Wild and crazy times for the development tree

[Jeremy Katz]

With the release of Fedora Core 5, the development tree is now open for
things to continue forward.  So if you've been following it purely to
get updates for the FC5 test releases, you'll probably want to grab the
fedora-release package from the FC5 release instead.  If you want to
keep testing and helping to develop things for Fedora Core 6, expect for
some fun to pop up as always.

Jeremy

Re: Wild and crazy times for the development tree

[Ralf Ertzinger]

Hi.

On Mon, 20 Mar 2006 11:59:57 -0500, Jeremy Katz wrote:

> to grab the fedora-release package from the FC5 release instead.  If
> you want to keep testing and helping to develop things for Fedora
> Core 6, expect for some fun to pop up as always.

Sooo... what are we going to break first?

Re: Wild and crazy times for the development tree

[Rahul Sundaram]

Ralf Ertzinger wrote:


Hi.

On Mon, 20 Mar 2006 11:59:57 -0500, Jeremy Katz wrote:

 


to grab the fedora-release package from the FC5 release instead.  If
you want to keep testing and helping to develop things for Fedora
Core 6, expect for some fun to pop up as always.
   



Sooo... what are we going to break first?

 


Try and see?

--
Rahul 

Re: Wild and crazy times for the development tree

[Peter Robinson]

> > to grab the fedora-release package from the FC5 release instead.  If
> > you want to keep testing and helping to develop things for Fedora
> > Core 6, expect for some fun to pop up as always.
>
> Sooo... what are we going to break first?

I'd vote on some X funnies at some stage as the 3D stuff that didn't
make FC5 comes in at some stage bring it on!!!

Pete

Re: Wild and crazy times for the development tree

[Arjan van de Ven]

On Mon, 2006-03-20 at 18:08 +0100, Ralf Ertzinger wrote:
> Hi.
> 
> On Mon, 20 Mar 2006 11:59:57 -0500, Jeremy Katz wrote:
> 
> > to grab the fedora-release package from the FC5 release instead.  If
> > you want to keep testing and helping to develop things for Fedora
> > Core 6, expect for some fun to pop up as always.
> 
> Sooo... what are we going to break first?
> 

I would love to see the -fstack-protector support for the kernel go in..

Re: Wild and crazy times for the development tree

[Paul F. Johnson]

Hi,

> > to grab the fedora-release package from the FC5 release instead.  If
> > you want to keep testing and helping to develop things for Fedora
> > Core 6, expect for some fun to pop up as always.
> 
> Sooo... what are we going to break first?

X is always guaranteed to raise a laugh ;-p

TTFN

Paul
-- 
"Logic, my dear Zoe, is merely the ability to be wrong with authority" -
Dr Who

Re: Wild and crazy times for the development tree

[Richard Hally]

Arjan van de Ven wrote:

On Mon, 2006-03-20 at 18:08 +0100, Ralf Ertzinger wrote:

Hi.

On Mon, 20 Mar 2006 11:59:57 -0500, Jeremy Katz wrote:


to grab the fedora-release package from the FC5 release instead.  If
you want to keep testing and helping to develop things for Fedora
Core 6, expect for some fun to pop up as always.

Sooo... what are we going to break first?



I would love to see the -fstack-protector support for the kernel go in..


What does the schedule look like for FC6?
I would like to see a 5 month schedule:
March 20 - April 20 == development
then
3 weeks;
freeze for test1 May 12;
test1 out around May 19;
3 weeks;
freeze for test2 June 9;
test 2 out around June 16;
3 weeks;
freeze for test3 July 7;
test3 out around July 14;
4 weeks (bug fixes only);
freeze for final around Aug 11;
FC6 around Aug 20!

With a schedule like this, when there is the inevitable slippage we will 
still be around 6 months between releases.

Re: Wild and crazy times for the development tree

[Arjan van de Ven]

On Mon, 2006-03-20 at 13:00 -0500, Richard Hally wrote:
> Arjan van de Ven wrote:
> > On Mon, 2006-03-20 at 18:08 +0100, Ralf Ertzinger wrote:
> >> Hi.
> >>
> >> On Mon, 20 Mar 2006 11:59:57 -0500, Jeremy Katz wrote:
> >>
> >>> to grab the fedora-release package from the FC5 release instead.  If
> >>> you want to keep testing and helping to develop things for Fedora
> >>> Core 6, expect for some fun to pop up as always.
> >> Sooo... what are we going to break first?
> >>
> > 
> > I would love to see the -fstack-protector support for the kernel go in..
> > 
> What does the schedule look like for FC6?
> I would like to see a 5 month schedule:

personally I think the current 9 month schedule wasn't too bad (ok the
end slipped too much but lets ignore that bit); it gives enough time to
do fundamental improvements. For fc6 it would be nice if boot speed was
further improved for example, and since initscripts are tricky and need
lots of testing... a bit of extra time would be neat

Re: Wild and crazy times for the development tree

[David Nielsen]

man, 20 03 2006 kl. 19:45 +0100, skrev Arjan van de Ven:

> personally I think the current 9 month schedule wasn't too bad (ok the
> end slipped too much but lets ignore that bit); it gives enough time to
> do fundamental improvements. For fc6 it would be nice if boot speed was
> further improved for example, and since initscripts are tricky and need
> lots of testing... a bit of extra time would be neat

I tend to agree, the 9 month cycle worked wonderfully, Fedora Core 5 is
by far the best release the Fedora Project has put out yet and as a
tester I enjoyed having that extra time to see new fundamental changes
take place and get bugs tracked down.

If the new init system can be delivered this cycle I would love to see
it, I don't see 9 month cycles being a price to pay, more likely it's a
benefit for most everybody (opinion taken from the most reliable source
in the world.. my backside).

But let's get a clear roadmap down this time, what features are
essential for the next cycle?

- David

Re: Wild and crazy times for the development tree

[Jeremy Katz]

On Mon, 2006-03-20 at 21:40 +0100, David Nielsen wrote:
> man, 20 03 2006 kl. 19:45 +0100, skrev Arjan van de Ven:
> > personally I think the current 9 month schedule wasn't too bad (ok the
> > end slipped too much but lets ignore that bit); it gives enough time to
> > do fundamental improvements. For fc6 it would be nice if boot speed was
> > further improved for example, and since initscripts are tricky and need
> > lots of testing... a bit of extra time would be neat
> 
> I tend to agree, the 9 month cycle worked wonderfully, Fedora Core 5 is
> by far the best release the Fedora Project has put out yet and as a
> tester I enjoyed having that extra time to see new fundamental changes
> take place and get bugs tracked down.

Realistically, I don't think the 9 month cycle really helped that much
except for one very specific case of the underlying installer changes.
And realistically, if it had been a six month cycle instead, those would
have been worked out then as well.  For everything else, if we had been
on a six month cycle, some of them would have made FC5 and some would
have made FC6.  But guess what, that's going to be true no matter _when_
you actually cut a release.  It's the price of doing releases more than
once every three years :-)

Jeremy

Re: Wild and crazy times for the development tree

[Paul F. Johnson]

Hi,

> But let's get a clear roadmap down this time, what features are
> essential for the next cycle?

A *much* reduced core size (5 CDs + rescue is getting a bit much) and
large reduction in the overall memory overhead. I know FC is fantastic,
but given you need 128Mb for a text only version and 512Mb for a desktop
environment, it's becoming hard to justify to the powers that be that
using Linux over WinXP is that good an option.

Slack 10 can run (text only) in 18Mb with a desktop environment in 64Mb
- Debian (sorry for swearing on this list) is a whole lot smaller than
FC in terms of memory again.

I think it was suggested that ISOs are made of FE. It may be time to do
this, but with quite a lot from FC moved to it. For example, gcc-gnat,
gfortran, objc and anything *not* mono/mcs (in other words, beagle,
fspot etc) should, IMHO, be in extras - and yes, I do use gfortran and
gnat. The OOo language packs should also be moved out - just keep in
french, german, spanish and any big userbases.

The same applies to Qt and KDE - quite a lot of the material in there
should be in extras, Qt (standard + devel) and some of the kde system
should be in Core. We also have a number of different database systems
in Core. Could a case not be made for trimming it down to MySQL and
postgresql with the rest again going to FE?

9 months is a long time for a release. Perhaps an interim release at the
5 month stage would be an advantage.

Obviously, these are just ideas *but* they do answer a growing number of
criticisms of FC.

TTFN

Paul
-- 
"ein zu starker starker Anblick kann Sie toten. Sie gegen gerade uber
den Rand mit dem festen Wissen des Wege vor Ihnen" - Linus Tordvals

Re: Wild and crazy times for the development tree

[Matthew Miller]

On Mon, Mar 20, 2006 at 09:40:10PM +0100, David Nielsen wrote:
> I tend to agree, the 9 month cycle worked wonderfully, Fedora Core 5 is
> by far the best release the Fedora Project has put out yet and as a
> tester I enjoyed having that extra time to see new fundamental changes
> take place and get bugs tracked down.

I was intrigued by your blog suggestion of alternating 9 month devel / 4
month stabilization cycles.  But I also think
your own objection (people will not take the 9-month release seriously
enough) is pretty strong.

As someone crazy enough to try to use Fedora for real work, the 9 month
cycle is awesome.


-- 
Matthew Miller   mat...@mattdm.org  
Boston University Linux  -->  

Re: Wild and crazy times for the development tree

[Jeremy Katz]

On Mon, 2006-03-20 at 21:15 +, Paul F. Johnson wrote:
> > But let's get a clear roadmap down this time, what features are
> > essential for the next cycle?
> 
> A *much* reduced core size (5 CDs + rescue is getting a bit much) and
> large reduction in the overall memory overhead. I know FC is fantastic,
> but given you need 128Mb for a text only version and 512Mb for a desktop
> environment, it's becoming hard to justify to the powers that be that
> using Linux over WinXP is that good an option.

We're actually probably significantly better than FC4 on memory overhead
for the most part.  Installation is a little higher overhead this time
around and most of the reason for bumping the recommendation.  

> Slack 10 can run (text only) in 18Mb with a desktop environment in 64Mb
> - Debian (sorry for swearing on this list) is a whole lot smaller than
> FC in terms of memory again.

It all depends on what you're defining as your desktop environment.  We
could do fvwm instead, but I really don't think that's what most people
want.

> I think it was suggested that ISOs are made of FE. It may be time to do
> this, but with quite a lot from FC moved to it. For example, gcc-gnat,
> gfortran, objc and anything *not* mono/mcs (in other words, beagle,
> fspot etc) should, IMHO, be in extras - and yes, I do use gfortran and
> gnat. The OOo language packs should also be moved out - just keep in
> french, german, spanish and any big userbases.

The problem is that a lot of this is split over src.rpms that *must* be
in Core.  And right now, there's no way to do that split.  Given where
some things stand, I think that we can start thinking about tools to
make it easier for building CD sets and shadow repositories in the FC6
timeframe and then start really rethinking a release after that.

> The same applies to Qt and KDE - quite a lot of the material in there
> should be in extras, Qt (standard + devel) and some of the kde system
> should be in Core. We also have a number of different database systems
> in Core. Could a case not be made for trimming it down to MySQL and
> postgresql with the rest again going to FE?

We don't really ship any substantial (ie, non-embedded) databases other
than mysql and postgresql that I know of.  And changing the embedded
database for a piece of software isn't a trivial thing.

Jeremy

Re: Wild and crazy times for the development tree

[David Woodhouse]

On Mon, 2006-03-20 at 16:32 -0500, Jeremy Katz wrote:
> The problem is that a lot of this is split over src.rpms that *must*
> be in Core.  And right now, there's no way to do that split.

That _really_ isn't a hard problem to solve. To have certain binary
packages built in the Core build system and then wind up in the Extras
repositories really isn't rocket science.

-- 
dwmw2

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Ralf Ertzinger wrote:

Hi.

On Mon, 20 Mar 2006 11:59:57 -0500, Jeremy Katz wrote:



to grab the fedora-release package from the FC5 release instead.  If
you want to keep testing and helping to develop things for Fedora
Core 6, expect for some fun to pop up as always.



Sooo... what are we going to break first?


