hibernate-devel — development

RE: [Hibernate-devel] New feature discussion (max query result list size)

[Gavin_King/Cirrus%<CI...@ci...>]

> I don't think it's a bad thing to start "breaking down" the Session
> interface like this, for the reasons you say:

>> On the _very_ positive side it allows easy and elegant extension
>> to setting other things like fetch sizes. Incidently, this is
>> the standard model used by persistence layers.

>...which arises from general good design principles, e.g. small
>special-purposes interfaces being generally better and more cleanly
>extensible than large general-purpose interfaces.

Yes, absolutely. I still don't think it is completely misguided to
have the find() and iterate() methods there on the Session, for
convenience. My thinking on this (much) earlier on was that

list = s.find("from foo in class bar.Foo where foo.name=?", "foo",
Hibernate.STRING);

is more compact than:

Query q = s.prepareQuery("from foo in class bar.Foo where foo.name=?");
q.setParameter(1, "foo", Hibernate.STRING);
list = q.execute();
q.close();

But certainly the second form is more flexible and arguably more elegant.
I think it makes perfect sense to support both, personally....

>However, Hibernate is a little different in that it assumes that its
clients
>are always Java code, whereas these other query languages have to deal
with
>being called from all sorts of clients, and can't easily expose a
consistent
>object API to all its possible clients.  Hibernate doesn't have that
>restriction, so it's less important to have a feature like this in the
query
>language itself.

I think theres an interesting issue here in that while certain things are
more directly and elegantly expressible in Java, others are much easier to
express in a query language. I have a problem with some persistence
solutions that reject the idea of embedding query strings in application
code and try to construct a query through manipulation of objects.
Usually this code is quite unreadable and logwinded.

I think this particular problem is a borderline case. There is a strong
argument somewhere that borderline cases should be done in Java code for
the sake of compile-time syntax checking, self-documentation, etc.....

RE: [Hibernate-devel] New feature discussion (max query result list size)

[Anton v. S. <an...@ap...>]

> > 3. for advanced usage of queries, provide a Query object with a
> > setMaxResults().

This sounds best to me.

I don't think it's a bad thing to start "breaking down" the Session
interface like this, for the reasons you say:

> On the _very_ positive side it allows easy and elegant extension
> to setting other things like fetch sizes. Incidently, this is
> the standard model used by persistence layers.

...which arises from general good design principles, e.g. small
special-purposes interfaces being generally better and more cleanly
extensible than large general-purpose interfaces.

I took a brief look at ODMG OQL to see whether that had a query-limiting
mechanism in the language, but I didn't find it(which doesn't mean it's not
there).  However, it does seem to me as though this requirement is
orthogonal to the business of producing a query, and doesn't necessarily
belong in the query language.  Putting it in the language might be nice, but
that would be icing on the cake, IMO.

BTW, I remembered that Microsoft Access uses "SELECT TOP n ...", so we're up
to 5 systems that have it in the query language.

However, Hibernate is a little different in that it assumes that its clients
are always Java code, whereas these other query languages have to deal with
being called from all sorts of clients, and can't easily expose a consistent
object API to all its possible clients.  Hibernate doesn't have that
restriction, so it's less important to have a feature like this in the query
language itself.

Anton

RE: [Hibernate-devel] New feature discussion (max query result list size)

[Anton v. S. <an...@ap...>]

> if you're going to change the language, their are at least 2
> other options :
>
> (a) oracle uses "WHERE ROWNUM<?"
>
> (b) informix uses "SELECT FIRST ?"

FWIW, Sybase/Microsoft T-SQL uses "SET ROWCOUNT n" to limit returned rows in
subsequent queries.  (later more Microsoft-specific SQL dialects may do it
differently.)

Re: [Hibernate-devel] New feature discussion (max query result list size)

[<no...@ex...>]

if you're going to change the language, their are at least 2 other options :

(a) oracle uses "WHERE ROWNUM<?"

(b) informix uses "SELECT FIRST ?"

