Re: successful installation with 2.3.5 boot-floppies

[Anthony Towns]

On Thu, Jun 14, 2001 at 12:19:47AM +1000, Anthony Towns wrote:
> (An additional problem is that the auric's recent problems (not master's,
> this is a whole new set from the last 24 hours or so) have screwed up both
> testing and the pool in general to an unknown degree. Depending on the state
> of your mirror, this may or may not be a problem. More news as it comes to
> hand.)

Should be better as of the next dinstall run. (seven hours away or so)

Cheers,
aj


-- 
Anthony Towns <[EMAIL PROTECTED]> 
I don't speak for anyone save myself. GPG signed mail preferred.

``_Any_ increase in interface difficulty, in exchange for a benefit you
  do not understand, cannot perceive, or don't care about, is too much.''
  -- John S. Novak, III (The Humblest Man on the Net)


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: successful installation with 2.3.5 boot-floppies

[Adam Di Carlo]

Erik Andersen <[EMAIL PROTECTED]> writes:

> On Tue Jun 12, 2001 at 08:37:21PM -0400, Adam Di Carlo wrote:
> > 
> > > they're allowed to have special instructions that you have to carry out
> > > very precisely ("Switch to VC 2, type this confusing sed command... Before
> > > rebooting, switch to VC 2, and chmod these directories like so..."),
> > > all they have to do is be usable to install a functional woody system.
> > 
> > Yah -- the bad bug still outstanding is the busybox tar problem, which
> > is causing a *VERY* insecure system to be produced.
> 
> This is not an outstanding bug.  It was fixed in busybox 1:0.51-7 which 
> was uploaded on June 4th.  busybox 1:0.51-8 was installed into the archive
> yesterday...

Yes, confirmed fixed.

-- 
.Adam Di [EMAIL PROTECTED]http://www.onshored.com/>


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: successful installation with 2.3.5 boot-floppies

[Anthony Towns]

On Tue, Jun 12, 2001 at 08:37:21PM -0400, Adam Di Carlo wrote:
> Anthony, I sense some impatience and/or hostility here.  What is it
> we can do to make you more happy?  

(For reference, most of the hostility is primarily due to non-Debian
related computer problems; the best thing you can do is either ignore it,
or kick me when I misapply my irritation. Further said non-Debian related
computer problems mean I haven't had time to look at much b-f's related
stuff recently. Next step is to see if we can get some CD images available
somewhere, along with any instructions that people might need to do installs
based on them. CD installs will be horrible atm: there's the b-f alpha-ness,
combined with probable problems with apt-cdrom, and who knows what else.
Still, at this point, that's fine, we're just trying to demonstrate that we
can do installs right now.)

(An additional problem is that the auric's recent problems (not master's,
this is a whole new set from the last 24 hours or so) have screwed up both
testing and the pool in general to an unknown degree. Depending on the state
of your mirror, this may or may not be a problem. More news as it comes to
hand.)

Cheers,
(aj)

-- 
Anthony Towns <[EMAIL PROTECTED]> 
I don't speak for anyone save myself. GPG signed mail preferred.

``_Any_ increase in interface difficulty, in exchange for a benefit you
  do not understand, cannot perceive, or don't care about, is too much.''
  -- John S. Novak, III (The Humblest Man on the Net)

 PGP signature

Re: successful installation with 2.3.5 boot-floppies

[Dale Scheetz]

Before everyone gets too bent out of shape, let me re-assure you all that
the debian-testing group is here to help, and now that we have releases to
work on, I am confident that the team will begin testing.

I've been pretty busy trying to find a job (finding work is always harder
than doing work...) but I should have some time to put toward the testing
of this latest release in the next day or two.

First, I want to thank Adam Di Carlo for his dedication to this job.
Without his efforts it is not clear where we would be with the boot
floppies, and I think we owe him a greatful Thank You.

To the debian-testing list/team I can only say, Go to:

  http://people.debian.org/~aph/debian/dists/woody/main/

and download and test as much of the boot floppies as possible.

I'll be testing the CD boot process, as that is not as complex as building
a complete CD. For those who can also toast a CD, here is the process:

1. Create an ARCHIVES subdirectory where the components to toast will be
   placed.

2. CD to ARCHIVES and create the subdirectory boot/

3. Copy the rescue image file into the ARCHIVES/boot/ directory.

4. Copy a rescue image file, the Drivers tarball, and the Base tarball
   into the ARCHIVES root directory for ease of installation.