All proprietary drivers?  ;o)



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

Re: Wild and crazy times for the development tree

[Dave Jones]

On Mon, Mar 20, 2006 at 04:42:09PM -0500, Mike A. Harris wrote:

 > >>to grab the fedora-release package from the FC5 release instead.  If
 > >>you want to keep testing and helping to develop things for Fedora
 > >>Core 6, expect for some fun to pop up as always.
 > >
 > >Sooo... what are we going to break first?
 > 
 > All proprietary drivers?  ;o)

Yawn, been there, done that :-P

Dave

-- 
http://www.codemonkey.org.uk

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Peter Robinson wrote:

to grab the fedora-release package from the FC5 release instead.  If
you want to keep testing and helping to develop things for Fedora
Core 6, expect for some fun to pop up as always.


Sooo... what are we going to break first?



I'd vote on some X funnies at some stage as the 3D stuff that didn't
make FC5 comes in at some stage bring it on!!!


Hehe.

There has been a fair bit of upstream X.Org stable branch commits
lately for the 2D ati driver, and X server - mostly benh's memory
map fixes and other various Radeon fixes.  There are some known
problems I believe, but once they're straightened out, I think there
is going to be a new upstream "ati" driver release from the stable
branch.

Adam Jackson is also working towards a new stable-branch Xorg server
release.

When each of these become available upstream as tarballed releases,
we'll merge them into rawhide to get the testing process going.  Down
the road once any regressions are worked out, we'll probably pull them
both into FC5-updates also.

R300 DRI support is quite unstable however, and will probably be held
back until it shows signs of greater stability.  When I say this, I
do so knowing that some people who have tried it have found it to be
fairly stable for their own hardware/situation.  My definition of
"stable" is "works stable for majority of people, with little to no
bug reports, and does not cause massive bugzilla influx of bugs when
the driver is provided".

So, R300 dri support will be something you need to compile for yourself
for the time being if desired.  Also, our kernel does not have the
R300+ PCI IDs in it anymore, so you may need to recompile your kernel
as well.  Not sure what davej's specific plans are for that, but I hope
he keeps the R300 PCI IDs out of the kernel at least until the r300
DRI driver is remotely useable large-scale.



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

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Richard Hally wrote:

Arjan van de Ven wrote:


On Mon, 2006-03-20 at 18:08 +0100, Ralf Ertzinger wrote:


Hi.

On Mon, 20 Mar 2006 11:59:57 -0500, Jeremy Katz wrote:


to grab the fedora-release package from the FC5 release instead.  If
you want to keep testing and helping to develop things for Fedora
Core 6, expect for some fun to pop up as always.


Sooo... what are we going to break first?



I would love to see the -fstack-protector support for the kernel go in..


What does the schedule look like for FC6?
I would like to see a 5 month schedule:
March 20 - April 20 == development


One month of development?

Hahahahaahah!


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

Re: Wild and crazy times for the development tree

[Dave Jones]

On Mon, Mar 20, 2006 at 04:49:42PM -0500, Mike A. Harris wrote:
 > So, R300 dri support will be something you need to compile for yourself
 > for the time being if desired.  Also, our kernel does not have the
 > R300+ PCI IDs in it anymore, so you may need to recompile your kernel
 > as well.  Not sure what davej's specific plans are for that, but I hope
 > he keeps the R300 PCI IDs out of the kernel at least until the r300
 > DRI driver is remotely useable large-scale.

not that it seems to matter. X seems to be doing a 'modprobe radeon'
when it sees anything ATI.  Chopping the IDs out stops it being
initialised, but it appears that X does _something_ differently
just by having the module loaded even if its not using it.

I can keep them chopped out for the time being, but the downside is
we're not going to get a lot of coverage testing, so how will we know
when its 'good enough' ?

Dave

-- 
http://www.codemonkey.org.uk

Re: Wild and crazy times for the development tree

[Stephen J. Smoogen]

On 3/20/06, Jeremy Katz  wrote:
> On Mon, 2006-03-20 at 21:40 +0100, David Nielsen wrote:
> > man, 20 03 2006 kl. 19:45 +0100, skrev Arjan van de Ven:
> > > personally I think the current 9 month schedule wasn't too bad (ok the
> > > end slipped too much but lets ignore that bit); it gives enough time to
> > > do fundamental improvements. For fc6 it would be nice if boot speed was
> > > further improved for example, and since initscripts are tricky and need
> > > lots of testing... a bit of extra time would be neat
> >
> > I tend to agree, the 9 month cycle worked wonderfully, Fedora Core 5 is
> > by far the best release the Fedora Project has put out yet and as a
> > tester I enjoyed having that extra time to see new fundamental changes
> > take place and get bugs tracked down.
>
> Realistically, I don't think the 9 month cycle really helped that much
> except for one very specific case of the underlying installer changes.
> And realistically, if it had been a six month cycle instead, those would

The problem with 6 month release cycles is developer burn-out. You had
3 extra months of fluff time that I saw lots of developers come in and
out and up-to-speed, and time for some developers to hand other stuff
off. In the previous release cycles.. I saw more people just bail out
after their 2nd release because they just didnt have any energy left
to devote and there was no time to do so. You can only push people
along a marathon schedule for so long before they break..

If you go back to a 6 month schedule, you will need to pair back how
many CD's are in the core set by at least 2 or 3 cd's. That would give
developers 6 weeks of fluff time to regain their energy after a
release and then be able to come back and say "you know this brick
wall we keep hitting on this project? Did you know there was a door 2
feet down?"

> have been worked out then as well.  For everything else, if we had been
> on a six month cycle, some of them would have made FC5 and some would
> have made FC6.  But guess what, that's going to be true no matter _when_
> you actually cut a release.  It's the price of doing releases more than
> once every three years :-)
>
> Jeremy
>
> --
> fedora-devel-list mailing list
> fedora-devel-list@redhat.com
> https://www.redhat.com/mailman/listinfo/fedora-devel-list
>


--
Stephen J Smoogen.
CSIRT/Linux System Administrator

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Paul F. Johnson wrote:

Hi,



But let's get a clear roadmap down this time, what features are
essential for the next cycle?



A *much* reduced core size (5 CDs + rescue is getting a bit much) and
large reduction in the overall memory overhead. I know FC is fantastic,
but given you need 128Mb for a text only version and 512Mb for a desktop
environment, it's becoming hard to justify to the powers that be that
using Linux over WinXP is that good an option.


Are you implying that the only reason one would choose Linux over
Windows XP, is because it has lower memory and system requirements?

That is one particular usage case scenario that might tilt the scales
of a given rollout to a Linux based solution over an XP based one, but
there are far far more other reasons for using Linux over XP than just
"lower memory/disk footprint".

Having expressed this though, I also agree with you completely, in that
it would be nice to see the overall memory footprint reduced in a sane
manner without tossing out desired functionality, etc.



Slack 10 can run (text only) in 18Mb with a desktop environment in 64Mb
- Debian (sorry for swearing on this list) is a whole lot smaller than
FC in terms of memory again.


Sure, different distributions focus on different areas, with some
levels of overlapping.  I don't think it is even possible at all
to make a complete one-size-fits all distribution however, because
different people have different and often conflicting
goals/requirements.



I think it was suggested that ISOs are made of FE. It may be time to do
this, but with quite a lot from FC moved to it. For example, gcc-gnat,
gfortran, objc and anything *not* mono/mcs (in other words, beagle,
fspot etc) should, IMHO, be in extras - and yes, I do use gfortran and
gnat. The OOo language packs should also be moved out - just keep in
french, german, spanish and any big userbases.


IMHO, moving more and more stuff out of Core and into Extras is an
overall good idea, so long as the infrastructure is present in
_advance_ to make it easy to install the stuff that has moved to
Extras, both at OS install time and later, and without requiring
mandatory network access.  ie:  Fedora Extras on CD, kindof like
powertools was before, but with anaconda support for that.

Anaconda support for additional arbitrary CDs would be nice.



The same applies to Qt and KDE - quite a lot of the material in there
should be in extras, Qt (standard + devel) and some of the kde system
should be in Core. We also have a number of different database systems
in Core. Could a case not be made for trimming it down to MySQL and
postgresql with the rest again going to FE?


I use KDE mainly, but with mostly GTK apps.  Moving KDE to Extras would
be ok I guess, so long as I can still choose to install it at install
time, and not have to mess around a lot post install.


9 months is a long time for a release. Perhaps an interim release at the
5 month stage would be an advantage.

Obviously, these are just ideas *but* they do answer a growing number of
criticisms of FC.


The only reason there might be growing numbers of criticisms of FC,
is that the userbase is expanding, so it makes sense that as the
number of users increase in volume that the number of both praises
and criticisms will increase in volume proportionately.

If we had a tug of war contest, in which 6 month cycle was the centre
of the tug of war starting point, and the team pulling to pull it
towards 4-5 months was on one side, and the team pulling on the
7-12 months was on the other side, I think we would clearly see the
"6 month flag" start out at the center, move a bit to the left, then
back to center and a bit to the right, and keep doing that more or
less indefinitely, until everyone passed out from exhaustion, or the
rope broke in half.

It makes more sense to base the release date off the features planned
for that release and an estimate of how much time is required to
obtain those features.  That includes in-house developed features,
and accounting for upstream release dates of various software that
are desired in the release.

That's much more important IMHO than arbitrary dates picked because
some people want a faster release cycle or longer release cycle for
their own individual preferences/desires.



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

Re: Wild and crazy times for the development tree

[Ian Pilcher]

Mike A. Harris wrote:
> 
> All proprietary drivers?  ;o)
> 

I can't help wondering...

What do you guys do when you want decent 3D performance?

-- 

Ian Pilcheri.pilc...@comcast.net

Re: Wild and crazy times for the development tree

[Paul F. Johnson]

> Hi,

>  
> > A *much* reduced core size (5 CDs + rescue is getting a bit much) and
> > large reduction in the overall memory overhead. I know FC is fantastic,
> > but given you need 128Mb for a text only version and 512Mb for a desktop
> > environment, it's becoming hard to justify to the powers that be that
> > using Linux over WinXP is that good an option.
> 
> Are you implying that the only reason one would choose Linux over
> Windows XP, is because it has lower memory and system requirements?

In part, yes. On a voluntary basis, I work for a local charity with some
pretty low end boxes (I think the highest spec machine is P3-366 or
thereabouts), all with 256Mb of memory. Linux runs fine on them.
However, due to the pointy haired boss (and he actually has pointy
hair!) reading distrowatch and seeing that 512Mb is recommended and has
said that given that, why can't he have his XP box and because of that,
we should have XP boxes. I know there is a massive problem with the
logic in that...

> That is one particular usage case scenario that might tilt the scales
> of a given rollout to a Linux based solution over an XP based one, but
> there are far far more other reasons for using Linux over XP than just
> "lower memory/disk footprint".

Yep. I completely agree. Memory is dirt cheap and if wasn't for me
giving my time free, the cost of Linux support would probably mean it is
more economical to go with the Borg and OOo.

> Having expressed this though, I also agree with you completely, in that
> it would be nice to see the overall memory footprint reduced in a sane
> manner without tossing out desired functionality, etc.

It would be interesting to see if the biggest of the hogs is the desktop
environment. I have noticed that while gnome is far better than previous
incarnations, it has somewhat become, well, bloated.

> > Slack 10 can run (text only) in 18Mb with a desktop environment in 64Mb
> > - Debian (sorry for swearing on this list) is a whole lot smaller than
> > FC in terms of memory again.
> 
> Sure, different distributions focus on different areas, with some
> levels of overlapping.  I don't think it is even possible at all
> to make a complete one-size-fits all distribution however, because
> different people have different and often conflicting
> goals/requirements.

True. If we leave aside Slack and concentrate on Debian, SuSE and
Mandriva (in otherwords, the closest thing to competition of the big
distros), Mandriva looks to have the same sort of memory requirements to
functionality ratio as FC, yet SuSE and Debian have smaller footprints
and arguably, greater functionality.

I'm not going to get into the argument over if Debian should be included
(as stable), I'm using it as not only is it one of the most revered, but
is the basis for the likes of Umbunto and Knoppix.

