Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-15 Thread Emmanuel Bernard
On Mon 2012-10-15 18:19, Hardy Ferentschik wrote: > > Basically for you a Service must allow for multiple implementations > > and a ServiceManager must offer a way to switch between various > > implementations. > > I orientate myself in this case at the ServiceLoader docs (especially since > we a

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-15 Thread Hardy Ferentschik
On 12 Jan 2012, at 5:29 PM, Emmanuel Bernard wrote: > On Fri 2012-10-12 15:13, Hardy Ferentschik wrote: >> >> On 12 Jan 2012, at 3:00 PM, Steve Ebersole wrote: >> >>> "Java services api" == ServiceLoader I assume? >> >> correct >> >>> Going on that assumption: >>> >>> No. ServiceLoader is

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Emmanuel Bernard
On Fri 2012-10-12 15:13, Hardy Ferentschik wrote: > > On 12 Jan 2012, at 3:00 PM, Steve Ebersole wrote: > > > "Java services api" == ServiceLoader I assume? > > correct > > > Going on that assumption: > > > > No. ServiceLoader is just a discovery mechanism. There still needs to be > > some

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Emmanuel Bernard
On Fri 2012-10-12 14:03, Sanne Grinovero wrote: > What you say makes sense, but why to we have an > JavaSerializationSerializationProvider then? > It's an alternative implementation which is not using Avro. > > AFAIK there currently is no way to select that one, hence the TODO I > had added recent

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Sanne Grinovero
On 12 October 2012 14:49, Hardy Ferentschik wrote: > It's likely that this will evolve; especially the Avro picking stuff you mention.. doesn't look like we finished that; in fact there is a TODO in SerializationProviderService. I'll check JIRA to see if we're tracking that. >>

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Hardy Ferentschik
>>> It's likely that this will evolve; especially the Avro picking stuff >>> you mention.. doesn't look like we finished that; in fact there is a >>> TODO in SerializationProviderService. I'll check JIRA to see if we're >>> tracking that. >> >> Right. I think we need to sort this out. Right now t

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Sanne Grinovero
On 12 October 2012 14:29, Hardy Ferentschik wrote: > > On 12 Jan 2012, at 2:58 PM, Sanne Grinovero wrote: > >> I was having similar doubts when recently converted the Serializer >> service to a Service. > > good :-) > >> The ServiceProvider can contain logic to make a choice about which >> Service

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Sanne Grinovero
On 12 October 2012 14:13, Hardy Ferentschik wrote: > > On 12 Jan 2012, at 3:00 PM, Steve Ebersole wrote: > >> "Java services api" == ServiceLoader I assume? > > correct > >> Going on that assumption: >> >> No. ServiceLoader is just a discovery mechanism. There still needs to be >> something tha

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Hardy Ferentschik
On 12 Jan 2012, at 2:58 PM, Sanne Grinovero wrote: > I was having similar doubts when recently converted the Serializer > service to a Service. good :-) > The ServiceProvider can contain logic to make a choice about which > Service implementation you're supposed to get, as it receives the > con

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Hardy Ferentschik
On 12 Jan 2012, at 3:00 PM, Steve Ebersole wrote: > "Java services api" == ServiceLoader I assume? correct > Going on that assumption: > > No. ServiceLoader is just a discovery mechanism. There still needs to be > something that, as you say, negotiates amongst the various discovered > imp

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Steve Ebersole
"Java services api" == ServiceLoader I assume? Going on that assumption: No. ServiceLoader is just a discovery mechanism. There still needs to be something that, as you say, negotiates amongst the various discovered implementations of a particular service. 2 well known ServiceLoader uses ar

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Sanne Grinovero
On 12 October 2012 13:55, Emmanuel Bernard wrote: > On Fri 2012-10-12 14:23, Hardy Ferentschik wrote: >> Hi, >> >> as part of my investigations for HSEARCH-1025 and HSEARCH-1026 I had a look >> at how services are implemented in Search. >> I thought I could make the statistics collector also a se

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Sanne Grinovero
Hi Hardy, I was having similar doubts when recently converted the Serializer service to a Service. The ServiceProvider can contain logic to make a choice about which Service implementation you're supposed to get, as it receives the configuration properties. In short how different alternatives are

Re: [hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Emmanuel Bernard
On Fri 2012-10-12 14:23, Hardy Ferentschik wrote: > Hi, > > as part of my investigations for HSEARCH-1025 and HSEARCH-1026 I had a look > at how services are implemented in Search. > I thought I could make the statistics collector also a service. Looking at > the code I am a little confused thou

[hibernate-dev] [HSearch] ServiceManager and services

2012-10-12 Thread Hardy Ferentschik
Hi, as part of my investigations for HSEARCH-1025 and HSEARCH-1026 I had a look at how services are implemented in Search. I thought I could make the statistics collector also a service. Looking at the code I am a little confused though. Let's look at the different pieces. First ServiceManager