[parisc-linux] Version 0.9 of PA-RISC Linux Released

[parisc-linux] Version 0.9 of PA-RISC Linux Released

[Matt Taggart]

Version 0.9 of PA-RISC Linux Released

On May 31, 2001 the PA-RISC Linux development community released version
0.9, the latest version of Linux for computers using Hewlett Packard's
PA-RISC processor. This release is the result of several years of work by
developers in the Free Software community including developers from The
Debian Project, Hewlett Packard, and Linuxcare.

- Latest Debian Installer 
- Support for 32 and 64 bit systems 
- Over 4000 Debian packages 
- Full development environment 
- Support for large memory (up to 16GB)
- Ability to use Debian's apt tool to update to the latest
    package versions

The 0.9 release install CDROM images and related documentation can be found at

http://parisc-linux.org/release-0.9/

Information about PA-RISC Linux is available at

http://parisc-linux.org/

Any questions or comments about the release should be directed to the
mailing list at

parisc-linux@parisc-linux.org

-- 
Matt Taggart        Linux Development Lab
taggart@fc.hp.com   HP Linux Systems Operation

Re: [parisc-linux] Version 0.9 of PA-RISC Linux Released

[R P Herrold]

On Thu, 31 May 2001, Matt Taggart wrote:

> The 0.9 release install CDROM images and related documentation can be found at
>
> http://parisc-linux.org/release-0.9/

Congratulations ...

website hint -- stolen from the FSF -- list the SIZE of the
images, to permit potential downloaders decide if it is worth
it.

hint 2:  add md5sums of the images -- both as gzipped, and of
the ISO for validation before burning, in the body of the
webpage -- must be external to the CD readme, for it would
change the values ...

-- Russ

Re: [parisc-linux] Version 0.9 of PA-RISC Linux Released

[Matthieu Delahaye]

> hint 2:  add md5sums of the images -- both as gzipped, and of
> the ISO for validation before burning, in the body of the
> webpage -- must be external to the CD readme, for it would
> change the values ...

Until the parisc-linux.org webpage is updated, you can find the
md5sums (of the gzipped format only) to the download page at our site:
http://mkhppa1.esiee.fr/en/download.html
Don't assume our md5sums are linked to the ISOs available at
parisc-linux.
We made ourself the compression (and the compression rate isn't the
same).
So they only correspond to the ISOs you can download from our site.

I'll update this afternoon (GMT+2) our site in order to be in
conformance
with the official site, and so provide ISO with exactly the same gziped
files. 
Each ISOs are 130Mo average.
-- 
Matthieu Delahaye
ESIEE Team
http://mkhppa1.esiee.fr/

Re: [parisc-linux] Version 0.9 of PA-RISC Linux Released

[Alan Cox]

> > http://parisc-linux.org/release-0.9/
> 
> Congratulations ...

Ditto.

Im slightly perturbed that random disk corruption under load isnt in release
errata notes as you are shipping a 2.4.0 kernel.

> website hint -- stolen from the FSF -- list the SIZE of the
> images, to permit potential downloaders decide if it is worth
> it.

Its a nice layout other than that. It doesnt render on some browsers however
because you have wrongly nested tables, an extra </table> and a <div> in the
wrong place.

But at the point we are debating the website I suspect it means there are a
shortage of useful things to pick holes in ;)

Re: [parisc-linux] Version 0.9 of PA-RISC Linux Released

[Andrew Shugg]

Congratulations, team, on making the release!

I had a quick look the files which I found here:

  http://puffin.external.hp.com/pub/parisc/cd-images/

A couple of questions arose:

- what's the 142MB "gg" file dated "01-Jun-2001 10:23"?  I mean, I know
  who gg is (hi grant) but as this file is of around the same vintage as
  all the 0.9 compressed ISOs I thought I'd see if it was related

- are all the 0.9 ISOs identical except for which kernel is booted by
  default?  eg does the 0.9-32bit-STI CD image also contain the kernels
  for the 32bit-serial and 64bit-serial installs?