> > I think it was suggested that ISOs are made of FE. It may be time to do
> > this, but with quite a lot from FC moved to it. For example, gcc-gnat,
> > gfortran, objc and anything *not* mono/mcs (in other words, beagle,
> > fspot etc) should, IMHO, be in extras - and yes, I do use gfortran and
> > gnat. The OOo language packs should also be moved out - just keep in
> > french, german, spanish and any big userbases.
> 
> IMHO, moving more and more stuff out of Core and into Extras is an
> overall good idea, so long as the infrastructure is present in
> _advance_ to make it easy to install the stuff that has moved to
> Extras, both at OS install time and later, and without requiring
> mandatory network access.  ie:  Fedora Extras on CD, kindof like
> powertools was before, but with anaconda support for that.

That would be a perfect solution and with another 9 months between FC5
and 6, is probably doable. 

I've noted the comments about the src.rpms and the "not rocket science"
to move them over to Extras once they've been built. I can see both
points of view there. I would say it would probably be easier for the
likes of gnat and gfortran as well as non-core mono pieces to have them
built in Extras than build in core and shift across.

> Anaconda support for additional arbitrary CDs would be nice.

Well, firstboot does have that sort of support, sort of...

> > The same applies to Qt and KDE - quite a lot of the material in there
> > should be in extras, Qt (standard + devel) and some of the kde system
> > should be in Core. We also have a number of different database systems
> > in Core. Could a case not be made for trimming it down to MySQL and
> > postgresql with the rest again going to FE?
> 
> I use KDE mainly, but with mostly GTK apps.  Moving KDE to Extras would
> be ok I guess, so long as I can still choose to install it at install
> time, and not have to mess around a lot post install.

I'm not suggesting a whole sale shift of KDE to FE, that would be silly
and very counter productive (just look what happened when it was
suggested that RH would be gnome only and how that was blown out of
proportion!). I'm saying th

Re: Wild and crazy times for the development tree

[Paul F. Johnson]

Hi,

> > All proprietary drivers?  ;o)

> What do you guys do when you want decent 3D performance?

Take viagra! ;-p

Seriously, this box I'm on now has a monster nVidia card. However, it's
my development box, so I'm not that worried. My son uses ATI, my other
half is on SiS and my laptop in Intel. All three are fine with 3D.

TTFN

Paul
-- 
"ein zu starker starker Anblick kann Sie toten. Sie gegen gerade uber
den Rand mit dem festen Wissen des Wege vor Ihnen" - Linus Tordvals

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Dave Jones wrote:

On Mon, Mar 20, 2006 at 04:49:42PM -0500, Mike A. Harris wrote:
 > So, R300 dri support will be something you need to compile for yourself
 > for the time being if desired.  Also, our kernel does not have the
 > R300+ PCI IDs in it anymore, so you may need to recompile your kernel
 > as well.  Not sure what davej's specific plans are for that, but I hope
 > he keeps the R300 PCI IDs out of the kernel at least until the r300
 > DRI driver is remotely useable large-scale.

not that it seems to matter. X seems to be doing a 'modprobe radeon'
when it sees anything ATI.


Yeah.  Not sure when exactly that changed, but it's probably because
there is now OSS support for Mach64/R128/Radeon for 3D I'm guessing.
We don't ship the mach64 DRI driver, or the R300+ one, but the X
server still seems to have static lists of stuff which is a bit brain
dead if you ask me.  ;o)



Chopping the IDs out stops it being
initialised, but it appears that X does _something_ differently
just by having the module loaded even if its not using it.


Indeed, it is looking that way.  We'll likely figure that out during
FC6 and hopefully get it fixed in Xorg 7.1


I can keep them chopped out for the time being, but the downside is
we're not going to get a lot of coverage testing, so how will we know
when its 'good enough' ?


I'd say at a minimum - when there's a new upstream release which
contains CVS checkins to the code in question as a minimum.  If
no fixes have went into the code that didn't work before, it's
most likely still broken.  ;o)

For the 2D-only side of things, I'd want to wait at least until
the new 2D radeon driver comes out (stable branch) and let it
get tested for a while in rawhide.  Then hopefully get any new
regressions resolved.  After that, enabling the kernel side
again to see if it breaks anything would be following scientific
method.  I'd prefer to avoid re-enabling 2-3 things simultaneously,
and then trying to guess which one caused breakage to occur though.

By the time a new stable 2D driver has come out and had a reasonable
shakedown period, then have a kernel DRM update which re-enables the
kernel bits, and another shakedown period - a new Mesa is likely
to be out, if not for the 6.4.x track, then the 6.5.x/6.6 track
heading towards Xorg 7.1.  I think it'd be good to wait until then
before considering re-enabling R300 DRI driver support in Mesa, as
we'd just be enabling known-broken code to do it before then, and
getting nothing useful other than an increase in bug reports
hitting Red Hat bugzilla, which should go to X.Org bugzilla instead.
;o)




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

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Ian Pilcher wrote:

Mike A. Harris wrote:


All proprietary drivers?  ;o)




I can't help wondering...

What do you guys do when you want decent 3D performance?


That'll likely vary greatly from person to person, rather than there
being one single unified answer.

To be honest, for my own personal needs, I generally don't need
mandatory OpenGL 3D acceleration support in Linux for 99.9%
of what I use Linux for day to day.  In fact, my primary desktop
system usually has DRI disabled, as this machine is used almost
entirely for 2D only usage, and having DRI enabled just leaves
another potential source of instability for me.

Whenever I do need/want 3D enabled support for something in Linux,
I fire up another machine and use it instead, leaving my primary
desktop in a stable DRI-disabled state.

To answer the "when do I want decent 3D performance" question,
I must first answer "When do I want accelerated 3D at all?".

99.9% of the time I never need 3D acceleration in Linux
personally.  I disable the screensaver completely, or else pick
the "Matrix" saver, or some other boring simple screensaver, and
I don't use any 3D software in Linux generally.  The only real
use of OpenGL/3D software that I have would be for video games,
or for fun eye-candy stuff.  I generally turn eye-candy type of
stuff completely off on my computers that are intended for
productive work use, and I don't install video games on them
either, for the same reason.  ;o)

So, what do I use OpenGL for?  More or less for video games,
in the rare occurances that I actually have time to play games.
However, and I hate to admit this but I'm being totally honest
here - when I do actually have some spare time to play computer
games, I want to spend _all_ of that time actually playing the
games themselves and enjoying myself doing so, and spend zero of
my time trying to get the games to actually work.  As such, I
personally just play games in Windows XP and be done with it.

Please do not take that as a suggestion for what everyone else
should do however.  I admire the people out there who have the
dedication and spare time to install 3rd party emulators such
as winex or whatever is the coolest thing nowadays, and then
fiddle with whatever settings are needed, fiddle with drivers
and whatnot to get the stuff working under Linux.  I know many
people out there have tonnes of cool games and other software
running in Linux using one or more emulation layers, and I have
done so myself in the past as well.  But I just got tired of
spending 4-8 hours of downloading and compiling various things
and tweaking them to "maybe" get something to work in Linux,
and not actually have any time to _use_ the game or whatever it
was after that.  Since "fun spare time" is a limited resource
to me, I've decided to just play the games I bolted out $60
for in the OS they were designed for - at least for the time
being, than to fight the fight.

Having said that, I do look forward to some time in the future,
hopefully this year, in which I'll have some extra spare cycles
to tackle the various emulators and whatnot out there and maybe
get some stuff running in Linux again.  ;o)  I do very much hate
using Windows for anything, but I bite my tongue and use it on
occasion just to make better use of my time.

Now...  if I _was_ actually trying to get a game or some other
accelerated OpenGL software to work in Linux, I would use
whatever the best card I have on hand with OSS driver support
was at the given point in time.  The FireGL 8800 is what I
generally use for that purpose when I do enable DRI, and it
generally works quite well.  It generally meets _my_ needs,
however being a few generations old, and not having all the
latest bells and whistles, it is almost definitely not going
to meet everyone else out there's needs.  If I found it too
slow for something, I'd perhaps drop in a 9800Pro and roll
myself a custom rebuild of Mesa with the r300 dri driver
enabled and fiddle a bit.  Again, I don't expect that this
type of situation would be an acceptable or desired solution
for many other people out there, but it is something I can
personally live with for the time being, without having to
resort to proprietary drivers.

If I did have applications that I wanted to run mandatorily,
which required accelerated 3D performance, if running them
in Linux on the FireGL 8800 didn't cut it for me, I'd run them
in XP, or probably not run them at all, and find a new hobby.

That's my take on things for _my_ situation only.  Again though,
each person's situation is different, and each person has different
wants and needs, and are willing to compromise in different areas.
Some people compromise with proprietary drivers, others use OSS
drivers and compromise features or stability, others, like me
sometimes compromise by using a different OS that does what I
need/want for the given task.  To each his own.

There simply isn't a good general one-size-fits-all solution that
makes everyone 

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Paul F. Johnson wrote:


A *much* reduced core size (5 CDs + rescue is getting a bit much) and
large reduction in the overall memory overhead. I know FC is fantastic,
but given you need 128Mb for a text only version and 512Mb for a desktop
environment, it's becoming hard to justify to the powers that be that
using Linux over WinXP is that good an option.


Are you implying that the only reason one would choose Linux over
Windows XP, is because it has lower memory and system requirements?


In part, yes. On a voluntary basis, I work for a local charity with some
pretty low end boxes (I think the highest spec machine is P3-366 or
thereabouts), all with 256Mb of memory. Linux runs fine on them.
However, due to the pointy haired boss (and he actually has pointy
hair!) reading distrowatch and seeing that 512Mb is recommended and has
said that given that, why can't he have his XP box and because of that,
we should have XP boxes. I know there is a massive problem with the
logic in that...


Well actually, I wouldn't completely disagree with his logic.  In
that particular case, if hardware requirements are the greatest
concern, and Fedora has high requirements in the area, it is entirely
possible that Fedora really isn't a good choice for the particular
problem case.  As much as many of us would _like_ to see Fedora be
the end all be all general purpose solution to everything ...  it
sometimes is not.  ;o)

No harm in trying other Linux distributions to see if you can find one
with lower system requirements that allow you to avoid using XP though.



That is one particular usage case scenario that might tilt the scales
of a given rollout to a Linux based solution over an XP based one, but
there are far far more other reasons for using Linux over XP than just
"lower memory/disk footprint".


Yep. I completely agree. Memory is dirt cheap and if wasn't for me
giving my time free, the cost of Linux support would probably mean it is
more economical to go with the Borg and OOo.


A few years ago I set up a Linux solution for a small business, which
was very happy with it for a few years.  I provided support for a
reasonable fee for several years and they didn't need to contact me
very often, so they were happy.  However, the OS eventually went EOL,
and I informed them they really needed to do an upgrade to something
more modern.  They didn't want to take the risk of destabilizing what
worked so well for so long, and so delayed acting on that.  When they
finally contacted me to see if I could update them, was around the
time Fedora came out.  Previously they were using Red Hat Linux, and
I told them I could no longer support their EOL'd RHL box, and that
it would cost them far less to go with RHEL than to be paying me to
update the box every 6 months or so.

They didn't want to pay the subscription for RHEL, even though they
were paying me more than that anyway, and I pointed out that it would
be cheaper for them to do this, and for Red Hat to provide them
support directly.

In similar failed-logic, they considered switching to a Windows based
solution instead, and I told them it'd cost them a lot more to do that,
and it'd cost them a lot more to have someone coming over daily/weekly
to fix the problems they'd have for their particular usage.  They
agreed it could potentially be more costly, but still didn't make the
logic connection that RHEL would be cheaper TCO to them.  They wanted
to keep me on their emergency dialer.  ;o)

In the end, I left the decision up to them to decide to go with RHEL
or Windows, and I passed the administration of the EOL'd box (which was
quite minimal at the time) on to a friend instead, so I wouldn't have
to deal with any eventual security breach or other headache which was
inevitably likely to happen.

In the end, they decided to not make any decision, and I believe they
may still be using the ancient 6.2 based solution.  ;o)

So there's logic, illogic, and an entire new thing which defies
comprehension..  irrational illogic perhaps we could call it.  ;o)

Eventually, the system will likely fall over, and my phone will ring
off the hook.  Thank technology for caller ID!  ;o)



Having expressed this though, I also agree with you completely, in that
it would be nice to see the overall memory footprint reduced in a sane
manner without tossing out desired functionality, etc.


