Re: [Hibernate] Configuration artifact precedence
But the drawback is that I need to delay all the binding operation till the actual buildSessionFactory() launch, and thus keeping a list of classes and Document to be processed. This consumes more memory (higher peak until the buildSessionFactory call), but its probably not a show stopper. That would be a pretty *big* memory peak - any reason why we just not take the first thing that hibernate gets configured with instead of this magic ? Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
RE: [Hibernate] Configuration artifact precedence
I don't understand what is the problem you are trying to solve? From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Emmanuel BernardSent: Thursday, 18 August 2005 8:54 PMTo: hibernate-devel@lists.sourceforge.netSubject: [Hibernate] Configuration artifact precedence Artifacts (ie hbm files and annotated classes) were not able to point to the same entity-name (== class name for annotations) in an AnnotationConfiguration object: a duplicate mapping exception were raised. I've added an hibernate.artifact property (we need to find a good name), to allow precedence from one artifact over an other.hibernate.artifact hbm, class (default) will prioritize hbm files over annotated classes for the same entity-name.hibernate.artifact class, hbm will prioritize annotated classes over hbm files for the same entity-name.hibernate.artifact class will ignore any hbm file passed to AnnotationConfigurationThis is a much more powerful version of hibernate.archive (slightly diff semantic but same final goal, I'll remove hibernate.archive).There are a limitation though, you cannot mix annotated classes and hbm files in a mapped class hierarchy.I'll add ejb3xml when implemented. I think this concept might come to Hibernate core at some point (at least the infrastructure code). But the drawback is that I need to delay all the binding operation till the actual buildSessionFactory() launch, and thus keeping a list of classes and Document to be processed. This consumes more memory (higher peak until the buildSessionFactory call), but its probably not a show stopper.Well, just to let you know and to help me to find a good configuration property name.
Re: [Hibernate] Configuration artifact precedence
When the users has both hbm.xml and annotations in his project (most likely doing their migration phase). I don't get why it needs special attention (besides just take what comes first), but if there is a reason for it then it sounds like hibernate.mappingprecedence would be a better name. But I don't like this requirment of queuing up all the dom4js to be able to decide it - will just give us trouble I think. /max I don't understand what is the problem you are trying to solve? From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Emmanuel Bernard Sent: Thursday, 18 August 2005 8:54 PM To: hibernate-devel@lists.sourceforge.net Subject: [Hibernate] Configuration artifact precedence Artifacts (ie hbm files and annotated classes) were not able to point to the same entity-name (== class name for annotations) in an AnnotationConfiguration object: a duplicate mapping exception were raised. I've added an hibernate.artifact property (we need to find a good name), to allow precedence from one artifact over an other. hibernate.artifact hbm, class (default) will prioritize hbm files over annotated classes for the same entity-name. hibernate.artifact class, hbm will prioritize annotated classes over hbm files for the same entity-name. hibernate.artifact class will ignore any hbm file passed to AnnotationConfiguration This is a much more powerful version of hibernate.archive (slightly diff semantic but same final goal, I'll remove hibernate.archive). There are a limitation though, you cannot mix annotated classes and hbm files in a mapped class hierarchy. I'll add ejb3xml when implemented. I think this concept might come to Hibernate core at some point (at least the infrastructure code). But the drawback is that I need to delay all the binding operation till the actual buildSessionFactory() launch, and thus keeping a list of classes and Document to be processed. This consumes more memory (higher peak until the buildSessionFactory call), but its probably not a show stopper. Well, just to let you know and to help me to find a good configuration property name. --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
Max Rydahl Andersen wrote: But the drawback is that I need to delay all the binding operation till the actual buildSessionFactory() launch, and thus keeping a list of classes and Document to be processed. This consumes more memory (higher peak until the buildSessionFactory call), but its probably not a show stopper. That would be a pretty *big* memory peak - any reason why we just not take the first thing that hibernate gets configured with instead of this magic ? We could hardcode that, but there is no reason to prefer hbm over ejb3xml for example. Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