5. Make the ISO image file using the command:

 mkisofs -r -T -v -o boot/ \
  -c boot/boot.catalog \
  -o TARGET/binary.iso \
  ARCHIVES

6. Toast TARGET/binary.iso onto CD media

7. Test.

I posted a sample report form on the debian-testing list a while back
which got no comment. I can only assume that it was satisfactory ;-)

Please use that form when reporting test results. Feel free to CC your
report to the boot floppies mailing list if there is information of
interest for these folks. I will try to make periodic composite reports to
that group, but the quicker they get the info the sooner it can be acted
upon.

If anyone in the testing group has any further questions, or needs another
copy of the sample testing report, feel free to contact me.

Waiting is,

On 12 Jun 2001, Adam Di Carlo wrote:

> Anthony Towns <[EMAIL PROTECTED]> writes:
> 
> > On Fri, Jun 08, 2001 at 12:04:49PM -0600, Matt Kraai wrote:
> > I was under the impression the boot-floppy disks were being based on
> > packages in sid (busybox, debootstrap, etc) and used to install packages
> > from woody (dpkg.deb, bash.deb, libc6.deb, etc). This still seems the
> > sensible thing to do, to me, but stuff I read seems to indicate the
> > opposite is happening?
> 
> No, if you're reading that, who-ever wrote it is wrong.
> 
> What we do is build using sid packages, as you say, until base
> freezes, and we let the testers choose which distro (stable, testing,
> unstable) they want to install.
> 
> With regards to telling the testers what to do, I feel that's the
> responsibility of the testing coordinator (is Dale Scheetz the
> coordinator?).  
> 
> If there is no testing coordinator, then that is going to be a problem
> that someone shoudl do something about.
> 
> 
> > The main goal for boot-floppies at the moment, btw, is much simpler
> > than what you appear to be thinking: what we need right now are some
> > boot-floppies in the archive accompanied by a list of do's, don't's and
> > workarounds that can be used by competent testers to do woody
> > installs.
> 
> AFAIK, unless I hear otherwise from testers, we are at that point the
> second that 2.3.5 is moved from incoming into woody (or sid I guess).
> 
> > These only need to be available for i386,
> 
> They are there for i386 and powerpc at least.  BenC said he'd build
> sparc shortly.  m68k might have to wait for 2.3.6, they seem to have a
> build problem.
> 
> > they're allowed to only work for
> > one method ("You can't use these floppies for CD installs, nor for
> > DHCP"),
> 
> Not many such limitations but I don't have any woody CDs yet to test
> with.
> 
> > they're allowed to have special instructions that you have to carry out
> > very precisely ("Switch to VC 2, type this confusing sed command... Before
> > rebooting, switch to VC 2, and chmod these directories like so..."),
> > all they have to do is be usable to install a functional woody system.
> 
> Yah -- the bad bug still outstanding is the busybox tar problem, which
> is causing a *VERY* insecure system to be produced.
> 
> > It'd be nice to start having "released" beta boot-floppies, asap. Even
> > without multiple architecture support, and whatever else.
> 
> We've had "released" versions since Sun,  8 Apr 2001 03:59:31 -0400.
> They just weren't working very well.
> 
> Anthony, I sense some impatience and/or hostility here.  What is it
> we can do to make you more happy?  
> 
> I'm trying to release the code more quickly -- 6 uploads since early
> april.  We are averaging about 10 days between releases  Do you
> have some expectations beyond taht which are not being met?
> 
> I could always use more qualified hack

Re: successful installation with 2.3.5 boot-floppies

[Erik Andersen]

On Tue Jun 12, 2001 at 08:37:21PM -0400, Adam Di Carlo wrote:
> 
> > they're allowed to have special instructions that you have to carry out
> > very precisely ("Switch to VC 2, type this confusing sed command... Before
> > rebooting, switch to VC 2, and chmod these directories like so..."),
> > all they have to do is be usable to install a functional woody system.
> 
> Yah -- the bad bug still outstanding is the busybox tar problem, which
> is causing a *VERY* insecure system to be produced.

This is not an outstanding bug.  It was fixed in busybox 1:0.51-7 which 
was uploaded on June 4th.  busybox 1:0.51-8 was installed into the archive
yesterday...