It would be interesting to see if the biggest of the hogs is the desktop
environment. I have noticed that while gnome is far better than previous
incarnations, it has somewhat become, well, bloated.


Indeed.  There is a tendency in OSS overall it seems for apps/libs/etc.
to sprout more and more dependencies over time just in general.  It is
very observable on the desktop, and in desktopish applications, but it
is also observable at the commandline and underlying OS level too.

Tough problem to solve I think.



I think it was suggested that ISOs are made of FE. It may be time to do
this, but with quite a lot from FC m

Re: Wild and crazy times for the development tree

[Ian Pilcher]

Mike A. Harris wrote:
> 
> Indeed.  There is a tendency in OSS overall it seems for apps/libs/etc.
> to sprout more and more dependencies over time just in general.  It is
> very observable on the desktop, and in desktopish applications, but it
> is also observable at the commandline and underlying OS level too.
> 
> Tough problem to solve I think.
> 

Partly due to the UNIX philosophy of compile-time feature selection.
It's solvable via dlopen, etc.; it would be nice to see the toolchains
help out here.

-- 

Ian Pilcheri.pilc...@comcast.net

Re: Wild and crazy times for the development tree

[Callum Lerwick]

On Mon, 2006-03-20 at 21:15 +, Paul F. Johnson wrote:
> A *much* reduced core size (5 CDs + rescue is getting a bit much)

IMHO its getting time to just forget CDs. Put out DVD isos only. (At the
very least, stop worrying about how many CDs core fills and worry about
fitting on one DVDR...)

Especially with yum backed network installs becoming available. If you
don't want to download and burn an entire DVD iso, just download a
10-50mb network install iso. This is what I do with debian on the rare
occasion I install a system from scratch. Since I only install a system
once every few years, its a huge waste to me to download and burn 5+
disks that I'm only going to use once.

Someone on IRC was arguing that in many parts of the world, bandwidth is
very expensive. I don't see how 5 CDs to download vs a DVD make a
difference here. The real question is, how common are DVD burners, and
most importantly, DVD readers in various parts of the world.

I suppose the thing to do is keep an eye on CD downloads vs DVD. Let the
community decide.


signature.asc
Description: This is a digitally signed message part

Re: Wild and crazy times for the development tree

[Callum Lerwick]

On Mon, 2006-03-20 at 20:23 -0500, Mike A. Harris wrote:
> Personally, I think it would be confusing as hell to have a single
> src.rpm generate binary rpms of which some are in Fedora Core and
> others are in Fedora Extras.  That would be made more confusing
> since Fedora has separate bugzilla product/etc. for Core and for
> Extras (and for Legacy).  Since bugzilla components themselves are
> based from the name of the src.rpm, one would have to file a bug
> report from a binary package from extras against Fedora Core for the
> src.rpm it was built from.
> 
> I can think of other complications and confusion that would ensue
> as well.  I'm not sure what to suggest to solve those issues however.

Less separation between core and extras. Like maybe none. (cough Debian
cough)


signature.asc
Description: This is a digitally signed message part

Re: Wild and crazy times for the development tree

[Dave Jones]

On Mon, Mar 20, 2006 at 10:04:46PM -0600, Callum Lerwick wrote:

 > Especially with yum backed network installs becoming available. If you
 > don't want to download and burn an entire DVD iso, just download a
 > 10-50mb network install iso. This is what I do with debian on the rare
 > occasion I install a system from scratch. Since I only install a system
 > once every few years, its a huge waste to me to download and burn 5+
 > disks that I'm only going to use once.
 > 
 > Someone on IRC was arguing that in many parts of the world, bandwidth is
 > very expensive. I don't see how 5 CDs to download vs a DVD make a
 > difference here. The real question is, how common are DVD burners, and
 > most importantly, DVD readers in various parts of the world.

DVD writers aren't anywhere near as commonplace as CD writers yet.
Looking around right now, I have 7 computers near me. CD writers outnumber
DVD writers 6:1.  (And the majority of the computers with CD writers are
less than 2 years old (two of them are <6 months old))

(ironically, the dvd writer is my 3 month old laptop)

I'm not claiming to be representative of the majority of Fedora users,
but given there are a lot of Fedora users less fortunate than myself,
I believe that discontinuing CD iso's would severely impact a huge portion
of our userbase.  It'd be interesting to see the download stats to see
if they agree with this.

Dave
-- 
http://www.codemonkey.org.uk

Re: Wild and crazy times for the development tree

[Ralf Corsepius]

On Mon, 2006-03-20 at 19:43 -0500, Mike A. Harris wrote:
> Ian Pilcher wrote:
> > Mike A. Harris wrote:
> > 
> >>All proprietary drivers?  ;o)
> >>
> > 
> > 
> > I can't help wondering...
> > 
> > What do you guys do when you want decent 3D performance?

Use the proprietary drivers ... :-)

> That'll likely vary greatly from person to person, rather than there
> being one single unified answer.

> Whenever I do need/want 3D enabled support for something in Linux,
> I fire up another machine and use it instead, leaving my primary
> desktop in a stable DRI-disabled state.
> 
> To answer the "when do I want decent 3D performance" question,
> I must first answer "When do I want accelerated 3D at all?".

> So, what do I use OpenGL for?  More or less for video games,

You are ignoring the fact, Linux has a strong user base in people with a
scientific/engineering/technical background ...

> Now...  if I _was_ actually trying to get a game or some other
> accelerated OpenGL software to work in Linux, I would use
> whatever the best card I have on hand with OSS driver support
> was at the given point in time. 

Whether you like it or not ... reality is different.

People are pragmatically using what they have/can get/are supplied with,
and will ditch the distro or even the OS if it doesn't suite to their
demands. Fortunately for Fedora, the proprietary drivers have worked
sufficiently well.

Ralf

Re: Wild and crazy times for the development tree

[Ralf Corsepius]

On Mon, 2006-03-20 at 17:29 -0500, Mike A. Harris wrote:
> Paul F. Johnson wrote:
> > Hi,
> > 
> > 
> >>But let's get a clear roadmap down this time, what features are
> >>essential for the next cycle?
> > 
> > 
> > A *much* reduced core size (5 CDs + rescue is getting a bit much) and
> > large reduction in the overall memory overhead. I know FC is fantastic,
> > but given you need 128Mb for a text only version and 512Mb for a desktop
> > environment, it's becoming hard to justify to the powers that be that
> > using Linux over WinXP is that good an option.
> 
> Are you implying that the only reason one would choose Linux over
> Windows XP, is because it has lower memory and system requirements?
No, not "the only reason", but one essential reason:

Linux's small memory footprint and it's free licensing had enabled
people to use Linux on "recycled HW".

Ralf

Re: Wild and crazy times for the development tree

[Bill Nottingham]

Dave Jones (da...@redhat.com) said: 
> I'm not claiming to be representative of the majority of Fedora users,
> but given there are a lot of Fedora users less fortunate than myself,
> I believe that discontinuing CD iso's would severely impact a huge portion
> of our userbase.  It'd be interesting to see the download stats to see
> if they agree with this.

DVD downloads outnumber CD downloads roughly 2:1 on x86, 5:1 on x86_64,
and 6:1 on ppc.

Bill

Re: Wild and crazy times for the development tree

[Dave Jones]

On Mon, Mar 20, 2006 at 11:35:47PM -0500, Bill Nottingham wrote:
 > Dave Jones (da...@redhat.com) said: 
 > > I'm not claiming to be representative of the majority of Fedora users,
 > > but given there are a lot of Fedora users less fortunate than myself,
 > > I believe that discontinuing CD iso's would severely impact a huge portion
 > > of our userbase.  It'd be interesting to see the download stats to see
 > > if they agree with this.
 > 
 > DVD downloads outnumber CD downloads roughly 2:1 on x86, 5:1 on x86_64,
 > and 6:1 on ppc.

Wow. I'm stunned.
I guess I should stop living in the past and get some DVD writables ;)

Dave

-- 
http://www.codemonkey.org.uk

Re: Wild and crazy times for the development tree

[Dimi Paun]

On Mon, 2006-03-20 at 23:39 -0500, Dave Jones wrote:
> Wow. I'm stunned.

Same here. I am in the same position as Dave, and I *thought*
this is rather typical. I would have been willing to put money
that CDs are way more popular than DVDs...

Good thing I didn't! :)

-- 
Dimi Paun 
Lattica, Inc.

Re: Wild and crazy times for the development tree

[Richard Hally]

Bill Nottingham wrote:
Dave Jones (da...@redhat.com) said: 

I'm not claiming to be representative of the majority of Fedora users,
but given there are a lot of Fedora users less fortunate than myself,
I believe that discontinuing CD iso's would severely impact a huge portion
of our userbase.  It'd be interesting to see the download stats to see
if they agree with this.


DVD downloads outnumber CD downloads roughly 2:1 on x86, 5:1 on x86_64,
and 6:1 on ppc.

Bill


Hi Bill,
How or where did you get those number?

Thanks.

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Callum Lerwick wrote:

On Mon, 2006-03-20 at 21:15 +, Paul F. Johnson wrote:


A *much* reduced core size (5 CDs + rescue is getting a bit much)



IMHO its getting time to just forget CDs. Put out DVD isos only. (At the
very least, stop worrying about how many CDs core fills and worry about
fitting on one DVDR...)

Especially with yum backed network installs becoming available. If you
don't want to download and burn an entire DVD iso, just download a
10-50mb network install iso. This is what I do with debian on the rare
occasion I install a system from scratch. Since I only install a system
once every few years, its a huge waste to me to download and burn 5+
disks that I'm only going to use once.

Someone on IRC was arguing that in many parts of the world, bandwidth is
very expensive. I don't see how 5 CDs to download vs a DVD make a
difference here. The real question is, how common are DVD burners, and
most importantly, DVD readers in various parts of the world.


I had an IRC discussion with a nice fellow from Jakarta a number of
months back, who explained how expensive it was to be on the internet.
He had a 56k connection, which cost him the equivalent of $20 USD for
the month.

My first thought was probably quite similar to what many of you are
currently thinking:  "Um, well that's very close to what it costs in
the US for access too, so is not really more expensive."

Except for the fact that this person was a full time computer
programmer, and told me that his monthly salary was the equivalent
of $200 US dollars.  Think about that.  His monthly internet connection
cost him 1/10th of his entire income for the month, just to get a
crappy 56k connection which ties up the phoneline as long as it is
connected.  I don't know if there were per-minute telephone fees
on top of that or not.

For each employed person reading this, take 1/10th of your income.
Would you pay that much money to have a 56k phone line based internet
connection each month?  Even if you did do so, would you want to
download the entire DVD ISO image which takes anywhere from 2-3 weeks
to download?  All this through a phone line which gets very frequently
disconnected?

I told him perhaps he should just purchase Fedora Core on CD instead,
and indicated there were many places online which you could order CDs.
He said it would be about $10, which again is like 5% of his monthly
income.  And that's a twice a year cost.  He said that buying Fedora
CDs locally was more expensive for "free software" than buying bootleg
copies of Windows XP down the street for $2-5 a pop.

That's only one single city in one country in the world.  There are
a great many of places in the world in which costs similar or higher
to this exist, and downloading large images is very prohibitive.

Another consideration worth mentioning is the OLPC project.



I suppose the thing to do is keep an eye on CD downloads vs DVD. Let the
community decide.


Living in North America, and having the pleasure of owning a DVD burner,
and having a huge stack of blank media, and a high speed Internet
connection (cablemodem) which costs me a very small fraction of my
income each month - I, like many North Americans do prefer to download
the DVD ISO images, and have no personal interest in the CD ISO images.
If the CD ISOs were to be dropped completely, it would bear practically
no consequence to me, as I have no use for them at all personally.

However, not having CD ISOs available would create a huge barrier for
many people in the world who are much less fortunate than you or I, who
either do not have high speed Internet, or who simply can't afford it,
or like the fellow I spoke to from Jakarta - who spend a massive
portion of their monthly income just to have the priveledge of using
open source software, because they believe it is the right thing to do.

Open source software (free software if you prefer) should be accessible
to all people, regardless of what country they live in, or their
personal financial circumstances.  Wherever possible it should not only
be free as in liberty, but also be available free of cost, or as close
to free of cost as possible.

