Thanks Reto. Hi all, I suggest the following steps for the next release: 1. Merge changes in parent from reunited to master and release parent 8 2. Merge all other modules from reunited to master and release them based on parent 8 3. Update some relevant packages in https://github.com/clerezza to use new master 4. Refactor schemagen and ontologies module and release them
What do you think? Cheers Hasan On Mon, Mar 25, 2019 at 1:16 PM Reto Gmür <[email protected]> wrote: > Hi Hasan > > I’ve pushed respective changes to the reunited branch. In my understanding > calling the method “reset” instead of “unset” would not be a problem but > just require some parameterization of @Reference. But I kept it as it is, > as the method will seldomly be “manually” invoked anyway. > > Cheers, > Reto > > From: Hasan <[email protected]> > Sent: Saturday, March 23, 2019 8:39 AM > To: [email protected]; Reto Gmür <[email protected]> > Subject: Re: [DISCUSS] Release the reunited branch as the master branch > version 8 > > I think we have to stick to the OSGI spec and use unsetXXX and not using > resetXXX. > (112.8.1 Component Annotations) > @Reto Gmür<mailto:[email protected]>, you can update the module as > suggested. And yes, I think it needs an argument. > > Thanks > Hasan > > On Mon, Mar 18, 2019 at 11:51 AM Reto Gmür <[email protected]<mailto: > [email protected]>> wrote: > I looked at the OSGi DS spec section 112.3.2 and it says "An event method > can take one or more parameters." - Which is quite dull given their example > just ignores the parameter: void unsetLog( LoggerFactory l ) { lf = null; } > > Reto > > -----Original Message----- > From: Hasan <[email protected]<mailto:[email protected]>> > Sent: Monday, March 18, 2019 6:24 AM > To: [email protected]<mailto:[email protected]> > Subject: Re: [DISCUSS] Release the reunited branch as the master branch > version 8 > > Hi Reto, all, > > On Thu, Mar 14, 2019 at 2:22 PM Reto Gmür <[email protected]<mailto: > [email protected]>> wrote: > > > Hi Hasan > > > > Yes, the renaming you propose makes sense to me. Are both names > > supported by the plugin generating the files in OSGI-INF or is some > > special annotation needed? Does the method need to have an argument? > > > > Sorry, I don't know. Maybe you can try or check? > > Hasan > > > > Cheers, > > Reto > > > > -----Original Message----- > > From: Hasan <[email protected]<mailto:[email protected]>> > > Sent: Wednesday, March 13, 2019 8:28 PM > > To: [email protected]<mailto:[email protected]> > > Subject: Re: [DISCUSS] Release the reunited branch as the master > > branch version 8 > > > > Hi Reto > > > > Without a serializer instance, a GraphWriter will have no use. Thus, I > > don't see a disadvantage of the code proposed. > > Not sure though whether the name unsetSerializer should be replaced > > with resetSerializer which would be more suitable in this case, > > because we are going to reset it to a default serializer rather than set > it to null. > > > > Cheers > > Hasan > > > > On Mon, Mar 11, 2019 at 10:19 AM Reto Gmür <[email protected] > <mailto:[email protected]>> wrote: > > > > > Hi Hasan > > > > > > Regarding GraphWriter: I created a branch reunited-non-ds-mbw to > > > illustrate what I mean. Do you see a disadvantage with that code > > proposal? > > > The same could be applied to GraphReader. > > > > > > Cheers, > > > Reto > > > > > > -----Original Message----- > > > From: Hasan <[email protected]<mailto:[email protected]>> > > > Sent: Sunday, March 10, 2019 3:59 PM > > > To: [email protected]<mailto:[email protected]> > > > Subject: Re: [DISCUSS] Release the reunited branch as the master > > > branch version 8 > > > > > > Hi Reto > > > > > > Thx for the input. > > > Comments inline > > > > > > On Sat, Mar 9, 2019 at 5:36 PM Reto Gmür <[email protected] > <mailto:[email protected]>> wrote: > > > > > > > Hi Hasan, all, > > > > > > > > I've just created a branch of the jena.* modules on > > > > https://github.com/clerezza and of > > > > https://github.com/linked-solutions/slds. > > > > > > > Great. > > > I assume we will fix the module version in the reunited branch to > 2.0.0. > > > Before that I'd like to have the parent version be changed from > > > 8-SNAPSHOT to 8. Do you agree? > > > What can we remove from the parent pom.xml and what can we update ? > > > > > > > > > > > The refactoring brings some good improvements so I'd like to see > > > > this released as soon as possible. > > > > > > > > > > This also means to make this the new master. > > > > > > > > > > In my opinion discussing the following points should nor block the > > > > release, but maybe we find a consensus easily so that we could > > > > incorporate this in the release > > > > > > > > - Do we need the .api package? Couldn't these types be directly in > > > > o.a.clerezza? After all Clerezza is mainly an API > > > > > > > So, what's the name of this artifact in the