I have no idea if there is a ANSI standard for this :-)


----- Original Message -----
From: <Gavin_King/Cirrus%CI...@ci...>
To: <hib...@li...>
Sent: Wednesday, June 19, 2002 10:06 AM
Subject: [Hibernate-devel] New feature discussion (max query result list
size)


> I just posted the following as a feature request. Its something I would
> very much like to see just after 1.0
>
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
> It would be very useful to allow the application to limit the size of a
> query result list. The three obvious ways to do this would be:
>
> 1. provide overloaded versions of find() and iterate() that take an int
> argument
>
> 2. allow a LIMIT clause in the query language (like MySQL)
>
> 3. for advanced usage of queries, provide a Query object with a
> setMaxResults().
>
> approach (1) has the disadvantage of causing an explosion of methods on
the
> Session interface - and a further possible explosion if we ever decide to
> extend this functionality to allowing offsets, fetch sizes, etc....
>
> approach(2) leaves slightly open the question of how to bind a dynamic
> value to the LIMIT value .... the best way to do this would be with a "?"
> placeholder that looks just like the JDBC in parameters. Also this
approach
> complexifies the query language which is somewhat undesirable.
>
> approach (3) requires a single new prepareQuery() method of the Session
> interface. However, it also requires a change in the was users think about
> queries. On the _very_ positive side it allows easy and elegant extension
> to setting other things like fetch sizes. Incidently, this is the standard
> model used by persistence layers.
>
> Given the preceding discussion, I would be ruling out (1) in favor of (2)
> or (3) - which are not mutually exclusive, actually.
>
> Thoughts?
>
>
> --------------------------------------------------------------------------
--
>                    Bringing you mounds of caffeinated joy
>                    >>>     http://thinkgeek.com/sf    <<<
>
> _______________________________________________
> Hibernate-devel mailing list
> Hib...@li...
> https://lists.sourceforge.net/lists/listinfo/hibernate-devel
>

[Hibernate-devel] New feature discussion (max query result list size)

[Gavin_King/Cirrus%<CI...@ci...>]

I just posted the following as a feature request. Its something I would
very much like to see just after 1.0

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

It would be very useful to allow the application to limit the size of a
query result list. The three obvious ways to do this would be:

1. provide overloaded versions of find() and iterate() that take an int
argument

2. allow a LIMIT clause in the query language (like MySQL)

3. for advanced usage of queries, provide a Query object with a
setMaxResults().

approach (1) has the disadvantage of causing an explosion of methods on the
Session interface - and a further possible explosion if we ever decide to
extend this functionality to allowing offsets, fetch sizes, etc....

approach(2) leaves slightly open the question of how to bind a dynamic
value to the LIMIT value .... the best way to do this would be with a "?"
placeholder that looks just like the JDBC in parameters. Also this approach
complexifies the query language which is somewhat undesirable.

approach (3) requires a single new prepareQuery() method of the Session
interface. However, it also requires a change in the was users think about
queries. On the _very_ positive side it allows easy and elegant extension
to setting other things like fetch sizes. Incidently, this is the standard
model used by persistence layers.

Given the preceding discussion, I would be ruling out (1) in favor of (2)
or (3) - which are not mutually exclusive, actually.

Thoughts?

[Hibernate-devel] 1.0rc1

[Gavin_King/Cirrus%<CI...@ci...>]

Go check it out!

I think we've discussed most of the changes here so no point listing them
now. Except for: at the last minute I added one-to-one associations :)

I guess everyone will also like the new Javadoc which is _much_ more
comprehensive.

Re: [Hibernate-devel] How do you get hold of a javax.transaction.Transaction?

[Gavin_King/Cirrus%<CI...@ci...>]

>I needed to do this using Orion as my application server, and here is
>how I did it:
>
>TransactionManager manager = (TransactionManager)new
>InitialContext().lookup("java:comp/UserTransaction");
>
>However, I don't know if it works the same on other applications
servers...

Just tried that and some variations in websphere (websphere lists
the UserTransaction as "jta/usertransaction") and no luck
unfortunately. The UserTransaction can't be cast to Transaction or
TransactionManager. I spent some serious time searching the web for
info on this and all I could turn up was that its possible to
do in Orion + JBoss but perhaps not in WebLogic + WebSphere....

might have to do without it.

thanks

Gavin

Re: [Hibernate-devel] How do you get hold of a javax.transaction.Transaction?

[Jon L. <jon...@xe...>]

I think this might have to be yet another configurable property.