So, as long as Fedora can be provided via CD ISO images for a reasonable
manpower cost in doing the work necessary to allow it to continue to
be viable, then it _should_ continue to be available in this manner
IMHO.  I think there are some creative ideas to allow this to be viable
for quite some time, and it seems that the Fedora Project is heading
in the right direction WRT this.

There is no really easy way to gauge how many people download the
CD images versus the DVD images, and equally no easy way to gauge
how many people install the OS from either based off how many
downloads there were.  Add bittorrent to the mix, and various other
download methods, other peer-to-peer networks, etc. and it's just
not possible to gauge.  It is quite clear to me however, that there
is a very significant number of people using CD based

Re: Wild and crazy times for the development tree

[Bill Nottingham]

Richard Hally (rha...@mindspring.com) said: 
> Bill Nottingham wrote:
> >Dave Jones (da...@redhat.com) said: 
> >>I'm not claiming to be representative of the majority of Fedora users,
> >>but given there are a lot of Fedora users less fortunate than myself,
> >>I believe that discontinuing CD iso's would severely impact a huge portion
> >>of our userbase.  It'd be interesting to see the download stats to see
> >>if they agree with this.
> >
> >DVD downloads outnumber CD downloads roughly 2:1 on x86, 5:1 on x86_64,
> >and 6:1 on ppc.
> >
> >Bill
> >
> Hi Bill,
>   How or where did you get those number?

http://torrent.fedoraproject.org:6969/

Bill

Re: Wild and crazy times for the development tree

[Ralf Corsepius]

On Mon, 2006-03-20 at 23:39 -0500, Dave Jones wrote:
> On Mon, Mar 20, 2006 at 11:35:47PM -0500, Bill Nottingham wrote:
>  > Dave Jones (da...@redhat.com) said: 
>  > > I'm not claiming to be representative of the majority of Fedora users,
>  > > but given there are a lot of Fedora users less fortunate than myself,
>  > > I believe that discontinuing CD iso's would severely impact a huge 
> portion
>  > > of our userbase.  It'd be interesting to see the download stats to see
>  > > if they agree with this.
>  > 
>  > DVD downloads outnumber CD downloads roughly 2:1 on x86, 5:1 on x86_64,
>  > and 6:1 on ppc.

Don't trust any statistics you didn't forge yourself ;)

What did you count? Successful complete downloads or accesses?

The DVD image is the first a recursive download would access, so if you
counted accesses you'd erroniously count retries (which are likely to
happen due to the size of the image).

> I guess I should stop living in the past and get some DVD writables ;)
Don't forget about people with 
* low bandwidth/limited download contingents. To many of them, DVD
images aren't a real alternative, because downloading them would block
internet-access for a considerable amount of time.
* machines without DVD-ROMs still require CDs.

Ralf

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Callum Lerwick wrote:

On Mon, 2006-03-20 at 20:23 -0500, Mike A. Harris wrote:


Personally, I think it would be confusing as hell to have a single
src.rpm generate binary rpms of which some are in Fedora Core and
others are in Fedora Extras.  That would be made more confusing
since Fedora has separate bugzilla product/etc. for Core and for
Extras (and for Legacy).  Since bugzilla components themselves are
based from the name of the src.rpm, one would have to file a bug
report from a binary package from extras against Fedora Core for the
src.rpm it was built from.

I can think of other complications and confusion that would ensue
as well.  I'm not sure what to suggest to solve those issues however.


Less separation between core and extras. Like maybe none. (cough Debian
cough)


Hmm, that's actually an interesting point.  ;o)  Perhaps in the
future we could move away from the current model of repository
level separation, to a more responsibility oriented level of
separation or somesuch.  In other words, there are reasons why
things are the way they are now, which led to the solution of
having Core and Extras seprate.  However, that doesn't mean it
is the only way for us to achieve whatever goals are needed.

Perhaps the separation could be rethought out.  I guess the first
step would be to clearly define what the specific reasons are for
Fedora Core being separate from Extras right now.  If we can
define those reasons completely and clearly, they could be
"MUST-HAVE" features used as a basis for a new concept of
putting things together perhaps..

Just open brainstorming...  ;o)

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

Re: Wild and crazy times for the development tree

[Callum Lerwick]

On Mon, 2006-03-20 at 23:11 -0500, Dave Jones wrote:
> DVD writers aren't anywhere near as commonplace as CD writers yet.
> Looking around right now, I have 7 computers near me. CD writers outnumber
> DVD writers 6:1.  (And the majority of the computers with CD writers are
> less than 2 years old (two of them are <6 months old))
> 
> (ironically, the dvd writer is my 3 month old laptop)

Yeah, but you own at least one, and that's the point. Extend this to
"friends with DVD writers". How many people don't have one?

Was Red Hat *ever* available on floppies? Yet we all still used it long
before CD writers became common... (Although RH was being sold for
profit on CD at the time...)


signature.asc
Description: This is a digitally signed message part

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Ralf Corsepius wrote:

On Mon, 2006-03-20 at 19:43 -0500, Mike A. Harris wrote:

Whenever I do need/want 3D enabled support for something in Linux,
I fire up another machine and use it instead, leaving my primary
desktop in a stable DRI-disabled state.

To answer the "when do I want decent 3D performance" question,
I must first answer "When do I want accelerated 3D at all?".



So, what do I use OpenGL for?  More or less for video games,


You are ignoring the fact, Linux has a strong user base in people with a
scientific/engineering/technical background ...


No, I'm not ignoring any facts at all.  My email clearly stated what
_I_ am using, and why, for my _own_ reasons.  I also clearly stated
that other people's situations are likely very different, and that
different people have different needs.  People with a scientific/
engineering/technical background are no exception.  _EVERY_ person's
needs is likely to be different in some way from the next guy.

My comments were not intended to recommend to people they should do what
_I_ decide have decided to do for me.  I clearly stated that also.

I was merely answering the question asked, which was "what do _I_
use OpenGL for".  And yes, I am a scientific/engineering/technical
person.



Now...  if I _was_ actually trying to get a game or some other
accelerated OpenGL software to work in Linux, I would use
whatever the best card I have on hand with OSS driver support
was at the given point in time. 


Whether you like it or not ... reality is different.


The email was about _MY_ chosen reality, not about your reality.
You can not change my own reality, nor my personal requirements or
preferences.  As stated, other people have different needs and
requirements which may be very different from what my own needs
and requirements are.  I did not in any way imply that everyone
should do what _I_ decide to do.  Don't put words in my mouth, or
read more into what I am saying, than what I am actually saying.



People are pragmatically using what they have/can get/are supplied with,
and will ditch the distro or even the OS if it doesn't suite to their
demands.


Yes, some people will indeed do that.  As I said, everyone has their
own individual perogatives.  Mine is different from yours, and yours
is different from the next guy.  There are likely to be "groups" of
people in the same boat, other groups of people in boat #2, boat #3,
and even likely to be overlaps between some of the different boats,
but in no way whatsoever is there a one-size-fits-all solution for
everyone.

I myself have not ditched the distro or even the OS.  I do however
utilize "that other OS" when I feel the need to do so, because Linux
currently does not suit my personal demands.



Fortunately for Fedora, the proprietary drivers have worked
sufficiently well.


That's not even completely true.  It depends on which proprietary driver
it is you are talking about, what specific hardware you have, what
type of displays you have attached, and what the support is like for
your specific desired configuration, what driver features you need/want,
and wether they're supported or not.  The proprietary drivers work well
for some people, and other people can't get them to work at all.
Sometimes they can't get them to work due to their own incompetence, or
simple human error in not following instructions.  Other times it is due
to incompatibilities between the proprietary software and the specific
kernel they're using, specific X server they're using, or the particular
motherboard they have, the BIOS they have, or some other combination
of factors.  I can completely guarantee that there is a whole host of
people out there who have had nothing but endless problems with
proprietary drivers out there from _any_ vendor, many of whom have
sworn to never use proprietary drivers again.  I know, I get to hear
about it all week long every week.

As I said before, there is no one-size-fits-all solution.  Some people
simply refuse to use proprietary Linux drivers for ideological reasons,
or on some other principle.  Others may have tried to use them and been
unable to get them to work, or found their specific hardware isn't
supported.

To each his own.


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

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Ralf Corsepius wrote:

On Mon, 2006-03-20 at 17:29 -0500, Mike A. Harris wrote:


Paul F. Johnson wrote:


Hi,




But let's get a clear roadmap down this time, what features are
essential for the next cycle?



A *much* reduced core size (5 CDs + rescue is getting a bit much) and
large reduction in the overall memory overhead. I know FC is fantastic,
but given you need 128Mb for a text only version and 512Mb for a desktop
environment, it's becoming hard to justify to the powers that be that
using Linux over WinXP is that good an option.


Are you implying that the only reason one would choose Linux over
Windows XP, is because it has lower memory and system requirements?


No, not "the only reason", but one essential reason:

Linux's small memory footprint and it's free licensing had enabled
people to use Linux on "recycled HW".


Absolutely.  A fraction of the entire userbase out there does indeed
choose linux due to it having generally lower hardware requirements than
other OS choices within their consideration.

And to various other users, hardware requirements are meaningless,
they choose their OS based on other factors, such as specific features
unique to the OS, or for ideolological reasons, or lower TCO, or any
of a number of other factors.

Generally lower hardware requirements is not on every single person's
radar when choosing an OS.  If everyone valued lower hardware
requirements higher than anything else, nobody out there would
be running any Microsoft product for years now.



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

Re: Wild and crazy times for the development tree

[David Nielsen]

tir, 21 03 2006 kl. 00:00 -0500, skrev Mike A. Harris:

> I told him perhaps he should just purchase Fedora Core on CD instead,
> and indicated there were many places online which you could order CDs.
> He said it would be about $10, which again is like 5% of his monthly
> income.  And that's a twice a year cost.  He said that buying Fedora
> CDs locally was more expensive for "free software" than buying bootleg
> copies of Windows XP down the street for $2-5 a pop.

Please note that in the spirit of the community and all that, a lot of
us "wealthy" westerners do spend a considerable amount of money shipping
totally free copies of Fedora around the world. If he or anyone wants a
copy but cannot download it, #fedora on irc.freenode.net is a good place
to find friendly people to help and every Linux forum I've ever been to
has had a sticky thread offering free copies of Linux mailed anywhere in
the world.

Getting Linux is not a problem if you utilise the community. It's one of
the reasons Linux is such a great thing to be part of.

- David

Re: Wild and crazy times for the development tree

[Ian Pilcher]

Dave Jones wrote:
> I guess I should stop living in the past and get some DVD writables ;)

Or forget about burning discs altogether.  Assuming that one is
upgrading (or parallel installing) a system that already has Linux
installed, there's no need.

I suppose that the one requirement that some people may not have is a
spare partition that can hold the ISO for a hard disk install.

-- 

Ian Pilcheri.pilc...@comcast.net

Re: Wild and crazy times for the development tree

[Callum Lerwick]

On Tue, 2006-03-21 at 00:10 -0500, Mike A. Harris wrote:
> > Less separation between core and extras. Like maybe none. (cough Debian
> > cough)
> 
> Hmm, that's actually an interesting point.  ;o)  Perhaps in the
> future we could move away from the current model of repository
> level separation, to a more responsibility oriented level of
> separation or somesuch.  In other words, there are reasons why
> things are the way they are now, which led to the solution of
> having Core and Extras seprate.  However, that doesn't mean it
> is the only way for us to achieve whatever goals are needed.
> 
> Perhaps the separation could be rethought out.  I guess the first
> step would be to clearly define what the specific reasons are for
> Fedora Core being separate from Extras right now.  If we can
> define those reasons completely and clearly, they could be
> "MUST-HAVE" features used as a basis for a new concept of
> putting things together perhaps..

Well, what makes Fedora different is its a fast moving, cutting edge
distribution. Fedora is more than happy to make drastic changes in a
short amount of time, stomping over numerous core packages. This is what
makes it different. (especially from Debian...)

Perhaps the core/extras separation is the price we pay to accomplish
this development style. Perhaps eliminating the separation will just
turn Fedora into slow moving Debian.

Or maybe it won't. I really don't know.

