Re: [agentic-ai-dev] Maven Coordinates/Package Names

[Reza Rahman <reza_rahman@xxxxxxxx>]

If the initial email and Andrew’s pretty thorough explanation is not sufficiently clear already, I suggest looking at the code. It mirrors both sets of information and does not have much room for ambiguity.

Everything is pretty much exactly what the dominant Jakarta norms/patterns are. There’s no real value proposition to deviate too much. As Andrew explains, we are actually perfectly fine to choose a very simple group ID too.

> Sent: Tuesday, March 10, 2026 at 4:25 PM
> From: "Estuardo Wyss via agentic-ai-dev" <agentic-ai-dev@xxxxxxxxxxx>
> To: "agentic-ai developer discussions" <agentic-ai-dev@xxxxxxxxxxx>
> Cc: "Estuardo Wyss" <estuardo.wyss@xxxxxxxxx>
> Subject: Re: [agentic-ai-dev] Maven Coordinates/Package Names
> 
> Hi there:
> 
> using next sintaxis are we repeating jakarta:
> 
> *jakarta*/
> └── ai
>     ├── agent (or whatever)
>     │   ├── *jakarta*.agentic-ai-api
>     │   └── *jakarta*-agentic-ai-tck
>     └── xy
>         ├── *jakarta*.xy-ai-api
>         └── *jakarta*.xy-ai-tck
> 
> El mar, 10 mar 2026 a la(s) 8:43 a.m., Piotr Żygieło via agentic-ai-dev (
> agentic-ai-dev@xxxxxxxxxxx) escribió:
> 
> > On Tue, 10 Mar 2026 at 14:52, Andrew Pielage via agentic-ai-dev
> > <agentic-ai-dev@xxxxxxxxxxx> wrote:
> > > So unless I'm overlooking something there's no explicit naming rules for
> > maven coordinates on the side of the spec committee that would prevent
> > using jakarta.ai as the group ID
> >
> > Thanks for checking.
> >
> > 1.a As I understand it - the proposed namespace on the sonatype side
> > would be jakarta.ai.
> >     Is that correct?
> >
> > 1.b OR - it would be several namespaces instead:
> > jakarta.ai.jakarta.agentic-ai-api, jakarta.ai.jakarta-agentic-ai-tck,
> > jakarta-agentic-ai-examples (which now are few already) ...
> >
> > 2. The structure of artifacts then:
> > jakarta/
> > └── ai
> >     ├── jakarta.agentic-ai-api
> >     └── jakarta-agentic-ai-tck
> >
> > (BTW - why TCK coordinates separate jakarta and agentic-ai-tck with
> > dash? Inject uses dot in both cases:
> > jakarta/inject/
> > ├── jakarta.inject-api
> > └── jakarta.inject-tck
> > )
> >
> > If 1a is correct - and there is *new* "XY AI" Spec to come - would it
> > share the `jakarta.ai` Sonatype namespace?
> >
> > Would this structure be expected one?
> > > jakarta-llm or jakarta-mcp along with
> > > jakarta-agentic-ai
> >
> > Suggests to me something like:
> >
> > jakarta/
> > └── ai
> >     ├── jakarta.agentic-ai-api
> >     ├── jakarta-agentic-ai-tck
> >     ├── jakarta.xy-ai-api
> >     └── jakarta.xy-ai-tck
> >
> > It might look fine, but...
> > I'm sure it's possible and I'm also pretty sure it's not desirable to
> > share access to namespace between SEPARATE specifications. Then it
> > shall be the hell of 1b.
> >
> > That's why I'd like to see jakarta.ai partitioned up front and have
> > (like is for jakarta/xml/)
> > jakarta/
> > └── ai
> >     ├── agent (or whatever)
> >     │   ├── jakarta.agentic-ai-api
> >     │   └── jakarta-agentic-ai-tck
> >     └── xy
> >         ├── jakarta.xy-ai-api
> >         └── jakarta.xy-ai-tck
> >
> > For N specs to have N namespaces (and not N*Z_i) to organize within
> > like the specs desire.
> >
> > But perhaps it's only me.
> >
> > > this will enable us to place all future AI work
> > > into one neat place such as jakarta-llm or jakarta-mcp along with
> > > jakarta-agentic-ai)
> >
> > I only wanted to challenge "one neat place", as I'm not confident that
> > would be the case.
> >
> > --
> > Piotrek
> > _______________________________________________
> > agentic-ai-dev mailing list
> > agentic-ai-dev@xxxxxxxxxxx
> > To unsubscribe from this list, visit https://accounts.eclipse.org
> >
> _______________________________________________
> agentic-ai-dev mailing list
> agentic-ai-dev@xxxxxxxxxxx
> To unsubscribe from this list, visit https://accounts.eclipse.org