message
> > is prohibited. If you have received this message in error,
> > you should destroy it and kindly notify the sender by reply
> > e-mail. Please advise immediately if you or your employer
> > do not consent to Internet e-mail for messages of this kind.
> > Opinions, conclusions and other information in this
> > message that do not relate to the official business of
> > Expert Information Services Pty Ltd ("The Company")
> > shall be understood as neither given nor endorsed by it.
> >
> > The Company advises that this e-mail and any attached
> > files should be scanned to detect viruses. The Company accepts no=20
> > liability for loss or damage (whether caused by negligence or not)=20
> > resulting from the use of any attached files.
> > **EIS******** End of Disclaimer **********
> >
> >
> >
> > -------------------------------------------------------
> > This sf.net email is sponsored by:ThinkGeek
> > Welcome to geek heaven.
> > http://thinkgeek.com/sf=20
> > _______________________________________________
> > hibernate-devel mailing list hib...@li...
> > https://lists.sourceforge.net/lists/listinfo/hibernate-devel
> >
>=20
>=20

[Hibernate] [ hibernate-Feature Requests-650587 ] unsaved-value for <property>

[SourceForge.net <no...@so...>]

Feature Requests item #650587, was opened at 2002-12-09 09:38
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=650587&group_id=40712

Category: None
Group: None
Status: Open
Priority: 5
Submitted By: Gavin King (oneovthafew)
Assigned to: Nobody/Anonymous (nobody)
Summary: unsaved-value for <property>

Initial Comment:
Currently only the <id> and <composite-id> elements
permit the unsaved-value attribute. There are a couple
of circumstances where that is *not* convenient;
particularly in the case of assigned identifiers. So we
should generalize the model and allow unsaved-value for
 <property> or <key-property>.

----------------------------------------------------------------------

>Comment By: Gavin King (oneovthafew)
Date: 2003-01-06 14:35

Message:
Logged In: YES 
user_id=384580

yep, thats the idea. Its basically pretty easy to implement
this . Start by changing ClassPersister.isUnsaved() to have
the following signature:

public boolean isUnsaved(Serializable id, Object object)

That way the persister can also interrogate the object's
version property....

Do you want to create a patch against the Hibernate 2 source
tree?

----------------------------------------------------------------------

Comment By: Mark Woon (unkyaku)
Date: 2003-01-06 14:19

Message:
Logged In: YES 
user_id=435832

Gavin, this would potentially solve the cascade prolem for
objects that have an assigned identifiers.  It would be
great to be able to specify unsaved-value on either
<version> or <timestamp> instead of <id> for this purpose.


----------------------------------------------------------------------

Comment By: Gavin King (oneovthafew)
Date: 2003-01-06 12:54

Message:
Logged In: YES 
user_id=384580

This would be *particularly* useful for the <version> and
<timestamp> elements; ie.

<version name="version" unsaved-value="0"/>
<timestamp name="lastUpdated" unsaved-value="null"/>


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=650587&group_id=40712

[Hibernate] [ hibernate-Feature Requests-650587 ] unsaved-value for <property>

[SourceForge.net <no...@so...>]

Feature Requests item #650587, was opened at 2002-12-08 14:38
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=650587&group_id=40712

Category: None
Group: None
Status: Open
Priority: 5
Submitted By: Gavin King (oneovthafew)
Assigned to: Nobody/Anonymous (nobody)
Summary: unsaved-value for <property>

Initial Comment:
Currently only the <id> and <composite-id> elements
permit the unsaved-value attribute. There are a couple
of circumstances where that is *not* convenient;
particularly in the case of assigned identifiers. So we
should generalize the model and allow unsaved-value for
 <property> or <key-property>.

----------------------------------------------------------------------

Comment By: Mark Woon (unkyaku)
Date: 2003-01-05 19:19

Message:
Logged In: YES 
user_id=435832

Gavin, this would potentially solve the cascade prolem for
objects that have an assigned identifiers.  It would be
great to be able to specify unsaved-value on either
<version> or <timestamp> instead of <id> for this purpose.


----------------------------------------------------------------------

Comment By: Gavin King (oneovthafew)
Date: 2003-01-05 17:54

Message:
Logged In: YES 
user_id=384580

This would be *particularly* useful for the <version> and
<timestamp> elements; ie.