 -Erik

--
Erik B. Andersen   email:  [EMAIL PROTECTED], [EMAIL PROTECTED]
--This message was written using 73% post-consumer electrons--


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: successful installation with 2.3.5 boot-floppies

[Adam Di Carlo]

Anthony Towns <[EMAIL PROTECTED]> writes:

> On Fri, Jun 08, 2001 at 12:04:49PM -0600, Matt Kraai wrote:
> I was under the impression the boot-floppy disks were being based on
> packages in sid (busybox, debootstrap, etc) and used to install packages
> from woody (dpkg.deb, bash.deb, libc6.deb, etc). This still seems the
> sensible thing to do, to me, but stuff I read seems to indicate the
> opposite is happening?

No, if you're reading that, who-ever wrote it is wrong.

What we do is build using sid packages, as you say, until base
freezes, and we let the testers choose which distro (stable, testing,
unstable) they want to install.

With regards to telling the testers what to do, I feel that's the
responsibility of the testing coordinator (is Dale Scheetz the
coordinator?).  

If there is no testing coordinator, then that is going to be a problem
that someone shoudl do something about.


> The main goal for boot-floppies at the moment, btw, is much simpler
> than what you appear to be thinking: what we need right now are some
> boot-floppies in the archive accompanied by a list of do's, don't's and
> workarounds that can be used by competent testers to do woody
> installs.

AFAIK, unless I hear otherwise from testers, we are at that point the
second that 2.3.5 is moved from incoming into woody (or sid I guess).

> These only need to be available for i386,

They are there for i386 and powerpc at least.  BenC said he'd build
sparc shortly.  m68k might have to wait for 2.3.6, they seem to have a
build problem.

> they're allowed to only work for
> one method ("You can't use these floppies for CD installs, nor for
> DHCP"),

Not many such limitations but I don't have any woody CDs yet to test
with.

> they're allowed to have special instructions that you have to carry out
> very precisely ("Switch to VC 2, type this confusing sed command... Before
> rebooting, switch to VC 2, and chmod these directories like so..."),
> all they have to do is be usable to install a functional woody system.

Yah -- the bad bug still outstanding is the busybox tar problem, which
is causing a *VERY* insecure system to be produced.

> It'd be nice to start having "released" beta boot-floppies, asap. Even
> without multiple architecture support, and whatever else.

We've had "released" versions since Sun,  8 Apr 2001 03:59:31 -0400.
They just weren't working very well.

Anthony, I sense some impatience and/or hostility here.  What is it
we can do to make you more happy?  

I'm trying to release the code more quickly -- 6 uploads since early
april.  We are averaging about 10 days between releases  Do you
have some expectations beyond taht which are not being met?

I could always use more qualified hackers of course

Personally I would put a little concern in about the coordination of
the testing effort.  Is that in place?  I have no idea, maybe...

-- 
.Adam Di [EMAIL PROTECTED]http://www.onshored.com/>


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: successful installation with 2.3.5 boot-floppies

[David Whedon]

> I was under the impression the boot-floppy disks were being based on
> packages in sid (busybox, debootstrap, etc) and used to install packages
> from woody (dpkg.deb, bash.deb, libc6.deb, etc). This still seems the
> sensible thing to do, to me, but stuff I read seems to indicate the
> opposite is happening?

FWIW I agree with aj as to what should be happenning.  I just commited the
follwing:
--- extract_base.c  2001/06/08 17:39:16 1.40
+++ extract_base.c  2001/06/09 23:10:35
@@ -248,6 +248,8 @@
   {_("archived"), _("'slink' - Debian 2.1 Released June 1999"),
CHOICE_SLINK },
   };
 
+
+  problemBox(_("Boot-floppies testers: please install the 'testing'
distrubution (woody).  Finding issues with installing that distribution is more
important than finding issues with sid or potato, thanks."),  _("Important
Note."));
 #endif
 
   /* Sanity Check */

> The main goal for boot-floppies at the moment, btw, is much simpler
> than what you appear to be thinking: what we need right now are some
> boot-floppies in the archive accompanied by a list of do's, don't's and
> workarounds that can be used by competent testers to do woody installs.
> These only need to be available for i386, they're allowed to only work for
> one method ("You can't use these floppies for CD installs, nor for DHCP"),
> they're allowed to have special instructions that you have to carry out
> very precisely ("Switch to VC 2, type this confusing sed command... Before
> rebooting, switch to VC 2, and chmod these directories like so..."),
> all they have to do is be usable to install a functional woody system.
> 
> It'd be nice to start having "released" beta boot-floppies, asap. Even
> without multiple architecture support, and whatever else.

k, With that in mind I think we should be able to release 2.3.5 on Monday 
at the latest.  I'm going to do some i386 installs now and see what it looks
like.


