[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: From release notes for FC5T3 (web)

On Fri March 10 2006 4:54 am, Rahul Sundaram wrote:
> gb spam wrote:
> >>On another list I asked two simple questions of the anti faction:
> >>1) Would it be a big deal for the coders to reinstate the "Everything
> >> install" button?
> >
> >Rahul said a number of times that the cost was high.  When asked to
> >quantify what high meant, he neglected to answer.  My suspicion is
> >that it would be high on pride.
> Nothing to do with pride since I did give you the answer.  Its a choice
> between spending time on the list discussions or go back and do actual
> work and people have pointed out to me ( by swearing at me offlist) that
> the balance currently is tilted towards discussions. 

Let it be said that I, who wrote the original words, above, that gb spam 
replied to, never swore to you off list. It is also the case that if you 
responded to my point 1 above during that debate, I missed your response. I 
gave up on the debate at a certain point, precisely because I didn't see any 
responses to my two questions. My FC5 box shortly thereafter, blew up after a 
yum update, and I stopped reading the list entirely. 

> I have already 
> explained several times on the list that Anaconda code has been
> rewritten to use yum and this feature has to be *reimplemented* now and
> the Anaconda developers told me that the code to reimplement this would
> be significant and they dont see enough merits to do so.  Combine that
> with the effect of this "feature" and the time spend in fixing and  hand
> holding new users who choose to use this option innocently and we land
> up with a high cost I cited.

OK - maybe this is the valid explanation I was seeking - I say maybe, because 
I'm not enough of a coder to say authoritatively. 

> >>2) Was there still an objection if the button were hidden, and
> >> accompanied by large disclaimers?
> >>
> >>Not one of the vociferous anti-everything faction responded - this was
> >>reposted three times during a very long debate on the issue. My
> >> conclusion is that the decision had been taken, and the fact that
> >> numerous users disagreed with the decision was of no account. The debate
> >> was never truly a debate - the issue was decided without end-user input.
> >> So, I repeat, we're dealing with control-freaks - it's a bad sign for
> >> Fedora, in my view.
> Are the developers control freaks if they dont see implement every
> feature you wanted?. Contrary to perception, I also discussed option 2).
> The problem with hidden features (such as the previous expert option in
> Anaconda)  is that they provide a alternative code path that is rarely
> tested and turns out broken in subtle ways every now and then.  This is
> a merit based system in which people who do the actual work win in the
> end.

My characterization was based on long participation in that debate, in which 
there was repeated ignoring of my question. You say that a response was 
eventually given - OK. As I explained above, I gave up on the discussion 
after several days. Regarding your last sentence, it's obvious on one level 
that the developers have the final say - BUT, if you decide to shut out the 
voices of your end-users, you will not achieve your goals. That was my 
perception at a certain point, that the many developers on that list were 
freely tossing in their defenses of their decision, but they were not 
responding to my two simple questions...

Currently, I'm in the evaluation mode re: Linux distros. I currently run six 
on various boxes. As I move closer to implementing actual production 
machines, I watch these debates/discussions on the various lists like a hawk. 
Eventually, I will be in the same position as Aaron, and will have to settle 
on a distro which will be implemented on many machines - probably, Fedora 
will not wind up on the most critical machines, but, RedHat derivatives will, 
and in some cases RedHat itself. But, maybe not. It is with such concerns, 
that I look at these sorts of debates. Part of that meritocracy you mention 
involves coding, but part also involves a social element, how you respond to 
your users needs and desires is important...

Claude Jones
Bluemont, VA, USA

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]