<version name="version" unsaved-value="0"/>
<timestamp name="lastUpdated" unsaved-value="null"/>


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=650587&group_id=40712

RE: [Hibernate] Suggestion: Encapsulate SQLException in HibernateException

[Gavin K. <Gav...@ex...>]

Has this thread died?

I still havn't got a really good feeling for what other people
think about this issue....=20

> -----Original Message-----
> From: rmi...@po... [mailto:rmi...@po...]=20
> Sent: Saturday, 4 January 2003 1:46 AM
> To: hib...@li...
> Subject: [Hibernate] Suggestion: Encapsulate SQLException in=20
> HibernateException
>=20
>=20
> Hi,
>=20
> I was thinking about wrapping SQLException in=20
> HibernateException, so that we don't need to worry about=20
> catch HibernateException and SQLException. Since now the work=20
> is focused on some changes for Hibernate 2, this could be the=20
> right time to do this kind of change. This would simplify the=20
> error handling in the applications.
>=20
> Robson
>=20
>=20
>=20
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf=20
> _______________________________________________
> hibernate-devel mailing list hib...@li...
> https://lists.sourceforge.net/lists/listinfo/hibernate-devel
>=20


********** CAUTION - Disclaimer **********
This message may contain privileged and confidential
information. If you are not the intended recipient of this
message (or responsible for delivery of the message to
such person) you are hereby notified that any use,
dissemination, distribution or reproduction of this message
is prohibited. If you have received this message in error,
you should destroy it and kindly notify the sender by reply
e-mail. Please advise immediately if you or your employer
do not consent to Internet e-mail for messages of this kind.
Opinions, conclusions and other information in this
message that do not relate to the official business of
Expert Information Services Pty Ltd ("The Company")
shall be understood as neither given nor endorsed by it.

The Company advises that this e-mail and any attached
files should be scanned to detect viruses. The Company
accepts no liability for loss or damage (whether caused
by negligence or not) resulting from the use of any
attached files.
**EIS******** End of Disclaimer **********

[Hibernate] FW: [hibernate - Open Discussion] first crack at maven'ization

[Gavin K. <Gav...@ex...>]

Check this out :)

-----Original Message-----
From: SourceForge.net [mailto:no...@so...]=20
Sent: Monday, 6 January 2003 1:08 PM
To: no...@so...
Subject: [hibernate - Open Discussion] first crack at maven'ization



Read and respond to this message at:=20
https://sourceforge.net/forum/message.php?msg_id=3D1819063
By: savaki

I've got a first crack of mavenized version of hibernate.  How do I send
this to you?

Here's the project site that it currently generates:

http://www.indigoegg.com/hibernate/

The unit test results were based on running hibernate 1.2.1 against
MySQL 4.0.5

Cheers!


______________________________________________________________________
You are receiving this email because you elected to monitor this forum.
To stop monitoring this forum, login to SourceForge.net and visit:=20
https://sourceforge.net/forum/monitor.php?forum_id=3D128638


********** CAUTION - Disclaimer **********
This message may contain privileged and confidential
information. If you are not the intended recipient of this
message (or responsible for delivery of the message to
such person) you are hereby notified that any use,
dissemination, distribution or reproduction of this message
is prohibited. If you have received this message in error,
you should destroy it and kindly notify the sender by reply
e-mail. Please advise immediately if you or your employer
do not consent to Internet e-mail for messages of this kind.
Opinions, conclusions and other information in this
message that do not relate to the official business of
Expert Information Services Pty Ltd ("The Company")
shall be understood as neither given nor endorsed by it.

The Company advises that this e-mail and any attached
files should be scanned to detect viruses. The Company
accepts no liability for loss or damage (whether caused
by negligence or not) resulting from the use of any
attached files.
**EIS******** End of Disclaimer **********

[Hibernate] [ hibernate-Feature Requests-635016 ] lock() a transient object

[SourceForge.net <no...@so...>]

Feature Requests item #635016, was opened at 2002-11-08 02:26
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=635016&group_id=40712

Category: None
Group: None
Status: Open
>Priority: 1
Submitted By: Gavin King (oneovthafew)
Assigned to: Nobody/Anonymous (nobody)
Summary: lock() a transient object

Initial Comment:
It would be very useful to be able to associate a
transient object with the session (obtaining an
appropriate lock-level by, for example, a version
check). Currently the only way to do this is to use
update(). That would be overkill if we aren't updating
the object.