Oh, hang on, I see my second question is already answered by the footer
of the above HTTP directory listing, but I'll leave it in for the
archives I guess.  Answer is:


  You can change one 0.9 ISO into another by replacing the first 512
  bytes.  For example, to change the 32serial into the 64serial, do

  dd bs=512 conv=notrunc if=palinux-0.9-64serial.iso.sector0 \
  of=palinux-0.9-32serial.iso

  You probably want to rename the file too...


I see the same information is provided in a more brief form at the
bottom of the 0.9 README.  Oh well, pardon my ramblings.  ;)

Andrew.

--
Andrew Shugg <andrew@neep.com.au>                   http://www.neep.com.au/

"Just remember, Mr Fawlty, there's always someone worse off than yourself."
"Is there?  Well I'd like to meet him.  I could do with a good laugh."

Re: [parisc-linux] Version 0.9 of PA-RISC Linux Released

[Paul Bame]

= A couple of questions arose:
= 
= - what's the 142MB "gg" file dated "01-Jun-2001 10:23"?  I mean, I know
=   who gg is (hi grant) but as this file is of around the same vintage as
=   all the 0.9 compressed ISOs I thought I'd see if it was related

It was a temporary file I used while correcting the STI ISO.

= - are all the 0.9 ISOs identical except for which kernel is booted by
=   default?  eg does the 0.9-32bit-STI CD image also contain the kernels
=   for the 32bit-serial and 64bit-serial installs?

Yes.  the palinux-0.9-README (also in the ISOs) gives a clue how to use
the CD for other purposes.  There is a README in the cd-images directory
which tells how to turn one bootable image into another bootable image
by downloading only 512 bytes.

Re: [parisc-linux] Version 0.9 of PA-RISC Linux Released

[Grant Grundler]

Andrew Shugg wrote:
> - what's the 142MB "gg" file dated "01-Jun-2001 10:23"?  I mean, I know
>   who gg is (hi grant) but as this file is of around the same vintage as
>   all the 0.9 compressed ISOs I thought I'd see if it was related

Not mine.
-rw-r--r--   1 bame     cvsusers 148707328 Jun  1 10:23 gg

grant

Grant Grundler
parisc-linux {PCI|IOMMU|SMP} hacker
+1.408.447.7253

[parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Michael S.Zick]

Hi Folks -
Sub-Title this one: Adventures in running V0.9 on a HP9000/720 box:

1) I had been working with V0.6 and CVS snapshot of 05/25/2001 when this =
notice arrived.
Great - Gotta get that! -- 18 hours later, got it.  (very slow modem conn=
ection to a backwoods ISP).
2) Now make some working room - throw out (almost) all of that old stuff.
3) Mount new ISO image - What, no new ISO image! *&^%$#$%=20

V0.9 ISO - Test one: it deletes just fine.
Now repeat step 1).

xx hours later - it's here (I will not delete this copy, I know that work=
s).

Now, will it run "out of the box" - well, out of the things in the ISO an=
yway?

V0.9 ISO - Test two: Nope, not on a Model 720

Ok, so will it build "out of the box" and then run?

V0.9 ISO - Test three: Nope - the /devel directory is gone.

4) Turn loose ncFTP daemon clinet loose on cvs - get current snapshots,=20
being optimistic - get the new (hppa) native stuff also.

5) Meanwhile - what things of possible use can I say, well there are some=
 similarities
between the "boxed" 32STI vmlinux and what I have found to date:

File: linux/arch/parisc/kernel/setup.c
Function: void __init setup_console....
Approx. lines 332-335
=09printk ->
=09arg[0] -> free_memory start
=09arg[1] -> passed command line (from palo/ipl/ipl.c - Funciton: static =
void join
=09arg[2] -> initrd start
=09arg[3] -> initrd end
The values shown for arg[0] and arg[1] are the same - does this indicate =
a problem?
The passed command line (after copy from arg[1]) prints correctly.

File: linux/drivers/char/console.c
Function: void __init con_init
Approx. lines 2455-2457
=09call to register_console appears to never return -=20
The 9000/720 (at least mine) has a SGC display reported by the inventory =
list.
Changing the passed command line (with IPL editor) to: console=3DttyB
(which specifies the PDC console functions) seems to by-pass the above co=
de
and things proceed until this kernel gets into the same troubles that I w=
as working on:

I had been trouble shooting a number of different things in the LASI-SEM7=
00 scsi
driver for the 53c7xx (yes, the 9000/720 has the oldest controller availa=
ble).

