Re: Fedora more successful, developer-wise, than Ubuntu

2007-12-24 Thread Jeff Spaleta
On Dec 24, 2007 2:51 PM, William Cattey <[EMAIL PROTECTED]> wrote:
> apt
> package management system makes a lot of the work to switch from our
> present tightly integrated OS + alternate versions of packages +
> additional packages to alternate versions and additional packages
> layered on a pre-installed system.

And you can't accomplish this with yum in a fedora ecosystem of repositories?


> It is product focus, however that makes Ubuntu attractive to me and
> many at MIT.  With Fedora and Enterprise, the two choices are three
> year old functionality, or bleeding edge functionality.  Example: in
> November Fedora 8 came out with important laptop power management
> functionality, but it also shipped with a broken rewrite of
> NetworkManager, and an alpha version of BIND.  Fedora 6 shifted to
> not getting any attention, and the clock to get off Fedora 7 started
> EVEN BEFORE FEDORA 8 FUNCTIONALITY WAS STABLE!

The NM in F8 is not "broken".. in any event there is the older network
stack that does not rely on NM for usage cases that NM isn't ready
for.  NM works for more network situations for my laptop in F8 did it
did in F7.

> At particular points a particular version is flagged for longer term
> support, but the primary focus is not on exploring new functionality,
> or on back porting popular functionality to an ancient codebase
> tailored to customers afraid of change. It is simply:  Produce a
> usable experience with a balance of stable code and new
> functionality.  What Fedora-based spin has this focus?

Fedora Legacy was tried ealier and had a 2 year live cycle .. and
while there was interest in using such a distribution from a usage
standpoint, there was not enough people interested in contributing to
actually doing to back ports.  When legacy was decommissioned the
Fedora lifecycle was adjust to give approximately 13 months of
updates, along individuals to 'skip' a Fedora release.  There's
absolutely nothing stopping another community attempt at 2 year
target.


-jef

-- 
Fedora-marketing-list mailing list
Fedora-marketing-list@redhat.com
https://www.redhat.com/mailman/listinfo/fedora-marketing-list


Re: Fedora more successful, developer-wise, than Ubuntu

2007-12-24 Thread William Cattey
It turns out that 1000 seats at MIT are switching to Ubuntu from  
Enterprise.
They will not be using LTS, because they expect to need more recent  
hardware drivers.


They are switching to Ubuntu instead of Fedora even though they  
expect there will be an annual OS update of Ubuntu required, the apt  
package management system makes a lot of the work to switch from our  
present tightly integrated OS + alternate versions of packages +  
additional packages to alternate versions and additional packages  
layered on a pre-installed system.


It is product focus, however that makes Ubuntu attractive to me and  
many at MIT.  With Fedora and Enterprise, the two choices are three   
year old functionality, or bleeding edge functionality.  Example: in  
November Fedora 8 came out with important laptop power management  
functionality, but it also shipped with a broken rewrite of  
NetworkManager, and an alpha version of BIND.  Fedora 6 shifted to  
not getting any attention, and the clock to get off Fedora 7 started  
EVEN BEFORE FEDORA 8 FUNCTIONALITY WAS STABLE!


With Ubuntu, a single common code base is taken care of by different  
groups serving different clientele.  Canonical Inc. will take money  
from corporate customers who need help managing systems or getting  
new functionality.  Volunteers explore new functionality and move  
forward on the process of producing open source solutions.  At  
particular points a particular version is flagged for longer term  
support, but the primary focus is not on exploring new functionality,  
or on back porting popular functionality to an ancient codebase  
tailored to customers afraid of change. It is simply:  Produce a  
usable experience with a balance of stable code and new  
functionality.  What Fedora-based spin has this focus?


-Bill



William Cattey
Linux Platform Coordinator
MIT Information Services & Technology

N42-040M, 617-253-0140, [EMAIL PROTECTED]
http://web.mit.edu/wdc/www/


On Dec 24, 2007, at 5:38 PM, Jon Stanley wrote:

On Dec 24, 2007 4:19 PM, Jeroen van Meeuwen <[EMAIL PROTECTED]>  
wrote:



I don't think Ubuntu LTS gives you the latest and greatest unless you
upgrade, does it?


I think that's the whole point.


Same with CentOS; although it might be supported longer then you are
going to use it, whenever you feel you want newer software you  
upgrade

to the next release. Meanwhile, it's stable.


