[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [tracecompass-dev] Important for contributors: Directory layout change in master
|
On 2015-06-18 02:19 PM, Matthew Khouzam wrote:
On 15-06-18 11:50 AM, Alexandre Montplaisir wrote:
On 2015-06-18 10:26 AM, Matthew Khouzam wrote:
Just to clarify, if some people are wondering why the patches went in
without a proper review, the answer is, they were reviewed. They were
pushed atomically as a branch outside of gerrit. This is not the
standard way of working though, we pretty much always review on gerrit
as we are striving for as much transparency as possible in the project.
The Git repo is just as public and visible as Gerrit. I fail to see
how it lacks transparency, as you seem to imply.
I am not implying, but rather stating that gerrrit IMO is more
transparent as it keeps a record of the review process, whereas git
records the content, in patch form. When a patch is submitted, it does
not simply go into the branch, but we have at least two committers OK
the patch. The process of review is public and transparent, we don't go
into a Dr Strangelove like war room (also we can't afford one) to review
code and merge it without a record of what happened.
Every -1 with subsequent iteration is important as it shows not only
what gets in but also what does not and why.
Cheers,
Matthew
That particular branch has been:
1) Discussed on the mailing list:
http://dev.eclipse.org/mhonarc/lists/tracecompass-dev/msg00126.html
2) Pushed to Gerrit, reviewed, and +2'ed on there:
https://git.eclipse.org/r/#/c/50222/
(Marc-André just +2'ed the tip of the branch, for simplicity's sake).
There were comments and according changes all along the way. I don't
know how you can be more transparent than that. Why does it bother you
so much that I used a different channel for the final submit (atomic
"git push" vs. submit,submit,submit on Gerrit) ?