The driver was reporting "unable to obtain consistant DMA memory" - among=
 other
strange and mysterious things.
Only recently did I notice that the ethernet driver is reporting the same=
 thing -

The printer port driver is reporting (interrupt driven mode) - is the pri=
nter port on
the 720 DMA capable?

So my next step will be to quit looking only at the SCSI driver and start=
 looking=20
for something common to the same problem reports from two different drive=
rs.

Note: All work is being done XC(x86) - the 720 disk was wiped clean befor=
e I got the
machine.  So I have to get it running before I can trouble shoot why it i=
sn't running=20
with the native mode toolset.

Yes - I am using the most recent XC-... things available from your site.

Any suggestions?
HELP !?!

PS: Is any work being done at the work station end?  Such as an 80960 X-c=
lient?  This
system came with two workstations - a HP 700/RX (model C2708A) and a HP
Envizex, aSeries (model C2731A).

PPS: A note on used equipment - I know this system was working (under HP-=
UX)
until it was shut down, the day before I bought it from a second-hand sto=
re.

Mike


On Thu, 31 May 2001, Matt Taggart wrote:
> Version 0.9 of PA-RISC Linux Released
>=20
> On May 31, 2001 the PA-RISC Linux development community released versio=
n
> 0.9, the latest version of Linux for computers using Hewlett Packard's
> PA-RISC processor. This release is the result of several years of work =
by
> developers in the Free Software community including developers from The
> Debian Project, Hewlett Packard, and Linuxcare.
>=20
> - Latest Debian Installer=20
> - Support for 32 and 64 bit systems=20
> - Over 4000 Debian packages=20
> - Full development environment=20
> - Support for large memory (up to 16GB)
> - Ability to use Debian's apt tool to update to the latest
>     package versions
>=20
> The 0.9 release install CDROM images and related documentation can be f=
ound at
>=20
> http://parisc-linux.org/release-0.9/
>=20
> Information about PA-RISC Linux is available at
>=20
> http://parisc-linux.org/
>=20
> Any questions or comments about the release should be directed to the
> mailing list at
>=20
> parisc-linux@parisc-linux.org
>=20
> --=20
> Matt Taggart        Linux Development Lab
> taggart@fc.hp.com   HP Linux Systems Operation
>=20
>=20
>=20
>=20
>=20
> -- =20
> To UNSUBSCRIBE, email to debian-hppa-request@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debi=
an.org

[parisc-linux] PA-RISC 0.9 release problem

[Alan Cox]

I hit one fundamental problem...

The CD dumps you into a pseudo gui interface which is not compatible with
the web console HP29xx emulation...

Alan

[parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Jayadeep Purushothaman]

Hi,

Is there a way someone can ship a CD for me? I am in India, so trying to
download the ISO images will take way too much bandwidth.

Thanks,
Jayadeep

Jayadeep Purushothaman,
Hewlett Packard (ISO),
29, Cunningham Road
Bangalore - 52

RE: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[MEYER,ALAN C. (HP-FtCollins,ex1)]

We plan to create a bunch of CDs for purposes like this.  We want to wait a
week or so, just to make sure the current ISOs  don't have some major
problems to be fixed.  Can you wait a bit?

Alan
 
> Is there a way someone can ship a CD for me? I am in India, 
> so trying to
> download the ISO images will take way too much bandwidth.
> 
> Thanks,
> Jayadeep
> 
> Jayadeep Purushothaman,
> Hewlett Packard (ISO),
> 29, Cunningham Road
> Bangalore - 52
> 
> 
> 
> _______________________________________________
> parisc-linux mailing list
> parisc-linux@lists.parisc-linux.org
> http://lists.parisc-linux.org/cgi-bin/mailman/listinfo/parisc-linux
> 

[parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Michael S.Zick]

Matthew -

Great idea - both sites should CLEARLY cross-link their information pages=
=2E

My own suggestions:
1) - leave the /devel directory on the CD image until at least
version 12.0.0 of pa-risc.linux.

