couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jan Lehnardt <>
Subject Re: [jira] [Commented] (COUCHDB-1342) Asynchronous file writes
Date Fri, 18 Nov 2011 01:08:32 GMT
Moving this out of JIRA.

> Thirdly, every time someone asks, "Can it wait till it's on trunk?", all I hear is, "Can
I ignore what you just said and commit this anyway?" If I point at something and say that
its broken its because I'm expecting the patch to change or an explanation of why I'm wrong.
And I'm fine being wrong. It happens quite often.

Paul, thanks for explaining how this ends up on your end (communication is a two way horse,
yay). I can only speak for myself, but I never ever intend to ignore the feedback. Hence my
latest reply in this regard to open blocking issues with the comments for the next release
off the particular branch a patch goes into. I see where your frustration is coming from now,
it wasn't clear before.

> But this pattern of submitting patches and asking for all concerns to be addressed after
the patch is in trunk is starting to get a bit annoying.

You are saying "all concerns" and I take it as a figure of speech, but the reverse, addressing
all concerns (ever so vague and little) is equally frustrating.

The solution is clear to me, we need to try and strike a balance. I don't think anyone argues
that having a patch mature in JRIA for a few days (or weeks) is bad and equally, once it hits
a sweet spot, that it can continue to simmer in a release branch or master and have further
kinks worked out as they show up in broader use of that branch.

Knowing when to make the cut is the tricky part of course so that we don't have patches rotting
on the one side and master erode into a mess on the other.

I don't think any sort of policy or process helps here, but we should collectively be aware
of the trade-offs we are making and have a natural urgency to get good patches into master
and stemming from that an urgency to help mature said patches in JIRA before they do hit master
(all modulo this is a volunteer effort, of course :).

I feel I am starting to state the obvious here, so I'll leave you with this.


View raw message