I needed to do this using Orion as my application server, and here is how I
did it:

TransactionManager manager = (TransactionManager)new
InitialContext().lookup("java:comp/UserTransaction");

However, I don't know if it works the same on other applications servers...

Jon...

----- Original Message -----
From: <Gavin_King/Cirrus%CI...@ci...>
To: <hib...@li...>
Sent: Friday, June 14, 2002 12:53 PM
Subject: [Hibernate-devel] How do you get hold of a
javax.transaction.Transaction?


> Does anyone know a reasonably portable way to get hold of a
> javax.transaction.Transaction or javax.transaction.TransactionManager
> from an application server JNDI server?
>
> If I knew how to do this, I could get the caching stuff working
> properly for situations where the JTA transaction is incomplete when
> we call Session.close(). We would need to register an instance
> of javax.transaction.Synchronization with the
> javax.transaction.Transaction instance.
>
> (At present, for the cache to work correctly, the transaction must
> be either fully committed or rolled back when you close the Session.)
>
> Gavin
>
>
> _______________________________________________________________
>
> Don't miss the 2002 Sprint PCS Application Developer's Conference
> August 25-28 in Las Vegas -
http://devcon.sprintpcs.com/adp/index.cfm?source=osdntextlink
>
> _______________________________________________
> Hibernate-devel mailing list
> Hib...@li...
> https://lists.sourceforge.net/lists/listinfo/hibernate-devel
>

[Hibernate-devel] How do you get hold of a javax.transaction.Transaction?

[Gavin_King/Cirrus%<CI...@ci...>]

Does anyone know a reasonably portable way to get hold of a
javax.transaction.Transaction or javax.transaction.TransactionManager
from an application server JNDI server?

If I knew how to do this, I could get the caching stuff working
properly for situations where the JTA transaction is incomplete when
we call Session.close(). We would need to register an instance
of javax.transaction.Synchronization with the
javax.transaction.Transaction instance.

(At present, for the cache to work correctly, the transaction must
be either fully committed or rolled back when you close the Session.)

Gavin

[Hibernate-devel] Hibernate.config() now in CVS

[Gavin_King/Cirrus%<CI...@ci...>]

I checked in the new configuration system last night, after testing it
out in websphere.

Check it out - tell me what sucks.

RE: [Hibernate-devel] XML SessionFactory configuration

[Gavin_King/Cirrus%<CI...@ci...>]

Hi Brad

>were u thinking of leaving the option to configure programmatically in?

Absolutely. There will always be people who need to do things their own
way. Remember we aren't making any assumptions about architecture here.

>using hibernate in the container, u then need multiple configuration
>files - each file needs a different name and then how do u to tell
>hibernate to use the correct file for a particular app; or u need one
>configuration file that can contain the configuration for each
>application - hibernate still needs to be told which app it is dealing
>with.

Actually in what I've implemented, this is okay, I think. (depending
upon exact arrangement of ClassLoaders.) You can configure multiple
SessionFactorys in the same file, with different JNDI names. Then the
application looks up its own SessionFactory in JNDI, specifying the
correct name for that application.

Otherwise, instead of calling Hibernate.configure(), perhaps we could
let you call:

  new Configuration("my/application/hibernate.cfg.xml").configure();

but that would be for special cases. I wish this stuff was in CVS so
you can have a look ... it will be there tonight :)

>ideally we want to have an application's configuration all in once place
>- including the configuration of third-party frameworks.  to achieve
>this we have writen a wrapper/utility class for hibernate that (among
>other things) integrates with our application's configuration framework.

I promise this code will continue to work well past version 1. Having
never recieved a single complaint/suggestion for the whole
Datastore/SessionFactory API, I'm quite convinced it is serving people
well. Hence I have no intention of changing/removing it.

peace

Gavin

RE: [Hibernate-devel] XML SessionFactory configuration

[Brad C. <bra...@wo...>]

were u thinking of leaving the option to configure programmatically in?

