Looks good Roland - I assume Pradeepto and the team maintaining fabric8
maven plugin is already in on it (cc devtools-build@)
/max
http://about.me/maxandersen
On 14 Jun 2018, at 7:59, Roland Huss wrote:
Hey Andrew,
exactly that idea is laid out in
https://github.com/fabric8io/fabric8-build where
we from the Fuse team (Nicola and myself) are planning to continue
there
soon after we discussed this on our F2F next week. The idea to extract
the
main abstractions (Generators, Enrichers, Profiles) into Maven
independent
code, also fixing issues in the way how resources descriptors are
created
(i.e. two separates passes through the pipelines for
Kubernetes/Openshift
separately instead of running only for Kubernetes and the mangling
that
resources as an afterthought for OpenShift objects).
This basic blocks should then be easily consumable by a Maven or
Gradle
plugin, any IDE and also Arquillian (cc Alex).
Happy about any feedback and contributions.
regards ...
... roland
On Thu, Jun 14, 2018 at 2:33 AM Andrew Lee Rubinger <a...@redhat.com>
wrote:
On Wed, Jun 13, 2018 at 7:51 PM, Burr Sutter <bsut...@redhat.com>
wrote:
On Wed, Jun 13, 2018 at 6:15 PM Gorkem Ercan <ger...@redhat.com>
wrote:
On 13 Jun 2018, at 16:46, Aslak Knutsen wrote:
Anything can be wrapped in a Maven plugin.
It would be preferable if it was not a Maven plugin, then it just
might be
usable outside of that target audience.
Sure, I meant whether we are assuming a maven plugin as a
deliverable?
We must have a maven plugin offering. Ideally gradle and npm as
well.
In my hopes and dreams, it's a generic Java library which is then
consumed
and leveraged by a Maven Plugin.
Like: you wouldn't put all your business logic in a servlet, so why
put
any in a Maven Plugin? :)
Roland made note in another thread that a fair bit of the current
capabilities could likely get removed.
Yes, but we need an inventory of these from those who are depending
upon
it. For instance the Fuse product.
S,
ALR
-aslak-
On Wed, Jun 13, 2018 at 11:22 PM Gorkem Ercan <ger...@redhat.com>
wrote:
On 11 Jun 2018, at 10:04, Andrew Lee Rubinger wrote:
On Sat, Jun 9, 2018 at 4:48 AM, Burr Sutter <bsut...@redhat.com>
wrote:
Based on the "other thread"...
It does seem to me that fabric8-maven-plugin is mission
critical to
our
developer experience.
Or at least some incarnation of a Java language-specific adaptor
to
oc, yes.
What do you mean by adaptor? Are we still assuming a maven plugin
or
something else?
S,
ALR
So, for the parties who have been "naysayers", can we drop the
"nays"
and
focus just on "making it more awesome"?
On Fri, May 25, 2018 at 3:06 PM, Max Andersen
<mande...@redhat.com>
wrote:
Roland - if you are seeing OpenShift.io based fixes that
don’t
make
sense please speak up. I didn’t see any feedback on the
roadmap
request for
inputs.
The intent is for sure not to diverge away. We want to work
together
on
this rather than see even more new forks.
And yes gofabric8 etc. is also not of use for us.
The whole images situation we should work on aligning as that
also
relate
to middleware etc.
/max
http://about.me/maxandersen
On 16 May 2018, at 18:44, Roland Huss <rh...@redhat.com>
wrote:
Hi Burr,
let me try to comment from a Fuse perspective:
* We are still committed to the fabric8 base images (s2i [1]
and
non-s2i
[2]) as they are the only community images available. We will
continue to
use them in the future FIS releases (at least upstream).
* The fabric8-maven-plugin's [3] scope has diverged from our
use
cases,
so we plan to explore new ways of providing a way to leverage
OpenShift
builds and deployments for the FIS quickstarts. I.e. fmp's
scope
gets
aligned currently to its usage in openshift.io, drifting away
from
being
a general purpose plugin. Also, many features (like gofabric8,
helm,
cluster installation, .. support) are of no use for us, so we
don't
want to
maintain them for no obvious reason.
We have some initial ideas, but nothing concrete. So happy to
start
a
discussion with Devtools, RHOAR and you how to align plans.
* Fabric8's docker-maven-plugin [4] will be still supported in
the
upstream (it has been productised only as an integral part of
f-m-p)
These our current plan but we too are interested in what the
dev
tools
team is planning on doing, of course.
regards ...
... roland
[1]: https://github.com/fabric8io-images/s2i
[2]: https://github.com/fabric8io-images/java
[3]: https://github.com/fabric8io/fabric8-maven-plugin
[4]: https://github.com/fabric8io/docker-maven-plugin
On Mon, May 14, 2018 at 9:45 PM Burr Sutter
<bsut...@redhat.com>
wrote:
Did we ever come to agreement on keeping F8 alive?
Specifically the
base docker images
maven plugin
bit.ly/msa-instructions still relies on these tools and that
is
still
one of our most successful assets (seen in
bit.ly/microservicesvideo)
_______________________________________________
Devtools mailing list
Devtools@redhat.com
https://www.redhat.com/mailman/listinfo/devtools
_______________________________________________
Devtools mailing list
Devtools@redhat.com
https://www.redhat.com/mailman/listinfo/devtools
_______________________________________________
Devtools mailing list
Devtools@redhat.com
https://www.redhat.com/mailman/listinfo/devtools
--
Twitter: @ALRubinger
_______________________________________________
Devtools mailing list
Devtools@redhat.com
https://www.redhat.com/mailman/listinfo/devtools
_______________________________________________
Devtools mailing list
Devtools@redhat.com
https://www.redhat.com/mailman/listinfo/devtools
_______________________________________________
Devtools mailing list
Devtools@redhat.com
https://www.redhat.com/mailman/listinfo/devtools
--
Twitter: @ALRubinger
_______________________________________________
Devtools mailing list
Devtools@redhat.com
https://www.redhat.com/mailman/listinfo/devtools
_______________________________________________
Devtools mailing list
Devtools@redhat.com
https://www.redhat.com/mailman/listinfo/devtools
_______________________________________________
Devtools mailing list
Devtools@redhat.com
https://www.redhat.com/mailman/listinfo/devtools