I am using the kernel snap-shot of 06/03/2001 - in the hopes
that it is close to what was used to build the kernels in V0.9 -
but that is only a guess on my part

2) - don't drop a built kernel into an image of even a pre-release
CD until it compiles error (and warning) free.

Some of the problems I have found and fixed were clearly
mentioned by the compiler.

3) - Note to others trying fix "Old Box" support -

Please use a command line such as:
date >0603_something_cmnd.txt ; \
date >0603_something_errs.txt ; \
make vmlinux (or whatever) >>0603_something_cmnd.txt 2>>0603_something_er=
rs.txt

Then tar-zip the linux/.config, ...cmnd.txt, and the ...errs.txt -
attach to e-mail directed to Mike <mszick@pfash.com>

The more paths through this code taken by the compiler, the faster=20
I should be able to find (and maybe fix) the "Old Box" support.

4) - Note to folks posting the daily kernel snapshots -
4.a) - Please make the "daily snapshots" contain only changed
files from the prior "daily snapshot".
4.b) - Please make a "complete" kernel snapshot only once
a week.
4.c) - Please include the three files mentioned above in
each and every snapshot.

Being able to diff those three files is a big aid in making
progress on these problems.

Mike

On Tue, 05 Jun 2001, Andrew Shugg wrote:
> Matthew Wilcox said:
> > Well, 720 is explicitly mentioned as not yet being supported.
>=20
> I'm not sure who is looking after the debian-hppa web page at
> /ports/hppa/ (glad to see it is being kept up to date though!) but
> perhaps whomever it is could point to the parisc-linux.org hardware
> compatibility page.
>=20
>   http://parisc-linux.org/release-0.9/systems.html
>   http://parisc-linux.org/hardware/supported.html
>=20
> Or perhaps its information could be paraphrased on the debian-hppa page
> for first-time visitors?  eg "for the curious amongst you, the followin=
g
> systems work: xxx, these maybe work yyy, and these definitely don't zzz=
"
>=20
> Andrew.
>=20
> --
> Andrew Shugg <andrew@neep.com.au>                   http://www.neep.com=
=2Eau/
>=20
> "Just remember, Mr Fawlty, there's always someone worse off than yourse=
lf."
> "Is there?  Well I'd like to meet him.  I could do with a good laugh."
>=20
>=20
> -- =20
> To UNSUBSCRIBE, email to debian-hppa-request@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debi=
an.org

Re: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Matthew Wilcox]

On Tue, Jun 05, 2001 at 07:24:31AM -0500, Michael S.Zick wrote:
> Great idea - both sites should CLEARLY cross-link their information pages.

if you have constructive suggestions for the website, please feel free to
post them.  I didn't like the original layout, and I think the redesign
made it harder to find information.  but i don't know how to make it
better, nor do i have time.

> I am using the kernel snap-shot of 06/03/2001 - in the hopes
> that it is close to what was used to build the kernels in V0.9 -
> but that is only a guess on my part

this is why i introduced the EXTRAVERSION.  it's -pa10, and that's still
head of the CVS tree.

> 2) - don't drop a built kernel into an image of even a pre-release
> CD until it compiles error (and warning) free.
> 
> Some of the problems I have found and fixed were clearly
> mentioned by the compiler.

let's see your patch then.

-- 
Revolutions do not require corporate support.

Re: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Michael S.Zick]

On Tue, 05 Jun 2001, Matthew Wilcox wrote:
>=20
> let's see your patch then.
>=20
> --=20
> Revolutions do not require corporate support.

Corporations shouldn't require users to drive them
to do their best.

Read on, gang, read on - when I get a clean
compile and a kernel that boots on my model 720...

I'll ask someone where to send the diff's...

Until then (maybe never) I'll continue to post my
progress notes that seem to directly apply to
questions being asked by others on the mailing lists.

Mike

Re: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Matthew Wilcox]

On Tue, Jun 05, 2001 at 02:16:05PM -0500, Michael S. Zick wrote:
> Corporations shouldn't require users to drive them
> to do their best.

