Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [orbit-dev] Declaring Build: S20161205183421

Yes, they are at odds. IMHO (and it is just one opinion) we should "mimic"
what the third-party provides. Not what some "fourth-party" provides. :)

Or, at least ask them, on their mailing list if they contributed them split-up
to maven central, or if they have any preference.

And, since the third party ships them all-together in one OSGi bundle, and since we have done that previously, that would seem to tip the balance in favor of all-in-one?

My guess (and just a guess, I've not looked) is there is not much difference in "size" of the bundles.
But I realize some would argue that more modular is always better.

Thanks for the history and explanations. I will leave up to you now and comment no more (on this case :)




On 12/08/2016 03:11 PM, Roland Grunberg wrote:
Yes, I see that the sources/binaries are shipped together on that page.
However, on maven central everything is split up. There's httpclient-cache,
httpclient-fluent, etc. These 2 approaches are at odds but I'd argue more
people are probably consuming through maven so diverging from that packaging
might be confusing.

As CQs, they were also filed separately (CQ 11712, 11713) likely because
they were based off of the maven built artifacts. Since they're already
approved it shouldn't be too much work to get them into orbit-recipes.


Cheers,




Back to the top