the option to configure by file would definitely be usefull in some
circumstances.  most people r probably aware of the problems with this
in the following scenario, but i'll babble on about it anyway.  when
deploying an application as some sort of archive (eg. war) and u have
staging and live servers, u don't want to change the archive file
between them (but the configuration is most likely different).  so then
the xml configuration file has to live outside the archive, which leaves
the global classpath to the container.  if u have multiple applications
using hibernate in the container, u then need multiple configuration
files - each file needs a different name and then how do u to tell
hibernate to use the correct file for a particular app; or u need one
configuration file that can contain the configuration for each
application - hibernate still needs to be told which app it is dealing
with.

ideally we want to have an application's configuration all in once place
- including the configuration of third-party frameworks.  to achieve
this we have writen a wrapper/utility class for hibernate that (among
other things) integrates with our application's configuration framework.
all our code calls this utility class to get a session, which worries
about configuration, etc.  this is obviously not a solution for
hibernate itself.  at this stage i have not come up with any ideas that
involve changing hibernate to improve the situation.

brad

> -----Original Message-----
> From: Gavin_King/Cirrus%CI...@ci...
> [mailto:Gavin_King/Cirrus%CI...@ci...]
> Sent: Wednesday, 12 June 2002 5:41 PM
> To: hib...@li...
> Subject: [Hibernate-devel] XML SessionFactory configuration
>=20
>=20
>=20
> I've been thinking about the problem of configuring a=20
> SessionFactory for
> a J2EE application where application code has no control over=20
> the system
> initialization procedure. I'm thinking that instead of forcing the
> application to do this programmatically, there could be a config file
> like:
>=20
>  <hibernate-configuration>
>=20
>    <property name=3D"show_sql">false</property>
>    <property name=3D"use_outer_join">true</property>
>    <property name
> =3D"jta.UserTransaction">java:comp/UserTransaction/</property>
>=20
>    <session-factory name=3D"/jndi/name">
>      <property name=3D"datasource">my/first/datasource</property>
>      <mapping resource=3D"eg/Foo.hbm.xml"/>
>      <mapping resource=3D"eg/Bar.hbm.xml"/>
>    </session-factory>
>=20
>    <session-factory>
>      <property name=3D"datasource">my/other/datasource</property>
>      <mapping file=3D"C:/mapping/my_mappings.hbm.xml"/>
>    </session-factory>
>=20
>  </hibernate-configuration>
>=20
> What I'm thinking of is that a call to
>=20
>     Hibernate.configure()
>=20
> would look for a resource called hibernate.cfg.xml and use that to
> configure and register SessionFactory(s) with JNDI. Subsequent
> calls would do nothing. That way application components could all
> call Hibernate.configure() when instantiated but we would guarantee
> that the factories would only be initialized once.
>=20
> I dont want to see an explosion of configuration files but I think
> this is reasonable.
>=20
> Opinions? Suggestions?
>=20
> I will check in some code sometime in the next couple of days for
> people to play with....
>=20
> Gavin.
>=20
>=20
> _______________________________________________________________
>=20
> Sponsored by:
> ThinkGeek at http://www.ThinkGeek.com/
> _______________________________________________
> Hibernate-devel mailing list
> Hib...@li...
> https://lists.sourceforge.net/lists/listinfo/hibernate-devel
>=20

Re: [Hibernate-devel] XML SessionFactory configuration

[Gavin_King/Cirrus%<CI...@ci...>]

I expressed this badly, of course. I was meaning to imply no hooks into
the system initialization procedure as opposed to hooks into the
component _instance_ initialization procedure. You want to configure
a single session factory thats shared between potentially many instances
of several components (whether these components be servlets or enterprise
beans). I think its reasonable to try and make that a bit easier for
people.

P.S. Does anyone here use Avalon. I've been having a look at their stuff
lately and kind of like it and I'm wondering about writing an adapter that
exposes hibernate as an Avalon component. But I've never used Avalon so
I would probably not get it right on my own......

> While I don't necessarily think this is a bad idea, I'm not sure I
> quite understand why you say a J2EE app has no control over
> initialization. What particular app model(s) did you have in mind here?
> WAR's for servlet engines? EJB apps? other server side models?

Re: [Hibernate-devel] XML SessionFactory configuration

[Paul S. <pau...@ne...>]

 
> I've been thinking about the problem of configuring a SessionFactory for
> a J2EE application where application code has no control over the system
> initialization procedure. I'm thinking that instead of forcing the
> application to do this programmatically, there could be a config file
> like:

While I don't necessarily think this is a bad idea, I'm not sure I quite 
understand why you say a J2EE app has no control over initialization. What 
particular app model(s) did you have in mind here? WAR's for servlet engines? 
EJB apps? other server side models?

