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

Re: Could tthere be an update iso distribution



On Fri, 2006-03-17 at 03:08 -0500, Mike A. Harris wrote:
> Leslie Satenstein wrote:
> > With kudzu broken, potential keyboard problems, as well as the 2054 
> > kernel for nvidia, I see no reason to download the first release.
> > Why push broken software unless a new iso in a very short time 
> > thereafter(2 weeks) is released.
> 
> That's easy enough to answer.  All software has bugs in it, or at
> a bare minimum - most software does.  The software included in
> Fedora Core is no exception, and never has been.  Every single
> OS release ever made by Red Hat, or for that matter by Debian,
> SuSE, Mandrake, Sun Microsystems, Microsoft, Apple, or any other
> OS or distribution you could care to name, has shipped software
> which contains bugs.
> 
> Some of the bugs are minor, while other bugs are larger in impact.
> Bugs are a simple fact of life in computer software, and there's
> really no alternative than to live with that fact.
> 
> It is entirely possible in theory, to delay any of the above
> operating system's release dates, to then "go and fix that one
> more critical bug", but after doing so, there is at least 1, and
> more likely 100 or more other "critical" bugs that aren't fixed
> also.  Why not fix them too?
> 
> The reason why, is that you then delay your OS release by another
> 2 weeks, then another, then a month, and eventually Debian users
> start chastizing you for trying to clone the Debian release model.
> 
> That was a joke.  Ok, only partly a joke.
> 
> Seriously though, the release just gets delayed and delayed, and
> one user after another jumps on the bandwagon claiming "You fixed
> critical issue #12345, but *MY* issue which is obviously much
> worse of a problem is not fixed!  You must delay and fix my issue
> also!".   After 1000 users do this, you end up not releasing an
> operating system ever period, because there is _always_ another
> critical bug left in the OS.
> 
> There will _always_ be bugs in the OS which are frustrating to
> someone out there.  Always be bugs to which someone can't believe
> we could possibly have been so careless to release the OS without
> fixing at _least_ that one issue.  But again, there isn't one
> person out there thinking that.  There are 10000 people out there
> thinking that, and each of them is thinking about a different issue.
> 
> Delaying the OS to fix every bug that "someone" out there will
> be frustrated about, or shocked senselessly that we could possibly
> have released without delaying to fix their issue, really _DOES_
> mean "Never ever ship an OS ever."
> 
> That is just a straight fact - wether or not everyone accepts it
> or not.  The OS _WILL_ always ship with bugs, some known ahead
> of time, and others discovered after the fact.  At a certain
> point you either ship the OS, bugs and all, or throw in the
> towel.
> 
> 
> > Suggest a core5.01 version that has the bugfixes for nvidea, keyboard 
> > and for kudzu.
> 
> By all means, please feel free to create your own custom ISO images
> with whatever fixes you'd like to have in them after they're released.
> The Fedora project encourages users to customize the distro, and that
> includes making custom bug-fixed ISO sets if desired.
> 
> 
> > Sorry to be against the release. But a reputation for quality is at 
> > stake. I do not want to tell others, "The Core 5 cd's that are in the 
> > books and magazines are faulty" Do not use them, do not purchase the 
> > book or magazine.
> > 
> > Better to be 4 days late, have the fixes, then to ship a faulty product.
> 
> That's where you're wrong.  This release is no different than FC4, FC3,
> FC2, FC1, any RHL release, or any RHEL release.  All OS releases have
> bugs in them as I've stated above.  Some more severe than others, but
> that is just a fact of life.  If FC5 contains an out of the box bug
> which completely prevents you from using the OS, that is unfortunate,
> but it will likely work for many other people.  Fixing your favourite
> 3 critical must-have-fixed bugs, may fix the OS for you, and a handful
> of other people also, but then there will be 1000 other critical bugs
> that are not fixed which cause the same grief for other people as you
> are potentially seeing.  Should we stop-ship the OS and _just_ fix
> the bugs you care about?  Or should we stop-ship the OS and fix _ALL_
> critical bugs that _anyone_ cares very strongly about?
> 
> Stopping shipping for one person, or even a few people is not a viable
> option, and isn't fair for everyone else out there who has equal or
> worse problems.  Stopping shipping for everyone with a serious problem,
> until all such problems are fixed, means delaying the OS for weeks,
> or months - most likely years, or never actually ever releasing the
> OS again ever.  Plus, while those "critical bugs" are being fixed, then
> all of a sudden the new upstream release of GNOME comes out, and someone
> else wants that included too, since the release is delayed anyway!
> Then there's a new upstream release of openoffice which fixes a lot
> of bugs, might as well include that too.  Then X11R7.1 comes out, then
> a new upstream kernel, then KDE, then gphoto, then ....  you get the
> idea.
> 
> All new OS releases are going to have instabilities.  Get used to that,
> because it is reality, and reality isn't going to vanish any time soon,
> wether the OS release date is delayed to fix an issue or two or not.

I am not sure if this was already said?::
===========================================================================
The Release manager plays a major role in deciding when the OS is ready
to ship.
We should encourage the users and developers to find/report/fix bugs.
When the product is ready it is ready.

In a competitive world the winner has optimized the following:
--- Time to market
--- Quality
--- Stability
--- Performance
--- Marketing Genius (Eye candy appeal, etc...)
--- Customer support
===========================================================================

So, let the users complain.  
The FC5 team will deliver a good product, I hope :)



Thanks,
Ernesto






> 
> 
> 
> 
> -- 
> Mike A. Harris  *  Open Source Advocate  *  http://mharris.ca
>                        Proud Canadian.
> 


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