RFE: Retire Fedora Core 4 only _after_ FC6 has been released.

[Mike A. Harris]

Peter Arremann wrote:

> Please don't take this as a flame - but why should users care about that part? 
> You have a very valid point considering amount of work that RH engineers need 
> to do, but to the outside workd, it sounds a lot like the "dog ate my 
> homework" kinda excuse... Why should a Fedora user care about how much time 
> RH spends on RHEL? After all, they are not using RHEL but Fedora - which RH 
> so often states is a community project... 

And Fedora Legacy is what enables the community to continue maintenance
of the OS releases, for as long as the community desires to have them
maintained, by enabling external contributors to do the work.

> What I can't judge is the amount of work that this creates. FC3 has received 
> 11 new srpms since beginning of the year. If you extend the "official" 
> support for FC3 by 2 or 3 more months until shortly after FC5 is out, how 
> much more work would that really be? 

The same amount of work it takes someone in the community to contribute
the same updates to the Fedora Legacy project roughly.  The community
is in theory infinite in size though, and thus is much more scaleable
to maintaining things indefinitely.  In contrast, while the Red Hat
developer base is steadily increasing, the workload also is steadily
increasing.  Finite time and close to static developer resources, mean
that there is a maximum amount of work that can be done in a given time
with the manhours available.  The more OS releases that are maintained
at any given point in time, the more those manhours are spent updating
old legacy OS releases, and the less time is spent working on
development.

How RHEL relates to all of this is simple.  Whatever RHEL work needs to
be done by Red Hat engineering, simply is that - it is work that needs
to be done, and will be done.  That work will get priority, as that is
part of the reason our customers pay us money.  Since the time it takes
to maintain RHEL comes out of the same human resources that are spent
working on Fedora, you take the total time available, subtract the time
that needs to be spent on RHEL development, support and maintenance, and
what is left over, is Fedora development time plus Fedora maintenance
time.

While Fedora users such as yourself need not know or care about all of
this, the bottom line, is manpower resources are limited, and RHEL and
Fedora work is shared by the same people.

By transfering Fedora releases to Fedora Legacy and thus into the hands
of the very capable community, maintenance is scaleable to whatever
amount of manhours the community volunteers to put into it.

Likewise, if nobody were to volunteer, and thus no updates were to
happen, would IMHO be a sign of insignificant interest in the given
OS release being maintained.


> I don't think that's the question. The question is more like "Do I want to be 
> considered a legacy?"... Absolutely non-technical, but honestly - the word 
> legacy has such a negative meaning in english for many people. In the 
> perception of most users legacy systems only receive the most essential 
> security updates and if you ask for support the techs will laugh and make 
> comments "anyone still using that?" 

I've never perceived it that way.  If this is a legitimate gripe
however, perhaps it is time for the Fedora Project to change that
misconception.  Perhaps Fedora Legacy should have it's name changed
to something that carries an unarguable positive connotation.

"Fedora Forever!"

> I agree with the original poswer that - no matter if technically there is a 
> good reason or not - people should be given a good choice of when to upgrade 
> and to what release. Most people simply don't see staying with a legacy 
> system as a choice. 

This argument can be carried out indefinitely though.  The original plan
was to support/maintain one single OS release at a time.  That was
extended by 3 months to give people time to upgrade to the new OS
release without having to be forced to do so immediately.  Now what
the original poster of this thread is saying is that this should be
extended to 2 OS releases.  If that were to happen, then an equal
number of people would complain that Fedora Core 3 was being transferred
to legacy "immediately" when Fedora Core 5 came out, and they'd beg
us to extend support/maintenance by a few months, perhaps to Fedora
Core 6 test2.  At that point someone would bring up the same argument
again that it should happen when Fedora Core 6 comes out instead of
at test2, and before you know it we now have 3 OS releases consuming
our resources.

Regardless of what anyone's own personal thoughts/feelings are about
how long support should live, someone else will think that it should
be supported longer, and someone else will think it should be supported
less and the resources devoted to new development.  We already thought
about that, and drew the line in the sand at one OS release, then
moved it out by a few months to accomodate some concerns that were
raised.

But hey, the OS is free of cost, you get what you pay for ultimately,
and nobody is paying for support for 2 or 3 OS release (or even one),
so that's a pretty good bargain to get a free OS with approximately
9 months of OS maintenance if you ask me.  Even better that you can
then upgrade, again free of cost to the new OS, or continue to use
the existing OS for free, and have it maintained by the community.
Even better - you can become part of that community and contribute
maintenance volunteerism to keep it alive indefinitely.   All without
spending a dollar.  Pretty damn good deal.

RHEL is of course available for people who want to have a commercially
supported OS for long periods of time.  There are other options
available out there also, such as Debian, etc.

Just my $0.02 CAD...   YMMV



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