We could hardcode that, but there is no reason to prefer hbm over ejb3xml for example. when will we ever have a user add both a hbm.xml AND ejb3xml file ? And if we have - again, why not just choose the first one that gets added to the configuration by the user ? /max Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
Duplicate mapping exceptions while detecting artifacts in a .par file. Esp on smooth migration from hbm files to annotated classes. Gavin King wrote: I don't understand what is the problem you are trying to solve? *From:* [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] *On Behalf Of *Emmanuel Bernard *Sent:* Thursday, 18 August 2005 8:54 PM *To:* hibernate-devel@lists.sourceforge.net *Subject:* [Hibernate] Configuration artifact precedence Artifacts (ie hbm files and annotated classes) were not able to point to the same entity-name (== class name for annotations) in an AnnotationConfiguration object: a duplicate mapping exception were raised. I've added an /hibernate.artifact/ property (we need to find a good name), to allow precedence from one artifact over an other. /hibernate.artifact hbm, class/ (default) will prioritize hbm files over annotated classes for the same entity-name. /hibernate.artifact class, hbm/ will prioritize annotated classes over hbm files for the same entity-name. /hibernate.artifact class/ will ignore any hbm file passed to AnnotationConfiguration This is a much more powerful version of /hibernate.archive /(slightly diff semantic but same final goal, I'll remove hibernate.archive). There are a limitation though, you cannot mix annotated classes and hbm files in a mapped class hierarchy. I'll add ejb3xml when implemented. I think this concept might come to Hibernate core at some point (at least the infrastructure code). But the drawback is that I need to delay all the binding operation till the actual buildSessionFactory() launch, and thus keeping a list of classes and Document to be processed. This consumes more memory (higher peak until the buildSessionFactory call), but its probably not a show stopper. Well, just to let you know and to help me to find a good configuration property name. --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
Max Rydahl Andersen wrote: That would be a pretty *big* memory peak 13k per Document object => 13Mo for 1000 docs. Do you want to maintain manually a 1000 file list in your hibernate.cfg.xml or persistence.xml to save 13Mo temporary used! --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
When Christian migrate CaveatEmptor to HEM, the encounters issues with the discovery mechanism, because there were no notion of precedence. I think the idea of configurable precedence came from here. I have to admit I liked the idea to ease migrations. It's a bit like the BIOS boot ordering, you never used it, until you have to. The problem is that I discover them in the jar file, they are not added by the user. So I have no idea of the ordering. when will we ever have a user add both a hbm.xml AND ejb3xml file ? And if we have - again, why not just choose the first one that gets added to the configuration by the user ? /max Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
well I guess its the lesser of two evils ;) Max Rydahl Andersen wrote: That would be a pretty *big* memory peak 13k per Document object => 13Mo for 1000 docs. Do you want to maintain manually a 1000 file list in your hibernate.cfg.xml or persistence.xml to save 13Mo temporary used! --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
Yes I know the problem - and I also didn't find any hint in the spec about how they plan to define the precedence with respect to ejb3 xml in relation to annotations...any news on that ? And I guess having precedence as something you can change to allow for easier migration would be a good thing. As it is now Annotations/ejb3xml wins first - correct ? But couldn't we just say hbm.xml wins first and then users can remove the hbm.xml when they want the annotations to take over ? Or will that not have any affect because the annotation process occurs before users get the chance to add their hbm.xml files... Just thinking out loud... /max When Christian migrate CaveatEmptor to HEM, the encounters issues with the discovery mechanism, because there were no notion of precedence. I think the idea of configurable precedence came from here. I have to admit I liked the idea to ease migrations. It's a bit like the BIOS boot ordering, you never used it, until you have to. The problem is that I discover them in the jar file, they are not added by the user. So I have no idea of the ordering. when will we ever have a user add both a hbm.xml AND ejb3xml file ? And if we have - again, why not just choose the first one that gets added to the configuration by the user ? /max Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
