Re: [cross-project-issues-dev] [Maven/Tycho] Suggested groupId

[Mickael Istria <mickael.istria@xxxxxxxxxxxxxx>]

It seems like we should have different conventions depending on whether we build "simple jars" or OSGi-bundles.
For jetty as a jar, that could be
   <groupId>org.eclipse.jetty</groupId>
   <artifactId>jetty-server</artifactId>
but for an OSGi bundle/Eclipse plugin
   <groupId>org.eclipse.jetty</groupId>
   <artifactId>org.eclipse.jetty.server</artifactId>
...

I think that the debate is open wide, and this new problem (standard Java conventions and OSGi conventions that don't fit together) is not specific to Eclipse, but more to the whole Java world.
Please note that if we want to keep the bundle naming convention of Eclipse when using Tycho, we dont have choice: artifactId == Bundle-SymbolicName
That said, let's constrain the problem to my use case ;) For a SWTBot bundle, should I use
   <groupId>org.eclipse.swtbot</groupId>
   <artifactId>org.eclipse.swtbot.eclipse.finder</groupId>
Or
   <groupId>org.eclipse</groupId>
   <artifactId>org.eclipse.swtbot.eclipse.finder</groupId>
??

Le 10/11/2010 18:51, Jesse McConnell a écrit :

perhaps I should make it a bit clearer though, our normal release does
not using tycho, it uses the bundle-maven-plugin to produce osgi
bundles/jar files...

the features and plugins we work on that build using tycho do follow
the long artifactId conventions

jesse

--
jesse mcconnell
jesse.mcconnell@xxxxxxxxx



On Wed, Nov 10, 2010 at 11:49, Jesse McConnell
<jesse.mcconnell@xxxxxxxxx> wrote:
  

normal people use org.eclipse.<project> as the groupId and an
artifactId of something like jetty-server

anything other then that is not following normal maven conventions...

that being said there has been discussion about using that
org.eclipse:org.eclipse.jetty.jetty-server:<version> notation to allow
for reverse mappings to bundles...

Also, I'll just note that jetty maven artifacts are not signed in
maven central, we generate our signed artifacts afterwards as part of
a separate build process that operates off of the released tag and
generates the bits and pieces that helios and indigo builds consume.

the p2 repositories we publish do contain signed artifacts and follow
eclipse requirements

cheers,
jesse

--
jesse mcconnell
jesse.mcconnell@xxxxxxxxx



On Wed, Nov 10, 2010 at 11:37, Jeff McAffer <jeff@xxxxxxxxxxxxxxxxx> wrote:
    

There is also confusion over the artifact id.  Mostly I've seen people using org.eclipse.foo.bar (ie. the fully artifact file name without version and .jar) as the artifact name.  However, Chris points out that egit is using some other convention.

Jeff

On 2010-11-10, at 12:19 PM, Stephan Herrmann wrote:

      

So according to https://bugs.eclipse.org/bugs/show_bug.cgi?id=288644
the groupId must be org.eclipse and thus org.eclipse.egit is illegal.
But also several comments on https://bugs.eclipse.org/bugs/show_bug.cgi?id=283745
speak for org.eclipse.<project>, both as a wish and as the status-quo.

It seems, not everybody is convinced that the decision in bug 288644
is the way to go. And not everybody knows about the decision.

Stephan

On Wednesday, November 10, 2010 06:08:11 pm Oberhuber, Martin wrote:
        

See also

https://bugs.eclipse.org/bugs/show_bug.cgi?id=288644
<https://bugs.eclipse.org/bugs/show_bug.cgi?id=288644>
"Decide of group ids for Eclipse Maven artifacts"

Thanks,

--
Martin Oberhuber, Senior Member of Technical Staff, Wind River
direct +43.662.457915.85  fax +43.662.457915.6


________________________________

From: cross-project-issues-dev-bounces@xxxxxxxxxxx
[mailto:cross-project-issues-dev-bounces@xxxxxxxxxxx] On Behalf Of Chris
Aniszczyk
Sent: Mittwoch, 10. November 2010 18:06
To: Cross project issues
Subject: Re: [cross-project-issues-dev] [Maven/Tycho] Suggested groupId


On Wed, Nov 10, 2010 at 12:03 PM, Mickael Istria
<mickael.istria@xxxxxxxxxxxxxx> wrote:


     Hi all,

     I am working a little bit on trying Tycho for SWTBot build. And
I got a question of "good practice", if good practices with Tycho
already exist. The question is: What should be the groupId ?
     IMHO, the groupId should reflect the Bundle-Vendor / Provider
attribute, and AFAIAA the release train says the Provider should
"Eclipse.org - [Project]", for example "Eclipse.org - SWTBot". Then, I'd
go for org.eclipse.swtbot. Does this sound good, or should I prefer
org.eclipse ?



In EGit/JGit land we use 'org.eclipse.<project>'

 <groupId>org.eclipse.egit</groupId>
 <artifactId>egit-parent</artifactId>

So I'm for that approach.


          

_______________________________________________
cross-project-issues-dev mailing list
cross-project-issues-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/cross-project-issues-dev
        

_______________________________________________
cross-project-issues-dev mailing list
cross-project-issues-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/cross-project-issues-dev

      

_______________________________________________
cross-project-issues-dev mailing list
cross-project-issues-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/cross-project-issues-dev
  

--

Mickael Istria
R&D Engineer

email : mickael.istria@xxxxxxxxxxxxxx

This message and any attachment (the "message") is intended solely for the addressees and is confidential. If you receive this message by mistake, please delete it and notify the sender immediately. Any use not in accordance with its purpose, any out-spread or disclosure, either as a whole or partially, is prohibited except with formal approval. Internet cannot guarantee the integrity of this message, therefore BonitaSoft will not be liable for the message if modified.