> Just open brainstorming...  ;o)

Indeed.


signature.asc
Description: This is a digitally signed message part

Re: Wild and crazy times for the development tree

[Mike A. Harris]

David Nielsen wrote:

tir, 21 03 2006 kl. 00:00 -0500, skrev Mike A. Harris:



I told him perhaps he should just purchase Fedora Core on CD instead,
and indicated there were many places online which you could order CDs.
He said it would be about $10, which again is like 5% of his monthly
income.  And that's a twice a year cost.  He said that buying Fedora
CDs locally was more expensive for "free software" than buying bootleg
copies of Windows XP down the street for $2-5 a pop.



Please note that in the spirit of the community and all that, a lot of
us "wealthy" westerners do spend a considerable amount of money shipping
totally free copies of Fedora around the world. If he or anyone wants a
copy but cannot download it, #fedora on irc.freenode.net is a good place
to find friendly people to help and every Linux forum I've ever been to
has had a sticky thread offering free copies of Linux mailed anywhere in
the world.

Getting Linux is not a problem if you utilise the community. It's one of
the reasons Linux is such a great thing to be part of.


That's a wonderful idea!  I never even thought of that.  The cost of one
of us Westerners shipping a CD/DVD anywhere in the world is probably
about the same cost or close to it of someone buying it online give or
take a few bucks, but the bigger difference is that to me or you it is
the price of a few cups of coffee, whereas it's like making a car or
house payment to people in some parts of the world.


Perhaps a http://www.fedora-for-free.org or
http://www.fedora-philanthropy.org could be set up, in which people in
various parts of the world who can't afford to buy CD/DVDs or to easily
download them to go to get them sent to them for free, and other people
such as ourselves could contribute a few bucks each release to fund the
distribution.


Just an idea.

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

Re: Wild and crazy times for the development tree

[Ian Pilcher]

Ralf Corsepius wrote:
> People are pragmatically using what they have/can get/are supplied with,
> and will ditch the distro or even the OS if it doesn't suite to their
> demands. Fortunately for Fedora, the proprietary drivers have worked
> sufficiently well.

I have this gnawing feeling that ATI and nVIDIA are the twin rocks on
which desktop Linux is going to founder.

-- 

Ian Pilcheri.pilc...@comcast.net

Re: Wild and crazy times for the development tree

[Callum Lerwick]

On Tue, 2006-03-21 at 00:00 -0500, Mike A. Harris wrote:
> I had an IRC discussion with a nice fellow from Jakarta a number of
> months back, who explained how expensive it was to be on the internet.
> He had a 56k connection, which cost him the equivalent of $20 USD for
> the month.
> 
> My first thought was probably quite similar to what many of you are
> currently thinking:  "Um, well that's very close to what it costs in
> the US for access too, so is not really more expensive."
> 
> Except for the fact that this person was a full time computer
> programmer, and told me that his monthly salary was the equivalent
> of $200 US dollars.  Think about that.  His monthly internet connection
> cost him 1/10th of his entire income for the month, just to get a
> crappy 56k connection which ties up the phoneline as long as it is
> connected.  I don't know if there were per-minute telephone fees
> on top of that or not.
> 
> For each employed person reading this, take 1/10th of your income.
> Would you pay that much money to have a 56k phone line based internet
> connection each month?  Even if you did do so, would you want to
> download the entire DVD ISO image which takes anywhere from 2-3 weeks
> to download?  All this through a phone line which gets very frequently
> disconnected?

Its roughly the same amount of data to download a DVD vs 5 CDs. I fail
to see how the "bandwidth is expensive" argument applies.

Download cost is meaningless in the CDs vs DVDs argument. Its the same.

The question is, what does one do with the images once downloaded.

> I told him perhaps he should just purchase Fedora Core on CD instead,
> and indicated there were many places online which you could order CDs.
> He said it would be about $10, which again is like 5% of his monthly
> income.  And that's a twice a year cost.  He said that buying Fedora
> CDs locally was more expensive for "free software" than buying bootleg
> copies of Windows XP down the street for $2-5 a pop.

This is all a great example of what even a poor broke college student
like me takes for granted in the US, and is completely missing the
really useful bits of information:

1) So how *does* this person get Fedora, and keep it updated? Do they
download it at 56k? Do they buy CDs? What?

2) Do they have a DVD reader?


signature.asc
Description: This is a digitally signed message part

Re: Wild and crazy times for the development tree

[Callum Lerwick]

On Mon, 2006-03-20 at 23:37 -0600, Ian Pilcher wrote:
> I have this gnawing feeling that ATI and nVIDIA are the twin rocks on
> which desktop Linux is going to founder.

We're back in the early days of XFree86 all over again.

It'll get better.

So when is someone going to start reverse engineering nVidia hardware?
There's support in Utah-GLX that no one's yet bothered to port over to
DRI, and some guy has written drivers for BeOS, though I've been unable
to find the actual source for the driver. Just an unmodified copy of
Mesa...

http://dri.freedesktop.org/wiki/nVidia?action=highlight&value=CategoryHardware
http://www.haikunews.org/1050
http://web.inter.nl.net/users/be-hold/BeOS/NVdriver/download.html

It seems having well maintained binary drivers is a great deterrent to
the community reverse engineering its own. ATI hardware is actively
being reverse engineered, the Aureal Vortex cards got reverse
engineered, Broadcom and TI ACX1xx wireless has been reverse engineered.
But no nVidia support is being maintained that I know of...


signature.asc
Description: This is a digitally signed message part

Re: Wild and crazy times for the development tree

[Patrick Barnes]

On Monday 20 March 2006 23:39, "Mike A. Harris"  wrote:
> David Nielsen wrote:
> > tir, 21 03 2006 kl. 00:00 -0500, skrev Mike A. Harris:
> >>I told him perhaps he should just purchase Fedora Core on CD instead,
> >>and indicated there were many places online which you could order CDs.
> >>He said it would be about $10, which again is like 5% of his monthly
> >>income.  And that's a twice a year cost.  He said that buying Fedora
> >>CDs locally was more expensive for "free software" than buying bootleg
> >>copies of Windows XP down the street for $2-5 a pop.
> >
> > Please note that in the spirit of the community and all that, a lot of
> > us "wealthy" westerners do spend a considerable amount of money shipping
> > totally free copies of Fedora around the world. If he or anyone wants a
> > copy but cannot download it, #fedora on irc.freenode.net is a good place
> > to find friendly people to help and every Linux forum I've ever been to
> > has had a sticky thread offering free copies of Linux mailed anywhere in
> > the world.
> >
> > Getting Linux is not a problem if you utilise the community. It's one of
> > the reasons Linux is such a great thing to be part of.
>
> That's a wonderful idea!  I never even thought of that.  The cost of one
> of us Westerners shipping a CD/DVD anywhere in the world is probably
> about the same cost or close to it of someone buying it online give or
> take a few bucks, but the bigger difference is that to me or you it is
> the price of a few cups of coffee, whereas it's like making a car or
> house payment to people in some parts of the world.
>
> 
> Perhaps a http://www.fedora-for-free.org or
> http://www.fedora-philanthropy.org could be set up, in which people in
> various parts of the world who can't afford to buy CD/DVDs or to easily
> download them to go to get them sent to them for free, and other people
> such as ourselves could contribute a few bucks each release to fund the
> distribution.
> 
>
> Just an idea.
>

http://fedoraproject.org/wiki/FreeMedia

-- 
Patrick "The N-Man" Barnes
nma...@n-man.com

http://www.n-man.com/

Have I been helpful?  Rate my assistance!
http://rate.affero.net/nman64/
-- 



pgpn00RRguPnd.pgp
Description: PGP signature

Re: Wild and crazy times for the development tree

[Paul P Komkoff Jr]

Replying to Ralf Ertzinger:
> Hi.
> 
> On Mon, 20 Mar 2006 11:59:57 -0500, Jeremy Katz wrote:
> 
> > to grab the fedora-release package from the FC5 release instead.  If
> > you want to keep testing and helping to develop things for Fedora
> > Core 6, expect for some fun to pop up as always.
> 
> Sooo... what are we going to break first?

I vote for init.

-- 
Paul P 'Stingray' Komkoff Jr // http://stingr.net/key <- my pgp key
 This message represents the official view of the voices in my head

Re: Wild and crazy times for the development tree

[David Nielsen]

man, 20 03 2006 kl. 16:28 -0500, skrev Matthew Miller:
> On Mon, Mar 20, 2006 at 09:40:10PM +0100, David Nielsen wrote:
> > I tend to agree, the 9 month cycle worked wonderfully, Fedora Core 5 is
> > by far the best release the Fedora Project has put out yet and as a
> > tester I enjoyed having that extra time to see new fundamental changes
> > take place and get bugs tracked down.
> 
> I was intrigued by your blog suggestion of alternating 9 month devel / 4
> month stabilization cycles.  But I also think
> your own objection (people will not take the 9-month release seriously
> enough) is pretty strong.
> 
> As someone crazy enough to try to use Fedora for real work, the 9 month
> cycle is awesome.

I'd wager that the technology preview releases would actually get a fair
bit of attention if we continued the current agenda where large parts of
software is backported to the stable releases.

To sum up the point, we'd have 6 months of development time followed by
3 months of API stable bugfixing and updating (e.g. like what was done
with GNOME 2.14 in FC5) then we release that as the technology preview
Fedora, it would be considered stable. Then we follow that up with a 4
month API stable polish cycle.

Pros and cons of such an out of balance approach to releasing are many
but when I thought it up I had just had a lenghty conversation with one
of the FC developers who was telling me that with the 9 month cycle he
was fighting fatigue and burnout. It struck me as a user that the 9
month cycle was completely awesome but for developers it might be hard -
thus the need for faster paced releases and periods of light pressure.

We can all agree that major surgery like replacing the Init system,
reworking the installer, switching to modular X, switching GCC versions,
etc. requires more time than a 6 month cycle would allow for in terms of
testing (implementation I'm sure could be done in the 6 month cycle but
would it be well tested?). So we will need more cycles of this length,
depending on the amount of surgery that is planned in the near future. 

We also need to consider that Red Hat needs to make money, they do this
selling RHEL and services related to that product, this means they need
a stable product to base their work off. The polish cycle would, I hope,
serve well as a platform for this kind of work. I'm all for making Red
Hat' life easier, they kindly sponsor a lot of Fedora development that I
get with no strings attached - the least I can do for them is help
hammer Fedora as best I can.

I truly think that the 9/4 cycle could work well for us provided we
could build up an active testing community, it all depends on appealing
to people with QA experience or the will to learn these skills. If we
fail to do this having a polish cycle will not result in a massively
better product in the end - but then again if we don't have proper
testing of any cycle, Fedora will be shit regardless.

Is it time for me to lay down the crackpipe?

- David

Re: Wild and crazy times for the development tree

[Emmanuel Seyman]

On Tue, Mar 21, 2006 at 01:40:19AM -0600, Callum Lerwick wrote:
> 
> So when is someone going to start reverse engineering nVidia hardware?

Stéphane Marchesin has started doing this :
  http://nouveau.sourceforge.net/>

Emmanuel

Re: Wild and crazy times for the development tree

[Mike A. Harris]

Callum Lerwick wrote:


For each employed person reading this, take 1/10th of your income.
Would you pay that much money to have a 56k phone line based internet
connection each month?  Even if you did do so, would you want to
download the entire DVD ISO image which takes anywhere from 2-3 weeks
to download?  All this through a phone line which gets very frequently
disconnected?



Its roughly the same amount of data to download a DVD vs 5 CDs. I fail
to see how the "bandwidth is expensive" argument applies.


Because you don't actually have to download all 5 CDs to do an install,
unless we have recently made it mandatory that all of them are present
or something.  And downloads do not always succeed, resulting in having
to potentially redownload failed download in progress, or a corrupted
download.  Re-downloading one CD takes less time than redownloading
a DVD.


Download cost is meaningless in the CDs vs DVDs argument. Its the same.


Only if you download everything always, and have flawless downloads,
which I'm told is not the case.


The question is, what does one do with the images once downloaded.


Orthagonal issue to actually getting the images downloaded that one
wants in the first place.  It's easier and faster to do an absolute
minimal install, and then perhaps download individual packages
afterward with yum which you'd like to add on top of what you
installed already, than to download the whole DVD or 5 CDs, and then
install it.