So far i've heard nothing but talk from you.  i've been working on this
project for more than 2 years now, working for 3 different companies
during that time.  Sometimes I've been paid to work on it, sometimes
I haven't.  Sometimes I've worked on it entirely on my own time.  I want
to see something constructive from you before I'll accept your criticism.

> Read on, gang, read on - when I get a clean
> compile and a kernel that boots on my model 720...
> 
> I'll ask someone where to send the diff's...

Not the way it works.  Things will move on and you'll get left behind.
Send patches now and they'll (probably) get integrated.

-- 
Revolutions do not require corporate support.

Re: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Michael S.Zick]

On Tue, 05 Jun 2001, Matthew Wilcox wrote:

> I've been working on this
> project for more than 2 years now, working for 3 different companies
> during that time. =20

I'll be glad to match my 4 decades of experience against your 2 years any=
time.
=20
> Not the way it works.=20
> Send patches now and they'll (probably) get integrated.

Perhaps that is not how it works in your world, but consider -

The only way to validate a change as a potential patch is:

1) start with a clean (error and warning free) build that
boots and runs on the machine of interest.

2) hack, change, modify, whatever

3) when done, make another clean build that boots and runs
on the machine of interest.

4) evaluate the effects of the change(s).

5) If successful - then you MIGHT have something worth
submitting as a POTENTIAL patch.

Exactly the same 5 steps your developers take before they
commit a change to CVS.

> So far i've heard nothing but talk from you. =20

That's right - until it is a clean build, boots and runs -
everything is talk and speculation.

>  Things will move on and you'll get left behind.

So what?

The only changes I make outside of the parisc specific
branch of the source tree are to bring the code in line
with 2.4.4 (stable).

The changes I make inside of the parisc specific branch
of the tree will be (someday) found by your own developers.

Just a matter of do you want to duplicate the time and=20
trouble of finding them yourself, or read about them in
the mail archives.

If you are serious about my suggestions being counter-productive,
I will gladly spend the time I am taking to write them for other purposes=
=2E

Mike
Mike

Re: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Matthew Wilcox]

On Tue, Jun 05, 2001 at 03:22:20PM -0500, Michael S. Zick wrote:
> The only way to validate a change as a potential patch is:
> 
> 1) start with a clean (error and warning free) build that
> boots and runs on the machine of interest.
> 
> 2) hack, change, modify, whatever
> 
> 3) when done, make another clean build that boots and runs
> on the machine of interest.
> 
> 4) evaluate the effects of the change(s).
> 
> 5) If successful - then you MIGHT have something worth
> submitting as a POTENTIAL patch.
> 
> Exactly the same 5 steps your developers take before they
> commit a change to CVS.

the people with CVS access are hardly `my' developers.  and the process
you describe is only one form of development.  it's not applicable in a lot
of cases.

> That's right - until it is a clean build, boots and runs -
> everything is talk and speculation.

No.  if you post a patch which moves things in the right direction,
that's more than talk.  it's something which can be tested.

> So what?
> 
> The only changes I make outside of the parisc specific
> branch of the source tree are to bring the code in line
> with 2.4.4 (stable).

erm... if you're doing that, I'll be very surprised if you build a
working system.  Changes are needed to the generic VM which need to be
forward-ported and changed in a few places.  This is work which I don't
feel capable of doing.

> The changes I make inside of the parisc specific branch
> of the tree will be (someday) found by your own developers.

That's great, I love people who do work, keep their changes secret and
then expect everyone else to find them because they're `obvious'.

> Just a matter of do you want to duplicate the time and 
> trouble of finding them yourself, or read about them in
> the mail archives.

I'd love you to post them.  I said so earlier.

> If you are serious about my suggestions being counter-productive,
> I will gladly spend the time I am taking to write them for other purposes.

If all you're going to do is patch your own tree and write snide flames
to this mailing list, then yes, I'd rather you worked on a different
project.  If you're going to work cooperatively, then please hang around.

-- 
Revolutions do not require corporate support.