David


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: successful installation with 2.3.5 boot-floppies

[Anthony Towns]

On Fri, Jun 08, 2001 at 12:04:49PM -0600, Matt Kraai wrote:
> Yep.  If you are interested in the other known issues, you might
> want to look at my faq-o-matic entry[1] for the boot-floppies.  It
> lists the status of all the issues that I know about.

Some notes on the FAQ-o-matic questions:

* dhcp-client is in base in debootstrap 0.1.11 in sid

* gcc-3.0-base is only needed on some currently unreleased
  architectures, in particular hppa, iirc.

* groff-base is sid only, and you're more liable to get effective
  installs if you just focus on installing woody, which won't
  have it's dependencies changed too frequently

I was under the impression the boot-floppy disks were being based on
packages in sid (busybox, debootstrap, etc) and used to install packages
from woody (dpkg.deb, bash.deb, libc6.deb, etc). This still seems the
sensible thing to do, to me, but stuff I read seems to indicate the
opposite is happening?

The main goal for boot-floppies at the moment, btw, is much simpler
than what you appear to be thinking: what we need right now are some
boot-floppies in the archive accompanied by a list of do's, don't's and
workarounds that can be used by competent testers to do woody installs.
These only need to be available for i386, they're allowed to only work for
one method ("You can't use these floppies for CD installs, nor for DHCP"),
they're allowed to have special instructions that you have to carry out
very precisely ("Switch to VC 2, type this confusing sed command... Before
rebooting, switch to VC 2, and chmod these directories like so..."),
all they have to do is be usable to install a functional woody system.

It'd be nice to start having "released" beta boot-floppies, asap. Even
without multiple architecture support, and whatever else.

Cheers,
aj

-- 
Anthony Towns <[EMAIL PROTECTED]> 
I don't speak for anyone save myself. GPG signed mail preferred.

``_Any_ increase in interface difficulty, in exchange for a benefit you
  do not understand, cannot perceive, or don't care about, is too much.''
  -- John S. Novak, III (The Humblest Man on the Net)

 PGP signature

Re: successful installation with 2.3.5 boot-floppies

[Matt Kraai]

On Fri, Jun 08, 2001 at 10:23:50AM -0700, David Whedon wrote:
> Fri, Jun 08, 2001 at 10:26:57AM -0600 wrote:
> > To make this work, I had to make two minor changes to debootstrap.
> > After booting, I switched over to the shell on tty2 and opened
> > /usr/lib/debootstrap/scripts/sid with nano-tiny.  Then, in the
> > definition of base, I changed groff to groff-base and added
> > dhcp-client.
> 
> cool, I see bugs are already filed and ajt plans on getting a new debootstrap
> out this weekend.

Yep.  If you are interested in the other known issues, you might
want to look at my faq-o-matic entry[1] for the boot-floppies.  It
lists the status of all the issues that I know about.

Matt

1. http://opensource.lineo.com/~kraai/debian/fom-serve/cache/6.html


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: successful installation with 2.3.5 boot-floppies

[David Whedon]

Fri, Jun 08, 2001 at 10:26:57AM -0600 wrote:
> Howdy,
> 
> I'm happy to say that I am writing this email on a system on
> which I just installed sid using the 2.3.5 boot-floppies.  And
> there was much rejoicing.
That is great!

> To make this work, I had to make two minor changes to debootstrap.
> After booting, I switched over to the shell on tty2 and opened
> /usr/lib/debootstrap/scripts/sid with nano-tiny.  Then, in the
> definition of base, I changed groff to groff-base and added
> dhcp-client.

cool, I see bugs are already filed and ajt plans on getting a new debootstrap
out this weekend.

David


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

successful installation with 2.3.5 boot-floppies

[Matt Kraai]

Howdy,

I'm happy to say that I am writing this email on a system on
which I just installed sid using the 2.3.5 boot-floppies.  And
there was much rejoicing.

To make this work, I had to make two minor changes to debootstrap.
After booting, I switched over to the shell on tty2 and opened
/usr/lib/debootstrap/scripts/sid with nano-tiny.  Then, in the
definition of base, I changed groff to groff-base and added
dhcp-client.

This was done using the i386 compact disk set.

Matt


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]