Before you suggest doing a network install, think how well that would
work over a 56k or even slower - unreliable dialup line.  I did an
internet install over cable modem twice last week, and I get 200+Kb/s
and it took forever (and then failed due to install time bugs/problems.)



I told him perhaps he should just purchase Fedora Core on CD instead,
and indicated there were many places online which you could order CDs.
He said it would be about $10, which again is like 5% of his monthly
income.  And that's a twice a year cost.  He said that buying Fedora
CDs locally was more expensive for "free software" than buying bootleg
copies of Windows XP down the street for $2-5 a pop.


This is all a great example of what even a poor broke college student
like me takes for granted in the US, and is completely missing the
really useful bits of information:

1) So how *does* this person get Fedora, and keep it updated? Do they
download it at 56k? Do they buy CDs? What?

2) Do they have a DVD reader?


The person in question has downloaded some releases and described it
as a very horrible process.  Having myself previously downloaded Red
Hat Linux 4.2 through 6.2 via modem, the older releases at 14.4k and
RHL 6.0 and higher at around 33.6k, I can say that it took about 1-2
weeks to get everything, with very frequent disconnections and other
hassles, all while completely consuming my single telephone line and
thus preventing me from making or receiving phonecalls.  That really
sucked, so I feel the pain of those who do not have high speed.

Mind you, I was able to at least set up my downloading to be able to
resume from where it left off, but even I have ended up having a
downloaded ISO not match MD5sum and not work properly, to have to
redownload it.

(After that, I joined Red Hat, and was sent 7.0 on CDROM, and 6
 months later we got cable modem Internet access in town so I never
 had to feel the pain again since...  although waiting all day for
 a DVD ISO to download is still painful.  ;o)




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

Re: Wild and crazy times for the development tree

[Naoki]

Totally agree. DVD ISOs are far more convenient and eliminate the need
to argue over how many CDs FC should fit on.
I vote to stop pandering to old technology, CD-ROMs have been kicking
around for twenty years and are clearly not up to todays requirements.

The most compelling argument made so far was that of "If you only have
a 56k modem and internet usage is expensive and time consuming ", but
doesn't it take the same amount of time to download 3,4,(5?) CDs as it
does one DVD?  Callum said it best, if you really can't be downloading
a full DVD ISO then get the netboot iso and only download the exact
RPMs you need. That's even better than CD.


Callum Lerwick wrote:

  On Mon, 2006-03-20 at 21:15 +, Paul F. Johnson wrote:
  
  
A *much* reduced core size (5 CDs + rescue is getting a bit much)

  
  
IMHO its getting time to just forget CDs. Put out DVD isos only. (At the
very least, stop worrying about how many CDs core fills and worry about
fitting on one DVDR...)

Especially with yum backed network installs becoming available. If you
don't want to download and burn an entire DVD iso, just download a
10-50mb network install iso. This is what I do with debian on the rare
occasion I install a system from scratch. Since I only install a system
once every few years, its a huge waste to me to download and burn 5+
disks that I'm only going to use once.

Someone on IRC was arguing that in many parts of the world, bandwidth is
very expensive. I don't see how 5 CDs to download vs a DVD make a
difference here. The real question is, how common are DVD burners, and
most importantly, DVD readers in various parts of the world.

I suppose the thing to do is keep an eye on CD downloads vs DVD. Let the
community decide.
  

Re: Wild and crazy times for the development tree

[Hans Kristian Rosbach]

On Mon, 2006-03-20 at 23:29 -0600, Callum Lerwick wrote:
> On Mon, 2006-03-20 at 23:11 -0500, Dave Jones wrote:
> > DVD writers aren't anywhere near as commonplace as CD writers yet.
> > Looking around right now, I have 7 computers near me. CD writers outnumber
> > DVD writers 6:1.  (And the majority of the computers with CD writers are
> > less than 2 years old (two of them are <6 months old))
> > 
> > (ironically, the dvd writer is my 3 month old laptop)
> 
> Yeah, but you own at least one, and that's the point. Extend this to
> "friends with DVD writers". How many people don't have one?

This does not hold in all cases.

For example take any server hosting company, at least here the cd-roms
outnumber the dvd-roms approx 15:1 at the moment. All new servers gets
dvd-roms, but we don't want to exchange several hundred cd-roms into
dvd-roms. Besides I hardly ever need more than cd1 on those servers
due to the post-install script fetching most useful packages so I only
need to do a minimal install on all servers no matter what purpose they
will have.

And then at home, I have dvd-roms in all my workstations but my mother
(and many others that I help out at times) does not. So i would still
need to carry the cd version as that is universally usable. For that
reason I have never even downloaded a single FC dvd image yet.

My firewall and all development boxes have only cd-roms as most of them
are P2/P3/Xeon 700Mhz, and even my dual Xeon 2.8Ghz has a cd-rom.

-HK

Re: Wild and crazy times for the development tree

[Rudolf Kastl]

2006/3/21, Hans Kristian Rosbach :
> On Mon, 2006-03-20 at 23:29 -0600, Callum Lerwick wrote:
> > On Mon, 2006-03-20 at 23:11 -0500, Dave Jones wrote:
> > > DVD writers aren't anywhere near as commonplace as CD writers yet.
> > > Looking around right now, I have 7 computers near me. CD writers outnumber
> > > DVD writers 6:1.  (And the majority of the computers with CD writers are
> > > less than 2 years old (two of them are <6 months old))
> > >
> > > (ironically, the dvd writer is my 3 month old laptop)
> >
> > Yeah, but you own at least one, and that's the point. Extend this to
> > "friends with DVD writers". How many people don't have one?
>
> This does not hold in all cases.
>
> For example take any server hosting company, at least here the cd-roms
> outnumber the dvd-roms approx 15:1 at the moment. All new servers gets
> dvd-roms, but we don't want to exchange several hundred cd-roms into
> dvd-roms. Besides I hardly ever need more than cd1 on those servers
> due to the post-install script fetching most useful packages so I only
> need to do a minimal install on all servers no matter what purpose they
> will have.
>
> And then at home, I have dvd-roms in all my workstations but my mother
> (and many others that I help out at times) does not. So i would still
> need to carry the cd version as that is universally usable. For that
> reason I have never even downloaded a single FC dvd image yet.
>
> My firewall and all development boxes have only cd-roms as most of them
> are P2/P3/Xeon 700Mhz, and even my dual Xeon 2.8Ghz has a cd-rom.
>
> -HK
>
> --
> fedora-devel-list mailing list
> fedora-devel-list@redhat.com
> https://www.redhat.com/mailman/listinfo/fedora-devel-list
>

do you go to every server in your rack and pop in an install cd/dvd? :)

you serious?

regards,
rudolf kastl

Re: Wild and crazy times for the development tree

[sean]

On Tue, 21 Mar 2006 05:19:49 +0100
Ralf Corsepius  wrote:

> > > What do you guys do when you want decent 3D performance?
> 
> Use the proprietary drivers ... :-)

There are few cases where resorting to proprietary drivers is required.
There are open source drivers that provide good-enough 3d for the needs 
of many Linux users.

> You are ignoring the fact, Linux has a strong user base in people with a
> scientific/engineering/technical background ...

Engineers may have needs which can't be met by open source 3d drivers 
today.   Not sure who you're grouping into the scientific and technical 
categories though, i have a technical background and my needs are met 
perfectly well by open source 3d drivers.

> Whether you like it or not ... reality is different.

You should speak for yourself instead of imagining you have 
a better grasp of reality than everyone else :oP

> People are pragmatically using what they have/can get/are supplied with,
> and will ditch the distro or even the OS if it doesn't suite to their
> demands. Fortunately for Fedora, the proprietary drivers have worked
> sufficiently well.

Many people have been misinformed on this matter by others who are fixated 
on the latest-and-greatest graphics speed.  Personally I think it's time for 
a more rational discussion about the capabilities and performance levels 
actually needed by most people.

Anyone who values the flexibility and power offered by open source solutions
would do well to consider closely whether their 3d graphics needs honestly
require the use of proprietary drivers.

Cheers,
Sean

Re: Wild and crazy times for the development tree

[Hans Kristian Rosbach]

On Tue, 2006-03-21 at 12:50 +0100, Rudolf Kastl wrote:

> do you go to every server in your rack and pop in an install cd/dvd? :)
> 
> you serious?

Since we install about 1-3 servers per week, yes.
(And I don't have to do any of the carrying)

1. Server arrives at workshop desk
2. Hardware is installed/upgraded/dusted off
3. Bioses/firmwares updates
4. Memtest86+ overnight
5. OS Install (type depending on customers or our needs)
6. Server placed into rack

This lets us spot failing hardware such as capacitors or
hear noisy disks/fans. And with the KVM extender to my
office I can administrate the install remotely when I come
into my office after the overnight memtest anyways.

It might possibly be a bit more work than booting off PXE
but then we would have the added administration of PXE image
servers on each physical net. And just generally a cdrom is
a whole lot easier to debug if something fails. 

Installing minimal images over PXE would of course save
more time, but I would still have to make and maintain
all of them.

(Another benefit is that the firewall in that room is set
 up such that windows servers don't get viruses on them
 before windows update has run.)

-HK

Re: Wild and crazy times for the development tree

[Ralf Corsepius]

On Tue, 2006-03-21 at 07:06 -0500, sean wrote:
> On Tue, 21 Mar 2006 05:19:49 +0100
> Ralf Corsepius  wrote:
> 
> > > > What do you guys do when you want decent 3D performance?
> > 
> > Use the proprietary drivers ... :-)
> 
> There are few cases where resorting to proprietary drivers is required.
> There are open source drivers that provide good-enough 3d for the needs 
> of many Linux users.
Wishful thinking - Try finding a notebook without an ATI or Nvidia
graphics card ...

> > You are ignoring the fact, Linux has a strong user base in people with a
> > scientific/engineering/technical background ...
> 
> Engineers may have needs which can't be met by open source 3d drivers 
> today.   Not sure who you're grouping into the scientific and technical 
> categories though, i have a technical background and my needs are met 
> perfectly well by open source 3d drivers.
Mine are not - I am working on 3d simulations/animations/visualizations.
For my applications, the nvidia driver outperforms the nv driver by ca.
factor 10. On top of that, for the hardware I have available, the nv
driver had been non-functional on one machine before FC4.

> > Whether you like it or not ... reality is different.
> 
> You should speak for yourself instead of imagining you have 
> a better grasp of reality than everyone else :oP
ROTFL ... 

> > People are pragmatically using what they have/can get/are supplied with,
> > and will ditch the distro or even the OS if it doesn't suite to their
> > demands. Fortunately for Fedora, the proprietary drivers have worked
> > sufficiently well.
> 
> Many people have been misinformed on this matter by others who are fixated 
> on the latest-and-greatest graphics speed. 
I am not talking about squeezing the "max" out of the latest and
greatest graphics HW, I am talking about:
- Getting 3D/GL functional at all.
- Getting a reasonable 3D/GL performance.
- Getting access to advanced GL features.
all on moderate to old graphic NVidia cards.

>  Personally I think it's time for 
> a more rational discussion about the capabilities and performance levels 
> actually needed by most people.
If you want to get a feeling about what I am talking about, try
SceneViewer (From Inventor, in FE) with one of the models from the Large
Geometry Repository, or try the Coin Examples from sim.no (Not in FE).

