Hi all,
thanks for putting this discussion up ( I also started a similar one on private
which did not gain traction at all but I think its better to discuss it here,
either way).
For podlings there is no such thing as the attic so the natural way would be to
move it over to Github and we could simplay grant full access to all current
PPMC members.
This would not change anything, I guess as we would still be able to make
releases (under github not ASF Policy then) and keep on working but as the
github infra is quite nice (with issues and wiki and all) I don’t think that it
would suffer much.
My honest opinion is that it will be extremely hard to get this podling to
graduate anytime soon and that I currently don’t see that. So this could be a
valid alternative.
But ofc that’s just my opinion.
Julian
Am 19.01.20, 21:41 schrieb "Werner Keil" <werner.k...@gmail.com>:
Phil/all,
Thanks for the discussion.
I'm not sure, to what extent the ASF can really boost the project's
popularity and adoption?
There have been many uncertainties like the whole Config JSR issue,
although that's now dead, but it still isn't completely sure, if
MicroProfile Config would simply take its place or something inspired by it
as well as other approaches like Oracle's Helidon?
Most of it won't be effective until Jakarta EE 10, so another year since 9
with its refactoring job isn't even scheduled, but it may come out around a
year after Jakarta EE 8 (Sept '19) so with a possible annual release train
like Eclipse had before this would be around Sep 2021 at the earliest.
Until then the only thing would be to show TCK compatibility of MP Config
and (not so different from Helidon) also provide some more powerful
alternatives, including HOCON/Typesafe Config support.
"Attic" isn't that where Tamaya would end up anyway, should it be
discontinued?
I can tell from the experience of OpenDDR - DeviceMap (which graduated) and
back to its original brand, that this has not really hurt the project.
On the contrary, while we did not track downloads much in 2017, the second
half of 2018 alone (also did not watch it before) saw a massive boost in
downloads of the "device data" artifact up to 500 or 600k in 2018.
Last year it went back a bit, but it was still around 150k in 2019, roughly
10-12k per month on average (though it was not so evenly spread in reality)
We could not get download stats from apache.org with the argument, that
mirrors would not make that possible, so only the Maven repo downloads were
available which never exceeded 200-250 in the peak days of Apache
DeviceMap. Even in summer 2019 with a holiday gap OpenDDR data saw more
downloads.
As I have responsibilities for several Jakarta EE projects and the
platform/Spec Committee, I also can't promise to contribute much.
Beside Helidon which looks very healthy with steady contributions by Oracle
and a few non-Oracle committers, Apache alone has 2 efforts:
https://github.com/apache/commons-configuration (also extremely active and
healthy)
https://github.com/apache/geronimo-config an independent implementation of
MP Config which seems practically abandoned for about a year, most likely
because of several other implementations including SmallRye, Thorntail,
Quarkus or Helidon.
So I really can't say, if there is enough momentum to keep it alive in its
current form. And simply putting it on GH, may not make much difference
either.
Which commercial clients are using Tamaya at the moment? And would they
suffer, if it was archived as is?
Looking at all the different MP implementations alone, not to mention their
much bigger competitors like Spring or Micronaut (all of them contain
configuration of their own) I am really skeptical, that it makes sense
unless a major increase in contributors and contributions occur.
Maybe it could be better for those with a vested interest in configuration
to help either the mentioned Apache projects or the others outside ASF.
WDYT?
Werner
On Sun, Jan 19, 2020 at 8:13 PM P. Ottlinger <pottlin...@apache.org> wrote:
> Hi,
>
> Am 17.01.20 um 20:57 schrieb Anatole Tresch:
> > Honestly, I am opem to anything. Time actually is rare and I will not
> have
> > much in the next months either. We can think to move everything to GH
and
> > continue there as time allows...
>
> The disadvantage of that would be we loose the umbrella of ASF with all
> its infrastructure and regulations /such as attic/.
>
> > There is not much innovation in this topic to ne expected either.
> >
> > Other opinions?
>
> If we graduate we need to establish some sort of momentum and definitely
> need more contributions.
>
> Anatole and me not having that much time in the coming weeks means
> roughly no changes at all ....
>
> What do other contributors/list members think?
>
> Cheers,
> Phil
>
