On 9/21/2016 6:56 AM, Amrith Kumar wrote:
Of late I've been seeing a lot of rather questionable changes that
appear to be getting blasted out across multiple projects; changes that
cause considerable code churn, and don't (IMHO) materially improve the
quality of OpenStack.
I’d love to provide a list of the changes that triggered this email but
I know that this will result in a rat hole where we end up discussing
the merits of the individual items on the list and lose sight of the
bigger picture. That won’t help address the question I have below in any
way, so I’m at a disadvantage of having to describe my issue in abstract
terms.
Here’s how I characterize these changes (changes that meet one or more
of these criteria):
- Contains little of no information in the commit message (often just
a single line)
- Makes some generic statement like “Do X not Y”, “Don’t use Z”,
“Make ABC better” with no further supporting information
- Fail (literally) every single CI job, clearly never tested by the
developer
- Gets blasted across many projects, literally tens with often the
same kind of questionable (often wrong) change
- Makes a stylistic python improvement that is not enforced by any
check (causes a cottage industry of changes making the same correction
every couple of months)
- Reverses some previous python stylistic improvement with no clear
reason (another cottage industry)
I’ve tried to explain it to myself as enthusiasm, and a desire to
contribute aggressively; I’ve lapsed into cynicism at times and tried to
explain it as gaming the numbers system, but all that is merely
rationalization and doesn’t help.
Over time, the result generally is that these developers’ changes get
ignored. And that’s not a good thing for the community as a whole. We
want to be a welcoming community and one which values all contributions
so I’m looking for some suggestions and guidance on how one can work
with contributors to try and improve the quality of these changes, and
help the contributor feel that their changes are valued by the project?
Other more experienced PTL’s, ex-PTL’s, long time open-source-community
folks, I’m seriously looking for suggestions and ideas.
Any and all input is welcome, do other projects see this, how do you
handle it, is this normal, …
Thanks!
-amrith
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: [email protected]?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Nova has a 'how to get involved' wiki, which might be linked from the
developer docs somewhere:
https://wiki.openstack.org/wiki/Nova/Mentoring#How_to_Get_Involved
Basically has some high level how to get involved info for nova and some
of the low hanging fruit efforts going on, like python 3 support,
mox->mock conversion, unit test cleanup, config option cleanup, api-ref
docs cleanup, etc.
Obviously this doesn't work if no one sees it or reads it, or
understands it.
I'm not sure how to prevent someone from blasting the same low-value
code churn change across a dozen projects before asking any of those
teams if it's a good idea to begin with. I think that's what Tom Cruise
was supposed to be for in Minority Report...
--
Thanks,
Matt Riedemann
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: [email protected]?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
