Date: prev next · Thread: first prev next last
2011 Archives by date, by thread · List index


Hi Bjoern,

On Mon, 2011-09-26 at 19:47 +0200, Bjoern Michaelsen wrote:
This seems to me to be a bit overzealous. Invoking the impression that
development on branches is oldschool and obsolete is just wrong --

        There is no problem with working on branches, indeed - I'm actively
working on one (on and off at least). The problem / groupthink issue
from OO.o-land came more from trying to coerce people through ever more
restrictive hoops to get their code included; that is what I'm against.

Making no commits on master and every commit on a branch (like OOo did
unless you had the godlike RelEng rights) was way wrong of course.

        Quite; but it was way worse than that in the past - the philosophy of
quality-through-inflicting-pain-and-slowness-on-developers was acute,
and I want no trace of it here. And yes, it's possible to get too
fearful of process ;-) but I think that's a very good default position
to make people work v. hard to introduce more burdens.

Apply common sense. IMHO it could be quite healthy to have some more
work done on (publicly visible) branches(*).

        Sadly common sense is often not so common :-) there is always a good
reason for adding every incremental barrier to entry, and removing them
is hard.

That being said: A new warning is not a shooting offense IMHO (for
reasons you stated quite eloquently). Breaking master on you own
working platform however should induce a healthy amount of shame

        ;-) well, I don't know how ashamed we should be, if it is caught
quickly and fixed that's fine with me, but completely agreed that
slowing other people down significantly is really not-good (TM). The
more tools and analytics we can deploy to try to stop that the better.

Those 10-20 commits which are pushed to master as one. I dont think it
would hurt anyone, if this branch is visible as it grows commits instead
of being hidden on a local disc -- actually I think it would help.

        Ho hum; if the commits are for areas that are likely to be shared and
conflicting ( like fixing warnings ) I guess it might. Usually our
changes are far enough apart that it is not an issue - surely ? and it
seems unclear what the benefit vs. the pain pwrt. when re-synching
(around deleting old remote branches etc.) would be.

        All the best,

                Michael.

-- 
michael.meeks@suse.com  <><, Pseudo Engineer, itinerant idiot


Context


Privacy Policy | Impressum (Legal Info) | Copyright information: Unless otherwise specified, all text and images on this website are licensed under the Creative Commons Attribution-Share Alike 3.0 License. This does not include the source code of LibreOffice, which is licensed under the Mozilla Public License (MPLv2). "LibreOffice" and "The Document Foundation" are registered trademarks of their corresponding registered owners or are in actual use as trademarks in one or more countries. Their respective logos and icons are also subject to international copyright laws. Use thereof is explained in our trademark policy.