PaulS.


---------------------------------------------
This message was sent using Vianet's Web Based Emailer.
http://www.vianet.net.au

Re: [Hibernate-devel] XML SessionFactory configuration

[Jon L. <jon...@xe...>]

I think this makes sense, and I definately have a use for it.  I've
implemented more or less the same thing at the web application level where
with a ServletContext Listener that  listens for the ServletContextCreated
event and creates a SessionFactory and stores it in the application scope.

With this in place, I would no longer need this code, and I could share the
SessionFactory between the instances of the web application.

----- Original Message -----
From: <Gavin_King/Cirrus%CI...@ci...>
To: <hib...@li...>
Sent: Wednesday, June 12, 2002 9:40 AM
Subject: [Hibernate-devel] XML SessionFactory configuration


>
> I've been thinking about the problem of configuring a SessionFactory for
> a J2EE application where application code has no control over the system
> initialization procedure. I'm thinking that instead of forcing the
> application to do this programmatically, there could be a config file
> like:
>
>  <hibernate-configuration>
>
>    <property name="show_sql">false</property>
>    <property name="use_outer_join">true</property>
>    <property name
> ="jta.UserTransaction">java:comp/UserTransaction/</property>
>
>    <session-factory name="/jndi/name">
>      <property name="datasource">my/first/datasource</property>
>      <mapping resource="eg/Foo.hbm.xml"/>
>      <mapping resource="eg/Bar.hbm.xml"/>
>    </session-factory>
>
>    <session-factory>
>      <property name="datasource">my/other/datasource</property>
>      <mapping file="C:/mapping/my_mappings.hbm.xml"/>
>    </session-factory>
>
>  </hibernate-configuration>
>
> What I'm thinking of is that a call to
>
>     Hibernate.configure()
>
> would look for a resource called hibernate.cfg.xml and use that to
> configure and register SessionFactory(s) with JNDI. Subsequent
> calls would do nothing. That way application components could all
> call Hibernate.configure() when instantiated but we would guarantee
> that the factories would only be initialized once.
>
> I dont want to see an explosion of configuration files but I think
> this is reasonable.
>
> Opinions? Suggestions?
>
> I will check in some code sometime in the next couple of days for
> people to play with....
>
> Gavin.
>
>
> _______________________________________________________________
>
> Sponsored by:
> ThinkGeek at http://www.ThinkGeek.com/
> _______________________________________________
> Hibernate-devel mailing list
> Hib...@li...
> https://lists.sourceforge.net/lists/listinfo/hibernate-devel
>

[Hibernate-devel] XML SessionFactory configuration

[Gavin_King/Cirrus%<CI...@ci...>]

I've been thinking about the problem of configuring a SessionFactory for
a J2EE application where application code has no control over the system
initialization procedure. I'm thinking that instead of forcing the
application to do this programmatically, there could be a config file
like:

 <hibernate-configuration>

   <property name="show_sql">false</property>
   <property name="use_outer_join">true</property>
   <property name
="jta.UserTransaction">java:comp/UserTransaction/</property>

   <session-factory name="/jndi/name">
     <property name="datasource">my/first/datasource</property>
     <mapping resource="eg/Foo.hbm.xml"/>
     <mapping resource="eg/Bar.hbm.xml"/>
   </session-factory>

   <session-factory>
     <property name="datasource">my/other/datasource</property>
     <mapping file="C:/mapping/my_mappings.hbm.xml"/>
   </session-factory>

 </hibernate-configuration>

What I'm thinking of is that a call to

    Hibernate.configure()

would look for a resource called hibernate.cfg.xml and use that to
configure and register SessionFactory(s) with JNDI. Subsequent
calls would do nothing. That way application components could all
call Hibernate.configure() when instantiated but we would guarantee
that the factories would only be initialized once.

I dont want to see an explosion of configuration files but I think
this is reasonable.

Opinions? Suggestions?

I will check in some code sometime in the next couple of days for
people to play with....

Gavin.

