couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Noah Slater <nsla...@apache.org>
Subject Re: Experimenting as a way to decide WAS: [DISCUSS] Move more email to the commits list
Date Wed, 05 Feb 2014 11:55:19 GMT
An experiment should be well defined, should have an owner
(preferably), should have success/failure criteria, and should have an
end date.

As long as we have these things, I don't see any harm in experimenting
to see what works.

Communities (and other complex systems) are not like code. You cannot
write unit tests. You can't even debug them that well. Experimentation
is an important tool. Otherwise you'd never be able to make changes,
because nothing is certain or known beforehand.

On 5 February 2014 11:28, Benoit Chesneau <bchesneau@gmail.com> wrote:
> On Wed, Feb 5, 2014 at 11:16 AM, Alexander Shorin <kxepal@gmail.com> wrote:
>
>> On Wed, Feb 5, 2014 at 2:10 PM, Benoit Chesneau <bchesneau@gmail.com>
>> wrote:
>> > On Wed, Feb 5, 2014 at 11:03 AM, Alexander Shorin <kxepal@gmail.com>
>> wrote:
>> >
>> >> On Wed, Feb 5, 2014 at 1:53 PM, Noah Slater <nslater@apache.org> wrote:
>> >> > We wont know until we try. Perhaps that's why an experiment would be
a
>> >> > good idea?
>> >>
>> >> Experiments are fun (: and the only way to find the sweet spot or
>> >> ensure that it had been already found.
>> >>
>> >
>> > that why we advise to the babies to put there finger in the power-plug.
>> > It's funny ;)
>>
>> This is incorrect analogy since it produces different behaviour and
>> have very different side effects. You can always rollback ML changes,
>> but I'm not sure that you can do the same for the bad things in your
>> life. Anyway, offopic (:
>>
>>
> No not really,
>
> It's common to think that we can easily rollback on internet. In practice
> that is not that easy. You have to decide about deadlines, make sure people
> understood why. And at first convince people to use them. What do you do
> for people not convinced at first, are you handling both and old way? Are
> you forcing other to try? How do you convince people not convinced to try?
> What if at the end they didn't try? ....
>
> Just like some people says it's better to not put a protection on the plug,
> say to the baby it's bad and let see... Experimenting should be carefully
> discussed first imo and used in the last resort.
>
> - benoit



-- 
Noah Slater
https://twitter.com/nslater

Mime
View raw message