Wouldn't it make sense for it to throw an exception if the mappings are different, otherwise who cares? In migrating, if one makes a mistake it would be good to know - "xxx has two different configurations" and not pick one which has precedence (I'm coming from the standpoint of someone who may be migrating soon). Unless, of course, the configuration artifacts are always different between those mapped by Annotations/ejb3xml and those mapped by hbm.xml. It would also definately make sense if this was configurable. I'm thinking out loud too :P -JL - Original Message - From: "Max Rydahl Andersen" <[EMAIL PROTECTED]> To: "Emmanuel Bernard" <[EMAIL PROTECTED]> Cc: Sent: Friday, August 19, 2005 5:23 AM Subject: Re: [Hibernate] Configuration artifact precedence Yes I know the problem - and I also didn't find any hint in the spec about how they plan to define the precedence with respect to ejb3 xml in relation to annotations...any news on that ? And I guess having precedence as something you can change to allow for easier migration would be a good thing. As it is now Annotations/ejb3xml wins first - correct ? But couldn't we just say hbm.xml wins first and then users can remove the hbm.xml when they want the annotations to take over ? Or will that not have any affect because the annotation process occurs before users get the chance to add their hbm.xml files... Just thinking out loud... /max When Christian migrate CaveatEmptor to HEM, the encounters issues with the discovery mechanism, because there were no notion of precedence. I think the idea of configurable precedence came from here. I have to admit I liked the idea to ease migrations. It's a bit like the BIOS boot ordering, you never used it, until you have to. The problem is that I discover them in the jar file, they are not added by the user. So I have no idea of the ordering. when will we ever have a user add both a hbm.xml AND ejb3xml file ? And if we have - again, why not just choose the first one that gets added to the configuration by the user ? /max Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
RE: [Hibernate] Configuration artifact precedence
I must admit, that this makes sense to *me* -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Jordan Laughlin Sent: Friday, 19 August 2005 2:46 PM To: hibernate-devel@lists.sourceforge.net Subject: Re: [Hibernate] Configuration artifact precedence Wouldn't it make sense for it to throw an exception if the mappings are different, otherwise who cares? In migrating, if one makes a mistake it would be good to know - "xxx has two different configurations" and not pick one which has precedence (I'm coming from the standpoint of someone who may be migrating soon). Unless, of course, the configuration artifacts are always different between those mapped by Annotations/ejb3xml and those mapped by hbm.xml. It would also definately make sense if this was configurable. I'm thinking out loud too :P -JL - Original Message - From: "Max Rydahl Andersen" <[EMAIL PROTECTED]> To: "Emmanuel Bernard" <[EMAIL PROTECTED]> Cc: Sent: Friday, August 19, 2005 5:23 AM Subject: Re: [Hibernate] Configuration artifact precedence > > Yes I know the problem - and I also didn't find any hint in the spec about > how they plan to define the precedence with respect to ejb3 xml in > relation to annotations...any news on that ? > > And I guess having precedence as something you can change to allow for > easier migration would be a good thing. > > As it is now Annotations/ejb3xml wins first - correct ? > > But couldn't we just say hbm.xml wins first and then users can remove the > hbm.xml when they want the annotations > to take over ? Or will that not have any affect because the annotation > process occurs before users get the chance to > add their hbm.xml files... > > Just thinking out loud... > > /max > > >> When Christian migrate CaveatEmptor to HEM, the encounters issues with >> the discovery mechanism, because there were no notion of precedence. I >> think the idea of configurable precedence came from here. I have to >> admit I liked the idea to ease migrations. It's a bit like the BIOS boot >> ordering, you never used it, until you have to. >> The problem is that I discover them in the jar file, they are not added >> by the user. So I have no idea of the ordering. >> >>> when will we ever have a user add both a hbm.xml AND ejb3xml file ? >>> And if we have - again, why not just choose the first one that gets >>> added >>> to the configuration by the user ? >>> >>> /max >>> >>>>> >>>>>> Well, just to let you know and to help me to find a good >>>>>> configuration >>>>>> property name. >>>>> >>>>> >>>>> >>>>> What kind of values does it take ? >>>>> >>>>> >>>>> >>>> >>>> >>>> >>>> --- >>>> SF.Net email is Sponsored by the Better Software Conference & EXPO >>>> September 19-22, 2005 * San Francisco, CA * Development Lifecycle >>>> Practices >>>> Agile & Plan-Driven Development * Managing Projects & Teams * Testing >>>> & QA >>>> Security * Process Improvement & Measurement * >>>> http://www.sqe.com/bsce5sf >>>> ___ >>>> hibernate-devel mailing list >>>> hibernate-devel@lists.sourceforge.net >>>> https://lists.sourceforge.net/lists/listinfo/hibernate-devel >>> >>> >>> >>> >>> >> > > > > > --- > SF.Net email is Sponsored by the Better Software Conference & EXPO > September 19-22, 2005 * San Francisco, CA * Development Lifecycle > Practices > Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA > Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf > ___ > hibernate-devel mailing list > hibernate-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/hibernate-devel > --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