Correct - however security updates are backported to the "old"
versions of the software.


The Fedora Project moves in with EPEL, Extra Packages for Enterprise
Linux, perfectly suitable for a CentOS machine and with the same  
release

and 'support' cycle.


Not entirely sure what you mean here.  I think what was being called
for was a release whereby it's "supported" (with security updates,
etc) beyond the current 1 year, however perhaps not as much as the 7
years that RHEL is supported.

However, as Matthew said in the e-mail that came in as I was writing
this, there was little interest in Fedora Legacy when it existed.
What makes us think that there's more of a demand now?  It's either
the short, bleeding edge release cycle of Fedora as we know it, or the
long release cycle of RHEL.  Both serve different purposes.

-Jon

--
Fedora-marketing-list mailing list
Fedora-marketing-list@redhat.com
https://www.redhat.com/mailman/listinfo/fedora-marketing-list


--
Fedora-marketing-list mailing list
Fedora-marketing-list@redhat.com
https://www.redhat.com/mailman/listinfo/fedora-marketing-list


Re: Fedora more successful, developer-wise, than Ubuntu

2007-12-24 Thread Jeroen van Meeuwen

Jon Stanley wrote:

The Fedora Project moves in with EPEL, Extra Packages for Enterprise
Linux, perfectly suitable for a CentOS machine and with the same release
and 'support' cycle.


Not entirely sure what you mean here.  I think what was being called
for was a release whereby it's "supported" (with security updates,
etc) beyond the current 1 year, however perhaps not as much as the 7
years that RHEL is supported.



EPEL is a Fedora Project effort for Enterprise Linux providing extra 
packages and updates to those packages for the same amount of time the 
Enterprise Linux distribution it ships for is supported. I don't really 
know what their policy is on actively back-porting fixes for security 
issues, but as far as I know you can at least log bugs again an EPEL 
package until the end of the release's lifecycle.



However, as Matthew said in the e-mail that came in as I was writing
this, there was little interest in Fedora Legacy when it existed.
What makes us think that there's more of a demand now?  It's either
the short, bleeding edge release cycle of Fedora as we know it, or the
long release cycle of RHEL.  Both serve different purposes.



Yes, and the shorter release cycle for Fedora -at least in my opinion- 
helps us to do what we do best, moving forward, not "wasting" resources 
to what Enterprise Linux does best, being stable.


Kind regards,

Jeroen van Meeuwen
-kanarip

--
Fedora-marketing-list mailing list
Fedora-marketing-list@redhat.com
https://www.redhat.com/mailman/listinfo/fedora-marketing-list


Re: Fedora more successful, developer-wise, than Ubuntu

2007-12-24 Thread Jon Stanley
On Dec 24, 2007 4:19 PM, Jeroen van Meeuwen <[EMAIL PROTECTED]> wrote:

> I don't think Ubuntu LTS gives you the latest and greatest unless you
> upgrade, does it?

I think that's the whole point.

> Same with CentOS; although it might be supported longer then you are
> going to use it, whenever you feel you want newer software you upgrade
> to the next release. Meanwhile, it's stable.

Correct - however security updates are backported to the "old"
versions of the software.

> The Fedora Project moves in with EPEL, Extra Packages for Enterprise
> Linux, perfectly suitable for a CentOS machine and with the same release
> and 'support' cycle.

Not entirely sure what you mean here.  I think what was being called
for was a release whereby it's "supported" (with security updates,
etc) beyond the current 1 year, however perhaps not as much as the 7
years that RHEL is supported.

However, as Matthew said in the e-mail that came in as I was writing
this, there was little interest in Fedora Legacy when it existed.
What makes us think that there's more of a demand now?  It's either
the short, bleeding edge release cycle of Fedora as we know it, or the
long release cycle of RHEL.  Both serve different purposes.

-Jon

-- 
Fedora-marketing-list mailing list
Fedora-marketing-list@redhat.com
https://www.redhat.com/mailman/listinfo/fedora-marketing-list


Re: Fedora more successful, developer-wise, than Ubuntu

2007-12-24 Thread Matthew Miller
On Mon, Dec 24, 2007 at 04:02:39PM -0600, Jon Stanley wrote:
> and it's derivatives, a la CentOS.  Someone looking for a free as in
> {speech,beer} distribution with long term support I tend to point
> towards CentOS, but maybe there is middle ground between that and the
> current Fedora that we don't have.