Ralf

Re: Wild and crazy times for the development tree

[Rudolf Kastl]

2006/3/21, Hans Kristian Rosbach :
> On Tue, 2006-03-21 at 12:50 +0100, Rudolf Kastl wrote:
>
> > do you go to every server in your rack and pop in an install cd/dvd? :)
> >
> > you serious?
>
> Since we install about 1-3 servers per week, yes.
> (And I don't have to do any of the carrying)
>
> 1. Server arrives at workshop desk
> 2. Hardware is installed/upgraded/dusted off
> 3. Bioses/firmwares updates
> 4. Memtest86+ overnight
> 5. OS Install (type depending on customers or our needs)
> 6. Server placed into rack
>
> This lets us spot failing hardware such as capacitors or
> hear noisy disks/fans. And with the KVM extender to my
> office I can administrate the install remotely when I come
> into my office after the overnight memtest anyways.
>
> It might possibly be a bit more work than booting off PXE
> but then we would have the added administration of PXE image
> servers on each physical net. And just generally a cdrom is
> a whole lot easier to debug if something fails.
>
> Installing minimal images over PXE would of course save
> more time, but I would still have to make and maintain
> all of them.
>
> (Another benefit is that the firewall in that room is set
>  up such that windows servers don't get viruses on them
>  before windows update has run.)
>
> -HK
>
> --
> fedora-devel-list mailing list
> fedora-devel-list@redhat.com
> https://www.redhat.com/mailman/listinfo/fedora-devel-list
>

yup i was thinking of an install server. there are various approaches
as to how to do deal with it of course. well actually if you think
your present approach is best regarding your use case i wont hold you
back.

personally id go for the above mentioned pxe solution and push a hd
install then.

regards,
Rudolf Kastl

Re: Wild and crazy times for the development tree

[mbneto]

My wish list:
- break the packages in order to reduce dependencies so we can have a
lower footprint for a minimum install (and have a more granular
control of what is installed - security issues)
- optimize the memory usage so we can continue to use FC6,7 with
todays specs (even if reduced functionality).

>
> Sooo... what are we going to break first?
>

Re: Wild and crazy times for the development tree

[Jeff Spaleta]

On 3/20/06, Bill Nottingham  wrote:
> DVD downloads outnumber CD downloads roughly 2:1 on x86, 5:1 on x86_64,
> and 6:1 on ppc.

Is that for fc5?

I'd think the same statistic over the lifetime of a release would be
more interesting.  I'd argue the people who participate during the
release surge are a skewed sample of the overall population.

-jef

Re: Wild and crazy times for the development tree

[Nicu Buculei]

Jeff Spaleta wrote:

On 3/20/06, Bill Nottingham  wrote:

DVD downloads outnumber CD downloads roughly 2:1 on x86, 5:1 on x86_64,
and 6:1 on ppc.


Is that for fc5?

I'd think the same statistic over the lifetime of a release would be
more interesting.  I'd argue the people who participate during the
release surge are a skewed sample of the overall population.


Not only that, people who download at any time are a skewed sample of 
the overall population. A lot of people obtain their install media from 
other sources: the CD-Writer of a friend, a magazine, etc.


Also, the 2:1 ratio can be interpreted the other way: at least 33% users 
prefer CDs for installation, going DVD only would leave them in the cold.


--
nicu
my hats collection: http://fedora.nicubunu.ro/hats/
Open Clip Art Library: http://www.openclipart.org
my Fedora stuff: http://fedora.nicubunu.ro

Re: Wild and crazy times for the development tree

[Alan Cox]

On Mon, Mar 20, 2006 at 05:27:26PM -0600, Ian Pilcher wrote:
> Mike A. Harris wrote:
> > All proprietary drivers?  ;o)
> 
> I can't help wondering...
> What do you guys do when you want decent 3D performance?

I walk out of the front door, the resolution is excellent, the shadows are
superbly computed and you get exercise too 8)

On the more serious side there is a lack of DRI capability for very high
end gaming. R2xx will do all that I need however because highly detailed
simulations of killing people really don't appeal.

Alan

Re: Wild and crazy times for the development tree

[Alan Cox]

On Tue, Mar 21, 2006 at 12:02:03AM -0500, Bill Nottingham wrote:
> > Hi Bill,
> > How or where did you get those number?
> 
> http://torrent.fedoraproject.org:6969/

So its torrent stats only - thats fairly biased if so

Re: Wild and crazy times for the development tree

[Adam Jackson]

On Mar 20, 2006, at 6:27 PM, Ian Pilcher wrote:


Mike A. Harris wrote:


All proprietary drivers?  ;o)



I can't help wondering...

What do you guys do when you want decent 3D performance?


I write a working 3D driver.

- ajax

Re: Wild and crazy times for the development tree

[Josh Boyer]

Jeff Spaleta wrote:
> On 3/20/06, Bill Nottingham  wrote:
>> DVD downloads outnumber CD downloads roughly 2:1 on x86, 5:1 on x86_64,
>> and 6:1 on ppc.
>
> Is that for fc5?
>
> I'd think the same statistic over the lifetime of a release would be
> more interesting.  I'd argue the people who participate during the
> release surge are a skewed sample of the overall population.

That, and the numbers quoted are only for the torrent downloads.  Not
everyone uses the torrent.  However, it's probably the easiest to collect
sample set.  Tracking it on the mirrors would suck ;)

josh

Re: Wild and crazy times for the development tree

[Jeremy Katz]

On Tue, 2006-03-21 at 09:14 +0100, David Nielsen wrote:
> To sum up the point, we'd have 6 months of development time followed by
> 3 months of API stable bugfixing and updating (e.g. like what was done
> with GNOME 2.14 in FC5) then we release that as the technology preview
> Fedora, it would be considered stable. Then we follow that up with a 4
> month API stable polish cycle.

And then we have a release with basically extremely out of date software
for the next 9 months.  This ends up being pretty difficult to maintain
over the longer term.

> Pros and cons of such an out of balance approach to releasing are many
> but when I thought it up I had just had a lenghty conversation with one
> of the FC developers who was telling me that with the 9 month cycle he
> was fighting fatigue and burnout. It struck me as a user that the 9
> month cycle was completely awesome but for developers it might be hard -
> thus the need for faster paced releases and periods of light pressure.

Yep.  For all of the pressure that the 6 month cycle entails (and it
does), it definitely seemed less so than 9.  Mostly because the end is
more clearly in sight the entire time :)

> We can all agree that major surgery like replacing the Init system,
> reworking the installer, switching to modular X, switching GCC versions,
> etc. requires more time than a 6 month cycle would allow for in terms of
> testing (implementation I'm sure could be done in the 6 month cycle but
> would it be well tested?). So we will need more cycles of this length,
> depending on the amount of surgery that is planned in the near future. 

Like I've said, I don't think that a nine month cycle actually does
anything to significantly help here.  The problem is that then, if I
have something else pop up, it's easier to have it interrupt working on
whatever big new thing.  And it still ends up getting done at the last
minute.  I'm convinced this is a fundamental law of software
development :-)

Jeremy

Re: Wild and crazy times for the development tree

[Prarit Bhargava]

  And it still ends up getting done at the last

minute.  I'm convinced this is a fundamental law of software
development :-)



Parkinson's Law

http://www.wildenforcers.com/Images/boogaard/Boogard-Skate.jpg

P.

Re: Wild and crazy times for the development tree

[Prarit Bhargava]

Prarit Bhargava wrote:

  And it still ends up getting done at the last


minute.  I'm convinced this is a fundamental law of software
development :-)




Parkinson's Law

http://www.wildenforcers.com/Images/boogaard/Boogard-Skate.jpg



Oops.  Wrong link

http://www.bartleby.com/59/3/workexpandst.html

P.


P.

Re: Wild and crazy times for the development tree

[Stephen J. Smoogen]

On 3/21/06, Jeremy Katz  wrote:

> Like I've said, I don't think that a nine month cycle actually does
> anything to significantly help here.  The problem is that then, if I
> have something else pop up, it's easier to have it interrupt working on
> whatever big new thing.  And it still ends up getting done at the last
> minute.  I'm convinced this is a fundamental law of software
> development :-)
>

Well it is a fundamental law of life. The gain in 9 months here was
that I as a tester could let real life deal me its usual blows and I
could still do some testing and finding of stuff before we crashed.  I
did not have that luxury with FC2 and basically dropped doing anything
with it.. and only was able to pick up FC3 at around its release time.
I didnt look at FC4 until it came out. I will probably have to blow
off FC6 on a sixth month schedule and do some work with FC7.  I might
be a statistical outlayer.. but it seems to be a constant
question/reply from other people using Fedora here in Government land
(or  us guys over 35 land with kids land.)


> Jeremy
>
> --
> fedora-devel-list mailing list
> fedora-devel-list@redhat.com
> https://www.redhat.com/mailman/listinfo/fedora-devel-list
>


--
Stephen J Smoogen.
CSIRT/Linux System Administrator

Re: Wild and crazy times for the development tree

[Bill Nottingham]

Alan Cox (a...@redhat.com) said: 
> On Tue, Mar 21, 2006 at 12:02:03AM -0500, Bill Nottingham wrote:
> > > Hi Bill,
> > >   How or where did you get those number?
> > 
> > http://torrent.fedoraproject.org:6969/
> 
> So its torrent stats only - thats fairly biased if so

Yes, but it's what's immediately available. 5:1 for x86_64 is still
fairly significant.

Bill

Re: Wild and crazy times for the development tree

[Jonathan Dieter]

I'm currently living in a country where broadband is expensive, and, 
even if you have the money to pay for broadband, upload speeds are below 
64kbps.  Because bittorrent rewards uploading, I've found that 
downloading through bittorrent takes roughly double the time it takes me 
to download directly from the mirrors (and when my maximum download 
speed is 256kbps, it means the difference between a little over a day 
for FC5 and a little over two days).


I would suspect most people with slow upload links have this problem 
(though I could be completely wrong).  If that is the case, then the 
torrent statistics will be heavily biased towards users with reasonably 
fast upload links, and the users with slower links (who, like myself, 
are probably downloading the CD isos for reasons mentioned by Mike 
Harris in another posting) aren't accurately represented.


Bear in mind that this is all based on assumptions made on my 
experiences which may be completely incorrect.


Jonathan

Bill Nottingham said:

Alan Cox (a...@redhat.com) said:
> > On Tue, Mar 21, 2006 at 12:02:03AM -0500, Bill Nottingham wrote:
>>> > > > Hi Bill,
>>> > > > How or where did you get those number?
>> > >
>> > > http://torrent.fedoraproject.org:6969/
> >
> > So its torrent stats only - thats fairly biased if so

Yes, but it's what's immediately available. 5:1 for x86_64 is still
fairly significant.

Bill

Re: Wild and crazy times for the development tree

[Pedro Fernandes Macedo]

Callum Lerwick wrote:

Someone on IRC was arguing that in many parts of the world, bandwidth is
very expensive. I don't see how 5 CDs to download vs a DVD make a
difference here. The real question is, how common are DVD burners, and
most importantly, DVD readers in various parts of the world.

  
Please ignore that comment about bandwith.. I was a bit sleepy at the 
time and my brain was working in a weird way.
The issue really is DVD reader/burner availability as mentioned several 
times on this thread. Checking just two places directly connected to 
internet and technology that I have/had access to, I'd have to say that 
the number of DVD readers and burners is still very small, even in a 
major city here in Brazil (I live in the 3rd biggest city of the 
country). In my workplace (a company that specializes in tech support 
for other companies), we dont have a single DVD burner, we have two DVD 
readers that and all the other machines have CDs. In the university 
where I just finished my CS course, I know for sure of the existence of 
about 10 DVD burners in desktop computers/servers owned by the 
department, while most computers have a CD reader.


Another big issue is the availability of reliable DVD-/+R discs and 
their price. While in the US you can get a 50 pack for a relatively 
small price, here a 50 pack from a good brand is about 50 reais (about 
US$25,00. Just as a reference, the minimum wage here is R$300,00 or 
US$150,00) and you only can find those on ebay. You can also find some 
cheap discs for R$2,00 each, but then you better pray that it lasts 
enough time for you to remove the disk from the burner and put in the 
reader of the machine you want to install


As for the statistics cited by Bill Nottingham, I'm certain they're very 
favorable to DVD isos, specially if you consider that in the release 
notes and in the release announcement it's strongly suggested that 
bittorrent be used to download the DVD images since a few broken 
ftp/http clients and servers may not be able to handle such big files, 
while bittorrent works perfectly.


--
Pedro Macedo

Re: Wild and crazy times for the development tree

[John DeDourek]

Ian Pilcher wrote:


Ralf Corsepius wrote:
 


People are pragmatically using what they have/can get/are supplied with,
and will ditch the distro or even the OS if it doesn't suite to their
demands. Fortunately for Fedora, the proprietary drivers have worked
sufficiently well.
   



I have this gnawing feeling that ATI and nVIDIA are the twin rocks on
which desktop Linux is going to founder.

 


Or maybe low cost wireless cards that don't use firmware.