There is no hbm/classes listing needed in a persistence archive, no possibility to comment them. Max Rydahl Andersen wrote: It's a matter of taste, but keeping your hbm files in the same place they were make a lot of sens in a migration process. from a technical POV, there is no longer hardwire processing order. Its a matter of commenting it out from the hibernate.cfg.xml or your code - correct ? (unless you are thinking about the addJar() facility) /max Just thinking out loud... /max When Christian migrate CaveatEmptor to HEM, the encounters issues with the discovery mechanism, because there were no notion of precedence. I think the idea of configurable precedence came from here. I have to admit I liked the idea to ease migrations. It's a bit like the BIOS boot ordering, you never used it, until you have to. The problem is that I discover them in the jar file, they are not added by the user. So I have no idea of the ordering. when will we ever have a user add both a hbm.xml AND ejb3xml file ? And if we have - again, why not just choose the first one that gets added to the configuration by the user ? /max Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development LifecyclePractices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel
Re: [Hibernate] Configuration artifact precedence
And how do you compare a bunch (think inheritance) of hbms and a bunch of annotated classes? Even a single cfg file and an annotated class? By creating a specific comparison code? By creating the metamodel for both and comparing them? All that work for an exception... Gavin King wrote: I must admit, that this makes sense to *me* -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Jordan Laughlin Sent: Friday, 19 August 2005 2:46 PM To: hibernate-devel@lists.sourceforge.net Subject: Re: [Hibernate] Configuration artifact precedence Wouldn't it make sense for it to throw an exception if the mappings are different, otherwise who cares? In migrating, if one makes a mistake it would be good to know - "xxx has two different configurations" and not pick one which has precedence (I'm coming from the standpoint of someone who may be migrating soon). Unless, of course, the configuration artifacts are always different between those mapped by Annotations/ejb3xml and those mapped by hbm.xml. It would also definately make sense if this was configurable. I'm thinking out loud too :P -JL - Original Message - From: "Max Rydahl Andersen" <[EMAIL PROTECTED]> To: "Emmanuel Bernard" <[EMAIL PROTECTED]> Cc: Sent: Friday, August 19, 2005 5:23 AM Subject: Re: [Hibernate] Configuration artifact precedence Yes I know the problem - and I also didn't find any hint in the spec about how they plan to define the precedence with respect to ejb3 xml in relation to annotations...any news on that ? And I guess having precedence as something you can change to allow for easier migration would be a good thing. As it is now Annotations/ejb3xml wins first - correct ? But couldn't we just say hbm.xml wins first and then users can remove the hbm.xml when they want the annotations to take over ? Or will that not have any affect because the annotation process occurs before users get the chance to add their hbm.xml files... Just thinking out loud... /max When Christian migrate CaveatEmptor to HEM, the encounters issues with the discovery mechanism, because there were no notion of precedence. I think the idea of configurable precedence came from here. I have to admit I liked the idea to ease migrations. It's a bit like the BIOS boot ordering, you never used it, until you have to. The problem is that I discover them in the jar file, they are not added by the user. So I have no idea of the ordering. when will we ever have a user add both a hbm.xml AND ejb3xml file ? And if we have - again, why not just choose the first one that gets added to the configuration by the user ? /max Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testin
RE: [Hibernate] Configuration artifact precedence