----------------------------------------------------------------------

>Comment By: Gavin King (oneovthafew)
Date: 2003-01-06 12:55

Message:
Logged In: YES 
user_id=384580

Note that there ARE some conceptual difficulties with this.
Its most likely an undesirable feature.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=635016&group_id=40712

[Hibernate] [ hibernate-Feature Requests-650587 ] unsaved-value for <property>

[SourceForge.net <no...@so...>]

Feature Requests item #650587, was opened at 2002-12-09 09:38
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=650587&group_id=40712

Category: None
Group: None
Status: Open
>Priority: 5
Submitted By: Gavin King (oneovthafew)
Assigned to: Nobody/Anonymous (nobody)
Summary: unsaved-value for <property>

Initial Comment:
Currently only the <id> and <composite-id> elements
permit the unsaved-value attribute. There are a couple
of circumstances where that is *not* convenient;
particularly in the case of assigned identifiers. So we
should generalize the model and allow unsaved-value for
 <property> or <key-property>.

----------------------------------------------------------------------

>Comment By: Gavin King (oneovthafew)
Date: 2003-01-06 12:54

Message:
Logged In: YES 
user_id=384580

This would be *particularly* useful for the <version> and
<timestamp> elements; ie.

<version name="version" unsaved-value="0"/>
<timestamp name="lastUpdated" unsaved-value="null"/>


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=650587&group_id=40712

[Hibernate] [ hibernate-Feature Requests-612053 ] lockAll(), checkVersion()

[SourceForge.net <no...@so...>]

Feature Requests item #612053, was opened at 2002-09-20 19:57
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=612053&group_id=40712

Category: None
Group: None
>Status: Closed
Priority: 2
Submitted By: Gavin King (oneovthafew)
Assigned to: Gavin King (oneovthafew)
Summary: lockAll(), checkVersion()

Initial Comment:
* For the sake of performance it would be nice to be
able to pessimistically lock multiple objects at a time.

* It seems reasonable to allow a version check without
a pessimistic lock.

Perhaps a new Locking interface would make sense rather
than polluting Session. Or perhaps this stuff could be
added to Transaction.

----------------------------------------------------------------------

>Comment By: Gavin King (oneovthafew)
Date: 2003-01-06 12:50

Message:
Logged In: YES 
user_id=384580

Version checking has been possible for a while, via

session.lock(object, LockMode.READ)

There hasn't been much user interest in a lockAll() method.

----------------------------------------------------------------------

Comment By: Gavin King (oneovthafew)
Date: 2002-10-05 03:46

Message:
Logged In: YES 
user_id=384580

The locking API is now complete, excepting the lockAll() method.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=612053&group_id=40712

[Hibernate] [ hibernate-Feature Requests-567274 ] byte[] identifier, etc.

[SourceForge.net <no...@so...>]

Feature Requests item #567274, was opened at 2002-06-11 16:12
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=567274&group_id=40712

Category: None
Group: None
>Status: Closed
Priority: 3
Submitted By: Gavin King (oneovthafew)
Assigned to: Nobody/Anonymous (nobody)
Summary: byte[] identifier, etc.

Initial Comment:

There is a fairly straightforward change that could be 
made which would allow binary type identifiers and 
composite key classes that do not override 
equals/hashCode. Instead, the session could use the 
identifier Type to perform equals/hashCode.

----------------------------------------------------------------------

>Comment By: Gavin King (oneovthafew)
Date: 2003-01-06 12:44

Message:
Logged In: YES 
user_id=384580

Its unlikely that this feature will ever be needed. A
similar effect can be achieved with a "custom type
identifier" pattern and its likely that this change would
effect performance (even when NOT using id types that do not
override equals).

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=567274&group_id=40712

[Hibernate] [ hibernate-Feature Requests-512567 ] Mapping Tool

[SourceForge.net <no...@so...>]

Feature Requests item #512567, was opened at 2002-02-04 12:49
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=512567&group_id=40712

Category: None
Group: None
>Status: Closed
Priority: 1
Submitted By: Daniel Bradby (dbradby)
Assigned to: Nobody/Anonymous (nobody)
Summary: Mapping Tool

Initial Comment:
As I've seen with Cocobase, it comes with a tool to do 
the mappings visually (it had want to for that price). 