[Hibernate-devel] Feature for Lazy Initialization using Proxies

[Gavin_King/Cirrus%<CI...@ci...>]

Damn, I meant to send a mail to the list. Instead I sent it to Jon only.
I hope he doesnt mind me reposting his response with my mail quoted
at the bottom.....

Gavin

Jon wrote:

>Don't worry, I wouldn't check anything like this straight in...  I just
>wanted to see what the thoughts of everyone else was on this "feature".
>
>I'll go ahead an implement this for collections too and then submit it as
a
>patch for everyone to take a look at.
>
>I see your point about this being available for _disconnected_ versus
>_closed_ sessions, and I'll be happy to make my code make this distinction
>between the two cases.
>
>Jon...

----- Original Message -----
From: <Gavin_King/Cirrus%CI...@ci...>
To: "Jon Lipsky" <jon...@xe...>
Sent: Tuesday, June 11, 2002 12:29 PM
Subject: Re: [Hibernate-devel] Feature for Lazy Initialization using
Proxies


>
> A few thoughts on this, off the top of my head...
>
> 1. If its done for proxies, it should also be done for collections.
> 2. I'm happy for this to occur with a _disconnected_ Session, but I
> dunno about with a _closed_ Session.
> 3. There is a broader question of using lazy initialization with
> session disconnection/reconnection in that we can't do a version
> number check when we initialize. I think it's sort of okay, as long
> as the user aware of what (s)he is doing.....
> 4. Its a feature that should definately be disabled by default.
>
> Jon, please don't commit any code like this str8 into CVS. Would you
> please use the patch manager on the Sourceforge page, so we can all
> have a look over it before you commit it.....
>
> Thanks for all the help.
>
> Gavin
>
> >I modified the LazyInitializer so that instead of throwing an exception
> >because the session is closed, it will open a new connection and
reconnect
> >the session if the session, do the find, and the close the session
> again...  >This way I only have to load the objects I really need at
> runtime, and I >don't need to worry about managing the session.
>
> >If anyone has any comments on this kind of functionality, then I would
> >definately like to hear them.
>

Re: [Hibernate-devel] Feature for Lazy Initialization using Proxies

[Paul S. <pau...@ne...>]

Hi John,

I assume it's turned off by default: backwards compatibity is a wonderful thing.

PaulS :)

> Hi...
> 
> I patched my version of Hibernate with a feature that I think is useful,
> but before I committed it I wanted make sure everyone agrees with what I
> have done and if we should make this feature configurable, etc...
> 
> Normally if you open a session, load or find some object, then close the
> session you will be unable to access the objects with a many-to-one
> relation if it is setup to use lazy initialization.
> 
> I modified the LazyInitializer so that instead of throwing an exception
> because the session is closed, it will open a new connection and
> reconnect the session if the session, do the find, and the close the
> session again...  This way I only have to load the objects I really need
> at runtime, and I don't need to worry about managing the session.
> 
> If anyone has any comments on this kind of functionality, then I would
> definately like to hear them.
> 
> Thanks,
> Jon...
> 





---------------------------------------------
This message was sent using Vianet's Web Based Emailer.
http://www.vianet.net.au

[Hibernate-devel] Feature for Lazy Initialization using Proxies

[Jon L. <jon...@xe...>]

Hi...

I patched my version of Hibernate with a feature that I think is useful, =
but before I committed it I wanted make sure everyone agrees with what I =
have done and if we should make this feature configurable, etc...

Normally if you open a session, load or find some object, then close the =
session you will be unable to access the objects with a many-to-one =
relation if it is setup to use lazy initialization.

I modified the LazyInitializer so that instead of throwing an exception =
because the session is closed, it will open a new connection and =
reconnect the session if the session, do the find, and the close the =
session again...  This way I only have to load the objects I really need =
at runtime, and I don't need to worry about managing the session.

If anyone has any comments on this kind of functionality, then I would =
definately like to hear them.

Thanks,
Jon...

[Hibernate-devel] cirrus.hibernate.Lifecycle + other changes

[Gavin_King/Cirrus%<CI...@ci...>]

Since nobody screamed frantic disagreement at my suggestion for a new
Lifecycle interface, I've gone ahead and implemented it. I'd very much
appreciate if people would look over what I've got there in CVS and
make suggestions/criticisms.

http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/hibernate/Hibernate/cirrus/hibernate/Lifecycle.java?rev=1.1&content-type=text/vnd.viewcvs-markup

I also added Session.findIdentifiers() which returns object ids from
a query rather than the actual objects. This will often be useful,
especially for finder methods of BMP entity beans.