Well, Hibernate will throw an exception if you try to map the same class twice to the same entityName. -Original Message- From: Emmanuel Bernard [mailto:[EMAIL PROTECTED] Sent: Friday, 19 August 2005 5:22 PM To: Gavin King Cc: Jordan Laughlin; hibernate-devel@lists.sourceforge.net Subject: Re: [Hibernate] Configuration artifact precedence And how do you compare a bunch (think inheritance) of hbms and a bunch of annotated classes? Even a single cfg file and an annotated class? By creating a specific comparison code? By creating the metamodel for both and comparing them? All that work for an exception... Gavin King wrote: >I must admit, that this makes sense to *me* > > >-Original Message- >From: [EMAIL PROTECTED] >[mailto:[EMAIL PROTECTED] On Behalf Of >Jordan Laughlin >Sent: Friday, 19 August 2005 2:46 PM >To: hibernate-devel@lists.sourceforge.net >Subject: Re: [Hibernate] Configuration artifact precedence > >Wouldn't it make sense for it to throw an exception if the mappings are >different, otherwise who cares? In migrating, if one makes a mistake it >would be good to know - "xxx has two different configurations" and not >pick one which has precedence (I'm coming from the standpoint of >someone who may be migrating soon). Unless, of course, the >configuration artifacts are always different between those mapped by >Annotations/ejb3xml and those mapped by hbm.xml. It would also >definately make sense if this was configurable. > >I'm thinking out loud too :P > >-JL > >- Original Message - >From: "Max Rydahl Andersen" <[EMAIL PROTECTED]> >To: "Emmanuel Bernard" <[EMAIL PROTECTED]> >Cc: >Sent: Friday, August 19, 2005 5:23 AM >Subject: Re: [Hibernate] Configuration artifact precedence > > > > >>Yes I know the problem - and I also didn't find any hint in the spec >> >> >about > > >>how they plan to define the precedence with respect to ejb3 xml in >>relation to annotations...any news on that ? >> >>And I guess having precedence as something you can change to allow for >> >> > > > >>easier migration would be a good thing. >> >>As it is now Annotations/ejb3xml wins first - correct ? >> >>But couldn't we just say hbm.xml wins first and then users can remove >> >> >the > > >>hbm.xml when they want the annotations to take over ? Or will that not >>have any affect because the annotation >> >> > > > >>process occurs before users get the chance to add their hbm.xml >>files... >> >>Just thinking out loud... >> >>/max >> >> >> >> >>>When Christian migrate CaveatEmptor to HEM, the encounters issues >>> >>> >with > > >>>the discovery mechanism, because there were no notion of precedence. >>> >>> >I > > >>>think the idea of configurable precedence came from here. I have to >>>admit I liked the idea to ease migrations. It's a bit like the BIOS >>> >>> >boot > > >>>ordering, you never used it, until you have to. >>>The problem is that I discover them in the jar file, they are not >>> >>> >added > > >>>by the user. So I have no idea of the ordering. >>> >>> >>> >>>>when will we ever have a user add both a hbm.xml AND ejb3xml file ? >>>>And if we have - again, why not just choose the first one that gets >>>>added to the configuration by the user ? >>>> >>>>/max >>>> >>>> >>>> >>>>>>>Well, just to let you know and to help me to find a good >>>>>>>configuration property name. >>>>>>> >>>>>>> >>>>>> >>>>>>What kind of values does it take ? >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>>>--- >>>>>SF.Net email is Sponsored by the Better Software Conference & EXPO >>>>>September 19-22, 2005 * San Francisco, CA * Development Lifecycle >>>>>Practices Agile & Plan-Driven Development * Managing Projects & >>>>>Teams * >>>>> >>>>> >Testing > > >>>>>& QA >>>>>Security * Proces
Re: [Hibernate] Configuration artifact precedence
Well I guess if it's too difficult or slow to compare them, then issuing a warning along with the action Hibernate took (configurable somewhere) seems fine. -JL - Original Message - From: "Emmanuel Bernard" <[EMAIL PROTECTED]> To: "Gavin King" <[EMAIL PROTECTED]> Cc: "Jordan Laughlin" <[EMAIL PROTECTED]>; Sent: Friday, August 19, 2005 2:21 PM Subject: Re: [Hibernate] Configuration artifact precedence And how do you compare a bunch (think inheritance) of hbms and a bunch of annotated classes? Even a single cfg file and an annotated class? By creating a specific comparison code? By creating the metamodel for both and comparing them? All that work for an exception... Gavin King wrote: I must admit, that this makes sense to *me* -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Jordan Laughlin Sent: Friday, 19 August 2005 2:46 PM To: hibernate-devel@lists.sourceforge.net Subject: Re: [Hibernate] Configuration artifact precedence Wouldn't it make sense for it to throw an exception if the mappings are different, otherwise who cares? In migrating, if one makes a mistake it would be good to know - "xxx has two different configurations" and not pick one which has precedence (I'm coming from the standpoint of someone who may be migrating soon). Unless, of course, the configuration artifacts are always different between those mapped by Annotations/ejb3xml and those mapped by hbm.xml. It would also definately make sense if this was