group org.apache.clerezza? > > > > > > - I don't like the name .api.impl - Without the ".api" it would be a > > > bit > > > > better, still ".impl" is just very unspecific. > > > > > > > > > > If we know the name of the artifact in the previous question, eg > > > foo, we can call it foo.impl. > > > Or do I misunderstand something here? > > > > > > - It's great the the method in GraphWriter to set the Serializer is > > > now > > > > public. Now I can have the code: > > > > GraphWriter graphWriter = new GraphWriter(); > > > > graphWriter.setSerializer(Serializer.getInstance()); > > > > Before I needed to make a subclass to access the protected methods. > > > > However it seems that graphWriter could access the default > > > > serializer using .getInstance itself, if none is set. > > > > > > > > > > I think current master branch already has public method > setSerializer(). > > > > > > Cheers > > > Hasan > > > > > > > > > > Cheers, > > > > Reto > > > > > > > > > > > > -----Original Message----- > > > > From: Reto Gmür <[email protected]<mailto:[email protected] > >> > > > > Sent: Thursday, March 7, 2019 2:36 PM > > > > To: [email protected]<mailto:[email protected]> > > > > Subject: RE: [DISCUSS] Release the reunited branch as the master > > > > branch version 8 > > > > > > > > Hi Hasan > > > > > > > > Thanks a lot for the overview. > > > > > > > > While I think it shouldn't be a general requirement to release > > > > everything together, in this case it looks like it would make > > > > things easier to use version 2.0.0 in all modules. WDYT? > > > > > > > > Cheers, > > > > Reto > > > > > > > > -----Original Message----- > > > > From: Hasan <[email protected]<mailto:[email protected]>> > > > > Sent: Wednesday, March 6, 2019 9:54 PM > > > > To: [email protected]<mailto:[email protected]> > > > > Subject: Re: [DISCUSS] Release the reunited branch as the master > > > > branch version 8 > > > > > > > > Dear all > > > > > > > > I had a discussion with Reto regarding the versioning of the modules. > > > > Currently, all modules have version 8 after the refactoring. Reto > > > > suggested to keep previous version and increase it according to > > > > semantic versioning. > > > > So, since all modules have a breaking change, I will increase the > > > > major number. > > > > However, some modules are new or the result of a renaming. > > > > > > > > We have these modules in the reunited branch: > > > > > > > > * api (was org.apache.clerezza.commons-rdf:commons-rdf-api > > > > 0.3-SNAPSHOT) > > > > * api.impl (was > > > > org.apache.clerezza.commons-rdf:commons-rdf-impl-utils > > > > 0.3-SNAPSHOT) > > > > * ontologies (was org.apache.clerezza:rdf.ontologies > > > > 1.0.1-SNAPSHOT) > > > > * sparql (was sparql package in org.apache.clerezza:rdf.core > > > > 1.0.2-SNAPSHOT) > > > > * representation (was serializedform package in > > > > org.apache.clerezza:rdf.core 1.0.2-SNAPSHOT) > > > > * test.utils (was org.apache.clerezza:rdf.core.test > > > > 1.0.1-SNAPSHOT) > > > > * dataset (was access package in org.apache.clerezza:rdf.core > > > > 1.0.2-SNAPSHOT) > > > > * api.utils (is a merge of org.apache.clerezza:rdf.utils > > > > 1.0.1-SNAPSHOT with org.apache.clerezza:rdf.scala.utils > > > > 1.0.1-SNAPSHOT) > > > > * jaxrs.rdf.providers (was org.apache.clerezza:jaxrs.rdf.providers > > > > 1.0.1-SNAPSHOT) > > > > > > > > api will have version 1.0.0 > > > > api.impl will have version 1.0.0 > > > > ontologies will have version 2.0.0 sparql will have version 2.0.0 > > > > representation will have version 2.0.0 test.utils will have > > > > version > > > > 2.0.0 dataset will have version 2.0.0 api.utils will have version > > > > 2.0.0 jaxrs.rdf.providers will have version 2.0.0 > > > > > > > > What do you think? > > > > Any objections? > > > > > > > > Kind regards > > > > Hasan > > > > > > > > On Sat, Feb 16, 2019 at 11:12 AM Hasan <[email protected]<mailto: > [email protected]>> wrote: > > > > > > > > > Dear all > > > > > > > > > > The reunited branch of Clerezza ( > > > > > https://gitbox.apache.org/repos/asf/clerezza.git) brings back > > > > > the Clerezza common-rdf ( > > > > > https://gitbox.apache.org/repos/asf/clerezza-rdf-core.git) > > > > > into Clerezza (https://gitbox.apache.org/repos/asf/clerezza.git). > > > > > > > > > > At the same time we refactor Clerezza to > > > > > - remove a cyclic dependency between sparql and access package, > > > > > - have all core functionalities in the top level modules instead > > > > > of under a single rdf module, > > > > > - rename some modules and packages to reflect better their > > > functionality. > > > > > > > > > > The refactoring task is more or less complete. > > > > > > > > > > I think we should release the reunited branch in the near future > > > > > as version 8 of master. > > > > > > > > > > What do you think? > > > > > > > > > > Kind regards > > > > > Hasan > > > > > > > > > > > > > > > > > > > >
