I don't know if CTR is the norm at the ASF but I know there are some
pretty big adherents to it.

I'm normally an advocate of RTC and I think it's actually more
important when a community is small. I'm low on spare time right now
so let me see if I can find one of my old write ups to see if I still
believe what it says. The short version is that below a certain level
of activity we'll end up with essentially no community interaction.
Encouraging that interaction is part of why I've supported us taking
reviews from anyone as binding for RTC (rather than just committers).

FWIW I'd say right now I'd be -0 on going to CTR. I know you've got
some big changes you've been working through and my once-a-month or
whatever volunteer time to look at Yetus stuff isn't going to help
enough for that. If we decide to go the lazy consensus route, 24 hours
seems like too short of a window. E.g. we often see 72 hour windows on
ASF projects to try to account for timezone and work schedules.
On Thu, Nov 29, 2018 at 11:35 PM Allen Wittenauer
<[email protected]> wrote:
>
>
> I propose moving the Apache Yetus project to a Commit-then-Review model.  
> It's been evident for a very long time now that patch reviews are lacking 
> sufficient resources to move the project forward.  As a result, patches may 
> sit there for a very long time.
>
> I believe the only way to improve the situation is to make Apache Yetus 
> easier to use for those outside the ASF bubble to draw more interest in the 
> project.  The only way to do that is to change some of the core assumptions 
> of the code, write documentation to the website, and so on.  To do those 
> things in a way the whole project benefits is to commit them to the source 
> tree.  However, accomplishing those things cannot be done if patches are left 
> to rot.  Thus, we are in a chicken-and-egg scenario.
>
> Going to CTR would mean that the usual quality checks would happen either 
> post-commit, during release, or perhaps we have a window by which patches may 
> be committed lacking any input (say, 24 hours).
>
> Incidentally, reading through 
> https://www.apache.org/dev/new-committers-guide.html#guide-for-new-committers 
> it would appear that CTR is the norm in the ASF.
>
> Thoughts?

Reply via email to