There was, in Fedora Legacy. But there turned out to be too little interest.

-- 
Matthew Miller   [EMAIL PROTECTED]  
Boston University Linux  -->  

-- 
Fedora-marketing-list mailing list
Fedora-marketing-list@redhat.com
https://www.redhat.com/mailman/listinfo/fedora-marketing-list


Re: Fedora more successful, developer-wise, than Ubuntu

2007-12-24 Thread Jeroen van Meeuwen

Jon Stanley wrote:

On Dec 24, 2007 2:49 PM, Jeff Spaleta <[EMAIL PROTECTED]> wrote:


Point to specific functionality... open functionality that Fedora
doesn't have that we should.


I think that what is being referred to here is the Ubuntu 'LTS'
releases, that get long term support updates.  I 'm kind of on the
fence on whether or not we should do this - it means extra developer
cycles that may or may not exist, and the fact that we *do* have RHEL
and it's derivatives, a la CentOS.  Someone looking for a free as in
{speech,beer} distribution with long term support I tend to point
towards CentOS, but maybe there is middle ground between that and the
current Fedora that we don't have.



I don't think Ubuntu LTS gives you the latest and greatest unless you 
upgrade, does it?


Same with CentOS; although it might be supported longer then you are 
going to use it, whenever you feel you want newer software you upgrade 
to the next release. Meanwhile, it's stable.


The Fedora Project moves in with EPEL, Extra Packages for Enterprise 
Linux, perfectly suitable for a CentOS machine and with the same release 
and 'support' cycle.


Kind regards,

Jeroen van Meeuwen
-kanarip

--
Fedora-marketing-list mailing list
Fedora-marketing-list@redhat.com
https://www.redhat.com/mailman/listinfo/fedora-marketing-list


Re: Fedora more successful, developer-wise, than Ubuntu

2007-12-24 Thread Jon Stanley
On Dec 24, 2007 2:49 PM, Jeff Spaleta <[EMAIL PROTECTED]> wrote:

> Point to specific functionality... open functionality that Fedora
> doesn't have that we should.

I think that what is being referred to here is the Ubuntu 'LTS'
releases, that get long term support updates.  I 'm kind of on the
fence on whether or not we should do this - it means extra developer
cycles that may or may not exist, and the fact that we *do* have RHEL
and it's derivatives, a la CentOS.  Someone looking for a free as in
{speech,beer} distribution with long term support I tend to point
towards CentOS, but maybe there is middle ground between that and the
current Fedora that we don't have.

-- 
Fedora-marketing-list mailing list
Fedora-marketing-list@redhat.com
https://www.redhat.com/mailman/listinfo/fedora-marketing-list


Re: Fedora more successful, developer-wise, than Ubuntu

2007-12-24 Thread Jeff Spaleta
On Dec 23, 2007 7:46 PM, William Cattey <[EMAIL PROTECTED]> wrote:
> I am concerned that Fedora and Red Hat are losing mindshare in a way
> that a few years down the line will "kill the seed corn."  There's no
> nice middle ground between Red Hat Enterprise with well established
> functionality, and Fedora with bleeding edge functionality.  Well
> there is
> but it's called Ubuntu.

Point to specific functionality... open functionality that Fedora
doesn't have that we should.
If you are talking about nvidia or ati proprietary drivers out of the
box.. then that's a non-starter.  What piece of open technology is
Fedora not providing that let's people
"working with the same stuff that the guy next door runs on the Mac or
under Window"

It's real easy to speak in generalities and yet not actually say
anything significant.

>
> For the people already committed to the development community, Fedora
> is perfect just as it is.
> But for people not sure whether they want to use Linux, or join the
> developer community, Ubuntu works, but Fedora has issues.

Every distribution has issues.   What in the Fedora Desktop LiveCD doesn't work?

>
> Perhaps those who feel as Mukul Dharwadkar and myself will find a way
> to provide a particular spin on Fedora that will be an intermediate
> stage between Fedora as it is now, and Enterprise such that the
> developer community will build from the Fedora code base, not the
> less interesting, but more usable Debian/Ubuntu code base.

Again isn't the Fedora Desktop Livecd exactly what you want?

-jef

-- 
Fedora-marketing-list mailing list
Fedora-marketing-list@redhat.com
https://www.redhat.com/mailman/listinfo/fedora-marketing-list