We need something similar. These tools could generate 
mappings visually or through source parsing of doclets 
(thanks Paul). Also a way creating the database 
structure with this tool would be great.

I think we can get something simple done very soon.

----------------------------------------------------------------------

>Comment By: Gavin King (oneovthafew)
Date: 2003-01-06 12:41

Message:
Logged In: YES 
user_id=384580

We have had extensive discussions of a full rountripping
toolset based upon XDoclet and Middlegen. These are MUCH
more powerful ideas than a "mapping gui". This feature
request is now out of date.

----------------------------------------------------------------------

Comment By: Gavin King (oneovthafew)
Date: 2002-10-31 19:24

Message:
Logged In: YES 
user_id=384580

After having spent a little time playing with one of these
mapping tools, I've concluded that hand-editing the mapping
file is simply _much_ more efficient. Our mapping format is
quite readable (it was designed to be edited by hand).

I don't think this feature is desirable after all.

(Of course some parts of this feature request were done ages
ago.)

----------------------------------------------------------------------

Comment By: Christian Meunier (tlvenn)
Date: 2002-07-29 08:43

Message:
Logged In: YES 
user_id=583408

Take a look at JRelay too, their GUI is pretty good (better 
than cocobase imho)

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=428711&aid=512567&group_id=40712

[Hibernate] Tools (again)

[Gavin K. <Gav...@ex...>]

I'm not sure if people noticed, but I submitted an alternate Middlegen
plugin that generates hbm.xml, as discussed here. Theres a few
improvements I need to make to that, but I will try to get onto
"finishing" it in the next few days.

I also understand that Konstantin is adding composite id support to the
XDoclet module and that Max has very nearly finished the <meta> patch
for hbm2java.

To complete our full roundtripping I suggest the following additional
toolset features:

(1) create a <hbm2java> ant task (easy to do)

(2) add a @hibernate.meta tag to the XDoclet module

(3) add functionality for generation of XDoclet @hibernate.tags to
hbm2java

These new features will allow full reversability of <hibernatedoclet>
via <hbm2java> and vice versa.

Max has suggested that hbm2java could use a velocity template for code
generation. Particularly now that I have a bit of experience with
velocity, I certainly agree that it would be a nicer way of doing
things. However, from a practical point of view, I wouldn't want someone
to waste time on this if that would mean a big delay in delivering the
"finished" toolset. The #1 advantage I see to using a template is that
users would then be able to easily customize to template to suit their
own preferences. OTOH, if we wanted that, then we would probably need to
clean up the underlying "model" objects that represent the parsed
mapping document. So anyway, lets put that (the template) on the todo
list, but leave implementing (1) and (3), by whatever means, as the top
priority.

Thoughts?


********** CAUTION - Disclaimer **********
This message may contain privileged and confidential
information. If you are not the intended recipient of this
message (or responsible for delivery of the message to
such person) you are hereby notified that any use,
dissemination, distribution or reproduction of this message
is prohibited. If you have received this message in error,
you should destroy it and kindly notify the sender by reply
e-mail. Please advise immediately if you or your employer
do not consent to Internet e-mail for messages of this kind.
Opinions, conclusions and other information in this
message that do not relate to the official business of
Expert Information Services Pty Ltd ("The Company")
shall be understood as neither given nor endorsed by it.

The Company advises that this e-mail and any attached
files should be scanned to detect viruses. The Company
accepts no liability for loss or damage (whether caused
by negligence or not) resulting from the use of any
attached files.
**EIS******** End of Disclaimer **********

RE: [Hibernate] logo

[Gavin K. <Gav...@ex...>]

Ace! Blue Mars webdesign is excellent :)

I'd really appreciate their help. Let them know
we'd like something reasonably scalable so that
we can use the same design in the header of the
Wiki site as we use on the little "powered by"
button....

