Mike,
The short answer is “this is not as simple as replacing one class”. The longer
answer is that you may be interested in the work by Mark Payne and Sam Hjemfelt
to provide an additional runtime in NiFi (referred to as stateless NiFi [1]).
Hope this helps.
[1]
Adam,
I think your statements about “definitely a total sum of work that is greater”
are true for a specific audience. As someone who routinely reviews PRs and
handles release management tasks, I know where I would like to see
improvements. I also write feature code (though not as often
Adam,
To your point, we currently have the situation you described when
dependencies exist between NiFi and NiFi Registry, where one must be
released before the other.
In my opinion, it is better to be able to verify functionality in smaller
atomic change-sets and limited scope over a larger
Hello
I want to use Nifi's UI and APIs (i.e. registry) but not use the Nifi
scheduler, controller, runtime. I want to prototype hooking in my own
scheduler, controller, runtime. How involved would this be? Would it be more
involved than replacing the class
Andy - fair points. Note that by definition, the process you describe is
harder (requires more maneuvers). Maybe it's warranted/justified for the
desired integrity that you are after, but it's most definitely a total sum
of work that is greater.
Your registry example is really good. In your
I think by definition, a contribution _must_ fit into a single repository. This
will force developers to carefully consider the boundaries between modules and
build clean abstractions. If you are a new contributor, I would be surprised if
you are making a single (logical) contribution that
Bryan,
I think both of your points are actually closely related, and they somewhat
speak to my thoughts/concerns about splitting the repository.
I would argue that one PR that affects multiple modules in a single
repository is _easier_ to review than multiple PRs that affect single
modules. In
Two other points to throw out there...
1) I think something to consider is how the management of pull
requests would be impacted, since that is the main form of
contribution.
Separate repos forces pull requests to stay scoped to a given module,
making for more straight forward reviews. It also
Thanks Adam and Edward! This is exactly the type of discussion I was
hoping a detailed and specific proposal would generate, so thanks for
the input. I'll reply to each of you in turn:
Adam,
It is true that a repo-per-project approach is not required. I've
worked on projects that do it both ways
To be honest and to your point Joe, the thing that optimizes the RM duties
should probably be preferred in all of this. There is so much overhead for
the release manager, that lubricating the RM process probably trumps a lot
of my concerns. I think there's real concern for making the project
I think Nifi would really benefit from this. One thing I think that, should
be looked into is something I noticed while trying to get to grips with the
Nifi source, At the start of the year I did a small exercise to track how a
call from the API to start and stop a processor translates to a
I think the concerns around user management are valid, are they not?
Overhead in JIRA goes up (assigning rights to users in JIRA is
multiplied). Risk to new contributors is high, because each isolated
repository has its own life and code contribution styles. Maybe the actual
apache infra
to clarify user management for infra is not a prob. it is an ldap group.
repo creation is self service as well amd group access is tied to that.
release artifact is the source we produce. this is typically correlated to
a tag of the repo. if we have all source in one repo it isnt clear to me
Just as a point of discussion, I'm not entirely sure that splitting into
multiple physical git repositories is actually adding any value. I think
it's worth consideration that all the (good) changes being proposed are
done under a single mono-repository model.
If we split into multiple
14 matches
Mail list logo