configurable. I'm thinking out loud too :P -JL - Original Message - From: "Max Rydahl Andersen" <[EMAIL PROTECTED]> To: "Emmanuel Bernard" <[EMAIL PROTECTED]> Cc: Sent: Friday, August 19, 2005 5:23 AM Subject: Re: [Hibernate] Configuration artifact precedence Yes I know the problem - and I also didn't find any hint in the spec about how they plan to define the precedence with respect to ejb3 xml in relation to annotations...any news on that ? And I guess having precedence as something you can change to allow for easier migration would be a good thing. As it is now Annotations/ejb3xml wins first - correct ? But couldn't we just say hbm.xml wins first and then users can remove the hbm.xml when they want the annotations to take over ? Or will that not have any affect because the annotation process occurs before users get the chance to add their hbm.xml files... Just thinking out loud... /max When Christian migrate CaveatEmptor to HEM, the encounters issues with the discovery mechanism, because there were no notion of precedence. I think the idea of configurable precedence came from here. I have to admit I liked the idea to ease migrations. It's a bit like the BIOS boot ordering, you never used it, until you have to. The problem is that I discover them in the jar file, they are not added by the user. So I have no idea of the ordering. when will we ever have a user add both a hbm.xml AND ejb3xml file ? And if we have - again, why not just choose the first one that gets added to the configuration by the user ? /max Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.so
Re: [Hibernate] Configuration artifact precedence
Then you reduce the usability of auto discovery. Gavin King wrote: Well, Hibernate will throw an exception if you try to map the same class twice to the same entityName. -Original Message- From: Emmanuel Bernard [mailto:[EMAIL PROTECTED] Sent: Friday, 19 August 2005 5:22 PM To: Gavin King Cc: Jordan Laughlin; hibernate-devel@lists.sourceforge.net Subject: Re: [Hibernate] Configuration artifact precedence And how do you compare a bunch (think inheritance) of hbms and a bunch of annotated classes? Even a single cfg file and an annotated class? By creating a specific comparison code? By creating the metamodel for both and comparing them? All that work for an exception... Gavin King wrote: I must admit, that this makes sense to *me* -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Jordan Laughlin Sent: Friday, 19 August 2005 2:46 PM To: hibernate-devel@lists.sourceforge.net Subject: Re: [Hibernate] Configuration artifact precedence Wouldn't it make sense for it to throw an exception if the mappings are different, otherwise who cares? In migrating, if one makes a mistake it would be good to know - "xxx has two different configurations" and not pick one which has precedence (I'm coming from the standpoint of someone who may be migrating soon). Unless, of course, the configuration artifacts are always different between those mapped by Annotations/ejb3xml and those mapped by hbm.xml. It would also definately make sense if this was configurable. I'm thinking out loud too :P -JL - Original Message - From: "Max Rydahl Andersen" <[EMAIL PROTECTED]> To: "Emmanuel Bernard" <[EMAIL PROTECTED]> Cc: Sent: Friday, August 19, 2005 5:23 AM Subject: Re: [Hibernate] Configuration artifact precedence Yes I know the problem - and I also didn't find any hint in the spec about how they plan to define the precedence with respect to ejb3 xml in relation to annotations...any news on that ? And I guess having precedence as something you can change to allow for easier migration would be a good thing. As it is now Annotations/ejb3xml wins first - correct ? But couldn't we just say hbm.xml wins first and then users can remove the hbm.xml when they want the annotations to take over ? Or will that not have any affect because the annotation process occurs before users get the chance to add their hbm.xml files... Just thinking out loud... /max When Christian migrate CaveatEmptor to HEM, the encounters issues with the discovery mechanism, because there were no notion of precedence. I think the idea of configurable precedence came from here. I have to admit I liked the idea to ease migrations. It's a bit like the BIOS boot ordering, you never used it, until you have to. The problem is that I discover them in the jar file, they are not added by the user. So I have no idea of the ordering. when will we ever have a user add both a hbm.xml AND ejb3xml file ? And if we have - again, why not just choose the first one that gets added to the configuration by the user ? /max Well, just to let you know and to help me to find a good configuration property name. What kind of values does it take ? --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf ___ hibernate-devel mailing list hibernate-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hibernate-devel --- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 200