Thanks

> -----Original Message-----
> From: Christian Bauer [mailto:chr...@bl...]=20
> Sent: Monday, 6 January 2003 10:32 AM
> To: hib...@li...
> Subject: Re: [Hibernate] logo
>=20
>=20
> On 06 Jan (10:22), Gavin King wrote:
>=20
> > I'd love to have a logo but don't know anyone with the=20
> requisite gfx=20
> > skills ...
> >=20
> > Suggestions?
>=20
> We have designers in our company, maybe one of them has some=20
> time left.=20
> My girlfriend's also a graphics designer. I'll see what I can do and=20
> maybe come up with some suggestions for voting next week.
>=20
> --=20
> Christian Bauer
> tu...@in...
>=20
>=20
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf=20
> _______________________________________________
> hibernate-devel mailing list hib...@li...
> https://lists.sourceforge.net/lists/listinfo/hibernate-devel
>=20


********** CAUTION - Disclaimer **********
This message may contain privileged and confidential
information. If you are not the intended recipient of this
message (or responsible for delivery of the message to
such person) you are hereby notified that any use,
dissemination, distribution or reproduction of this message
is prohibited. If you have received this message in error,
you should destroy it and kindly notify the sender by reply
e-mail. Please advise immediately if you or your employer
do not consent to Internet e-mail for messages of this kind.
Opinions, conclusions and other information in this
message that do not relate to the official business of
Expert Information Services Pty Ltd ("The Company")
shall be understood as neither given nor endorsed by it.

The Company advises that this e-mail and any attached
files should be scanned to detect viruses. The Company
accepts no liability for loss or damage (whether caused
by negligence or not) resulting from the use of any
attached files.
**EIS******** End of Disclaimer **********

Re: [Hibernate] logo

[Christian B. <chr...@bl...>]

On 06 Jan (10:22), Gavin King wrote:

> I'd love to have a logo but don't know anyone with the
> requisite gfx skills ...
> 
> Suggestions?

We have designers in our company, maybe one of them has some time left. 
My girlfriend's also a graphics designer. I'll see what I can do and 
maybe come up with some suggestions for voting next week.

-- 
Christian Bauer
tu...@in...

RE: [Hibernate] logo

[Gavin K. <Gav...@ex...>]

I'd love to have a logo but don't know anyone with the
requisite gfx skills ...

Suggestions?


> do you have a hibernate logo i can use on=20
> http://boss.bekk.no/boss/middlegen/ when i commit the=20
> hibernate plugin? the logo.jpg that figured on the hibenrate=20
> site in pre-wiki days has gray background and not really=20
> suitable for "powered by hibernate".
>=20
> if you don't have one, i suggest you design one. i have=20
> feeling that many projects would like to have a "powered by=20
> hibernate" logo in the near future, and it's good pr for you.
>=20
> aslak
>=20
>=20
>=20
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf=20
> _______________________________________________
> hibernate-devel mailing list hib...@li...
> https://lists.sourceforge.net/lists/listinfo/hibernate-devel
>=20


********** CAUTION - Disclaimer **********
This message may contain privileged and confidential
information. If you are not the intended recipient of this
message (or responsible for delivery of the message to
such person) you are hereby notified that any use,
dissemination, distribution or reproduction of this message
is prohibited. If you have received this message in error,
you should destroy it and kindly notify the sender by reply
e-mail. Please advise immediately if you or your employer
do not consent to Internet e-mail for messages of this kind.
Opinions, conclusions and other information in this
message that do not relate to the official business of
Expert Information Services Pty Ltd ("The Company")
shall be understood as neither given nor endorsed by it.

The Company advises that this e-mail and any attached
files should be scanned to detect viruses. The Company
accepts no liability for loss or damage (whether caused
by negligence or not) resulting from the use of any
attached files.
**EIS******** End of Disclaimer **********

RE: [Hibernate] Hibernate 2 Transaction API

[Gavin K. <Gav...@ex...>]

> Not sure what benefits these methods add, since I don't think=20
> I'd ever=20
> use them...


Well, the benefit was supposed to be a slightly cleaner idiom
with no extra local variable holding the transaction, or
test for (tx!=3Dnull) in the catch block.


Nevertheless, upon reflection, I'm inclined to NOT make these
changes. Its likely that they would cause confusion for new
users and they certainly do "grow" the API without adding
functionality.

Thanks for your feedback.


********** CAUTION - Disclaimer **********
This message may contain privileged and confidential
information. If you are not the intended recipient of this
message (or responsible for delivery of the message to
such person) you are hereby notified that any use,
dissemination, distribution or reproduction of this message
is prohibited. If you have received this message in error,
you should destroy it and kindly notify the sender by reply
e-mail. Please advise immediately if you or your employer
do not consent to Internet e-mail for messages of this kind.
Opinions, conclusions and other information in this
message that do not relate to the official business of
Expert Information Services Pty Ltd ("The Company")
shall be understood as neither given nor endorsed by it.

