Taking fedora-devel into consideration [Was: Re: tab completion less useful now, due to sbin in path]

[Thorsten Leemhuis]

On 08.10.2008 21:31, Kevin Fenzi wrote:
> On Wed, 08 Oct 2008 12:48:54 +0200
> mcepl at redhat.com (Matej Cepl) wrote:
>> On 2008-10-07, 06:37 GMT, Thorsten Leemhuis wrote:
> [...]
> I don't speak for FESCo, but I do care what gets discussed here. 
> I however don't think it should just be a majority rule or 'he who
> shouts loudest' or even 'who we think "won" the discussion. 

Agreed, but nevertheless the process for getting things/features 
discussed/approved by FESCo partly tends to follow the 'follow the one 
who shouts loudest' mechanism(¹), because a whole lot of faith and trust 
is put into the one that drives the feature forward -- e.g. he's the one 
who shouts loudest in this case. But he's not doing it on purpose -- it 
just happens.


A totally made up example might help to express better what I'd like to say:

Say someone suggests to switch back to XFree86 for some reason (that's a 
bit to crazy for this example, but my mind didn't came up with something 
better). He writes a feature page and posts a proposal to fedora-devel 
(not all feature owners do that, which IMHO is wrong; posting the 
proposals to the list directly before they are being discussed would 
help a lot IMHO); some people will be in favor of the proposal (some 
people will be in favor even if things are crazy), but a lot of people 
will disagree (which definitely would be the right thing to do in the 
XFree86 example ;-) ). The discussion sooner or later will slow down; 
most subscribers will remember the debate as "most people disagree with 
it so this is not going to get realized in Fedora".

But now let's imagine that the one that want to do that change doesn't 
realize that it's wrong to do what he wants. He improves and modifies 
his proposal and one or two months later presents it to FESCo with the 
words "this was discussed on fedora-devel a few weeks ago; some people 
did not like it, hence I modified the proposal after I got all that 
feedback. Here is the improved version; please discuss and approve".

Most people (including FESCo members) after those one or two months will 
have forgotten most (if not all) of the arguments from the mailing list 
discussion; and only some people/FESCo members will look at the old 
discussion closer again (we are all busy and time is very rare); so it's 
easy to at least partly tend to trust the proposal (and its writer) that 
is up for FESCo discussion -- it's the classical "nuclear reactor" part 
in the bicycle shed example ( 
http://en.wikipedia.org/wiki/Color_of_the_bikeshed ).


The only way to prevent this problem: if you dislike something you need 
to watch "something" constantly and work against it as long as there is 
someone else that drives "something" forward. That's a whole lot of work 
  :-/

 > [...]

CU
knurd

(¹) note that his was quite similar in the old days when I was in 
FESCo/FESCO chair, so don't read this mail as a "knurd is complaining 
about FESCo again" mail ;-)