I've also tidied up Session, removing some deprecated methods finally,
in preparation for going out of beta. Hopefully the API is now pretty
stable. Are there any extra "utility" methods that would be helpful,
eg. Session.isPersistent(object)?

I've properly Javadoc'd all the extra little "extension" APIs like
ConnectionProvider, Type, IDGenerator, Dialect so people can feel
more comfortable making customizations. Of course, we can't commit
to keeping these interfaces as backward compatible as the
cirrus.hibernate package because some of them simply _will_ need
to change as functionality is added.

Re: [Hibernate-devel] 0.9.14

[Gavin_King/Cirrus%<CI...@ci...>]

>just wanted to contribute some code, but the latest I got from CVS
>doesn't seem to compile (missing files?).

Not sure. I think it should compile right now. At some stage in the past
48 hours I was missing a couple of new files for a few hours but I fixed
that yesterday. As Anton pointed out, you do need j2ee.jar for the
javax.sql.Datasource.....

>Did I miss something? I just did a module checkout, dropped a junit.jar
>in the 'lib' directory, and ran ant.

Can't comment on the Ant build script. There _may_ be something missing
there.

>BTW - releases don't seem to be labelled in CVS, can they be
>reconstructed based on date/time?

Ah. Yeah thats something I will have to start doing properly. From now on.
Sorry.

Gavin.

RE: [Hibernate-devel] 0.9.14

[Anton v. S. <an...@ap...>]

> just wanted to contribute some code, but the latest I got from CVS
> doesn't seem to compile (missing files?).
>
> Did I miss something? I just did a module checkout, dropped a junit.jar
> in the 'lib' directory, and ran ant.

There's a dependency in a few places in Hibernate and C3P0 on
javax.sql.DataSource, so adding jdbc2_0-stdext.jar (or equivalent) to the
build should do the trick.  As Gavin recently pointed out, the Ant script
was contributed and is only infrequently maintained.  However, it did work
when I tried it recently, with the addition of that library (which I had
forgotten about, otherwise I would have tried to fix it).

> BTW - releases don't seem to be labelled in CVS, can they be
> reconstructed based on date/time?

Don't know the answer to that... Gavin?

Anton

Re: [Hibernate-devel] 0.9.14

[Paul S. <pau...@ne...>]

Hi,

just wanted to contribute some code, but the latest I got from CVS 
doesn't seem to compile (missing files?).

Did I miss something? I just did a module checkout, dropped a junit.jar 
in the 'lib' directory, and ran ant.

BTW - releases don't seem to be labelled in CVS, can they be 
reconstructed based on date/time?

PaulS.

Re: [Hibernate-devel] Using streams for binary types

[Gavin_King/Cirrus%<CI...@ci...>]

> I think it might make sense to enable this by default since doing it this
> way should support every database.

Yeah, thats was what I was thinking.

I will run the tests on all platforms before the next release just to make
sure.

Re: [Hibernate-devel] Using streams for binary types

[Jon L. <jon...@xe...>]

I think it might make sense to enable this by default since doing it this
way should support every database.

Jon...

P.S. - Sorry about the spaces, I didn't know I did it...  I ask my
developers to do the same, so hopefully they won't find out I didn't follow
my own rules. :-)

----- Original Message -----
From: <Gavin_King/Cirrus%CI...@ci...>
To: "Jon Lipsky" <jon...@xe...>
Cc: <hib...@li...>
Sent: Friday, June 07, 2002 8:51 AM
Subject: Re: [Hibernate-devel] Using streams for binary types


>
> > I just checked in a change to add support for using streams when setting
> > binary types.
>
> Thanks Jon. :) I wonder if this option should be enabled by default.....
>
> Somewhere back in the dim dark past I did some experiments with stuff like
> this but I forget the details.
>
> Gavin
>
> P.S. Sorry to be a whingy nitpicker, but can we please keep using tabs
> instead of spaces for indenting the sourcecode.
>