The Company advises that this e-mail and any attached
files should be scanned to detect viruses. The Company
accepts no liability for loss or damage (whether caused
by negligence or not) resulting from the use of any
attached files.
**EIS******** End of Disclaimer **********

Re: [Hibernate] Hibernate 2 Transaction API

[Mark W. <mor...@SM...>]

One Ovthafew wrote:

>I'd like to propose a change to the Transaction API, for the purpose 
>of making the standard try/catch/finally idiom a bit cleaner. What I 
>would like is for the idiom to look like:
>
>Session s = factory.openSession();
>try {
>	//instantiate a new Transaction object, begin the transaction
>	s.beginTransaction();  
>	
>	// do some work
>	....
>	
>	// commit the transaction
>	s.transaction().commit();
>}
>catch (Exception e) {
>	// attempt to roll back the transaction
>	s.transaction().attemptRollback();
>	throw e;
>}
>finally {
>	s.close();
>}
>
>This will require the following changes:
>
>1. Change Session.beginTransaction() to keep a reference to the new 
>   Transaction instance the Session.
>

Shouldn't there be a Session.endTransaction() as well?  Or does 
Transaction.commit() implicitly end the Transaction (in which case you'd 
need a callback to let the Session know)?  Once you start using 
Transactions, do you expect users to not want to stop using them?  


>5. Add attemptCommit() and attemptRollback() methods to Transaction.
>   Rather than throwing an exception, these methods return boolean 
>   to indicate success or failure. In particular they return false 
>   if we did not successfully begin the transaction in the first place.
>
>6. Add Transaction.getLastException() to return any exception that
>   was squashed.
>  
>

Not sure what benefits these methods add, since I don't think I'd ever 
use them...


-Mark

[Hibernate] logo

[<asl...@ne...>]

do you have a hibernate logo i can use on
http://boss.bekk.no/boss/middlegen/ when i commit the hibernate plugin? the
logo.jpg that figured on the hibenrate site in pre-wiki days has gray
background and not really suitable for "powered by hibernate".

if you don't have one, i suggest you design one. i have feeling that many
projects would like to have a "powered by hibernate" logo in the near
future, and it's good pr for you.

aslak

[Hibernate] Hibernate 2 Transaction API

[One O. <1ov...@mb...>]

I'd like to propose a change to the Transaction API, for the purpose 
of making the standard try/catch/finally idiom a bit cleaner. What I 
would like is for the idiom to look like:

Session s = factory.openSession();
try {
	//instantiate a new Transaction object, begin the transaction
	s.beginTransaction();  
	
	// do some work
	....
	
	// commit the transaction
	s.transaction().commit();
}
catch (Exception e) {
	// attempt to roll back the transaction
	s.transaction().attemptRollback();
	throw e;
}
finally {
	s.close();
}

This will require the following changes:

1. Change Session.beginTransaction() to keep a reference to the new 
   Transaction instance the Session.
   
2. Add Session.transaction(). This method returns the Transaction 
   object currently associated with the Session.

5. Add attemptCommit() and attemptRollback() methods to Transaction.
   Rather than throwing an exception, these methods return boolean 
   to indicate success or failure. In particular they return false 
   if we did not successfully begin the transaction in the first place.

6. Add Transaction.getLastException() to return any exception that
   was squashed.


Suggestions? Votes?


---------------------------------------------------------------------
NEW to mBox, receive faxes to any email address!
Find out more http://www.mbox.com.au/fax

Re: [Hibernate] Length of Hibernate-generated aliases in SQL

[Mark W. <mor...@SM...>]

Mark Woon wrote:

> And I'd be more than happy to implement the fix, if I can just figure 
> out which methods are actually producing the offending SQL.  I'll try 
> looking at it later tonight.  I'm getting really tired of renaming my 
> db columns... 


I think I've got a reasonable fix for this problem.  Patch submitted to 
SourceForge 
(http://sourceforge.net/tracker/index.php?func=detail&aid=662501&group_id=40712&atid=428708).


-Mark