Re: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Alan Cox]

> On Tue, Jun 05, 2001 at 03:22:20PM -0500, Michael S. Zick wrote:
> > The only way to validate a change as a potential patch is:
> > 1) start with a clean (error and warning free) build that
> > boots and runs on the machine of interest.
> > 2) hack, change, modify, whatever
> > 3) when done, make another clean build that boots and runs
> > on the machine of interest.
> > 4) evaluate the effects of the change(s).
> > 5) If successful - then you MIGHT have something worth
> > submitting as a POTENTIAL patch.
> > Exactly the same 5 steps your developers take before they
> > commit a change to CVS.
> 
> the people with CVS access are hardly `my' developers.  and the process
> you describe is only one form of development.  it's not applicable in a lot
> of cases.

A lot of the Linux development occurs on much more of a 'directed chaos'
approach, and that's an approach gaining favour in software engineering 
circles too. If it looks roughly right you throw it in or run it past someone
with a good eye for code in that area and throw it in. If as is normally the
case the patch is fine it works very efficiently. When the patch turns out to
be wrong it is rare that there are multiple patches likely to cause the reported
symptoms.

Also you can run continual automated testing using tools like cerberus. Yes
my Software Engineering lecturer would never have approved - but it works.

Alan

Re: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Andrew Shugg]

Michael S.Zick said:
> I'll be glad to match my 4 decades of experience against your 2 years
> anytime.

Willy didn't say anything about his experience, but the time that that
he has put into this particular project.  Perhaps you meant to say you'd
like to put your 40 years of experience together with his two years of
work in order to help the project?  Sadly, that's probably stretching
your words well beyond the "benefit of the doubt".  =(

> If you are serious about my suggestions being counter-productive, I
> will gladly spend the time I am taking to write them for other purposes.

I don't know - or honestly care - who you are or which horse you rode in
on.  You're the first person I've seen to bring this attitude in to this
project and in my position as list-lurker and linux-hppa fiddler I'd be
perfectly happy if you were the last.

Andrew.

--
Andrew Shugg <andrew@neep.com.au>                   http://www.neep.com.au/

"Just remember, Mr Fawlty, there's always someone worse off than yourself."
"Is there?  Well I'd like to meet him.  I could do with a good laugh."

Re: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Michael S.Zick]

Fine by me.=20
I don't have to be told to goa away more than twice.

Mike

On Tue, 05 Jun 2001, Andrew Shugg wrote:
> Michael S.Zick said:
> > I'll be glad to match my 4 decades of experience against your 2 years
> > anytime.
>=20
> Willy didn't say anything about his experience, but the time that that
> he has put into this particular project.  Perhaps you meant to say you'=
d
> like to put your 40 years of experience together with his two years of
> work in order to help the project?  Sadly, that's probably stretching
> your words well beyond the "benefit of the doubt".  =3D(
>=20
> > If you are serious about my suggestions being counter-productive, I
> > will gladly spend the time I am taking to write them for other purpos=
es.
>=20
> I don't know - or honestly care - who you are or which horse you rode i=
n
> on.  You're the first person I've seen to bring this attitude in to thi=
s
> project and in my position as list-lurker and linux-hppa fiddler I'd be
> perfectly happy if you were the last.
>=20
> Andrew.
>=20
> --
> Andrew Shugg <andrew@neep.com.au>                   http://www.neep.com=
=2Eau/
>=20
> "Just remember, Mr Fawlty, there's always someone worse off than yourse=
lf."
> "Is there?  Well I'd like to meet him.  I could do with a good laugh."
>=20
>=20
> _______________________________________________
> parisc-linux mailing list
> parisc-linux@lists.parisc-linux.org
> http://lists.parisc-linux.org/cgi-bin/mailman/listinfo/parisc-linux

Re: [parisc-linux] Re: Version 0.9 of PA-RISC Linux Released

[Andrew Shugg]

Michael S.Zick said:
> Fine by me. 
> I don't have to be told to goa away more than twice.
> 
> Mike

Ah, but no-one actually has told you to go away.  Look back on your four
decades of experience or your even longer life and reflect on all you've
learned in that time about how pride and professionalism are rarely seen
together.  (Again, pardon my presumption, but I assume you did learn!)

Andrew.

--
Andrew Shugg <andrew@neep.com.au>                   http://www.neep.com.au/

"Just remember, Mr Fawlty, there's always someone worse off than yourself."
"Is there?  Well I'd like to meet him.  I could do with a good laugh."