Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Slackware12-CD not bootable - problems to create bootdisk

4 views
Skip to first unread message

KarinV

unread,
Sep 14, 2007, 4:38:40 AM9/14/07
to
Hello,
I received the 6 CDs from Slackware12, but installation-CD isn't
bootable. There is no 'bare.i', 'install.1' and 'install.2' to find
for this version.
If I boot with boot 10- and root 10-disk or Slackware10-CD, I can
start installation, but it breaks with the package 'glibc' and says:
'FATAL: you need to be running a 2.6.x Kernel in order to upgrade to
this version of glibc.'
But I didn't found any possibility to create special boot- and
rootdisk for slackware12.
What can I do?
I hope, anybody can help me!

Greetings from germany.

Karin

christian

unread,
Sep 14, 2007, 6:17:31 AM9/14/07
to
KarinV wrote:
> I received the 6 CDs from Slackware12, but installation-CD isn't
> bootable. [...]

this might help:

extra/btmgr-3.7/btmgr-3.7_1-i386-1.tgz

Douglas Mayne

unread,
Sep 14, 2007, 8:37:20 AM9/14/07
to
On Fri, 14 Sep 2007 01:38:40 -0700, KarinV wrote:

> Hello,
> I received the 6 CDs from Slackware12, but installation-CD isn't
> bootable. There is no 'bare.i', 'install.1' and 'install.2' to find
> for this version.

<snip>


> I hope, anybody can help me!
>
> Greetings from germany.
>
> Karin
>

If your CD (hardware) is bootable, but you are looking for "the old
standard" kernel, bare.i, then you'lll need to read this:
http://tinyurl.com/36v958

That file (above) is displayed when F2 is pressed at boot. It states
that setup will use either the huge.s or hugesmp.s kernel; there's no
bare.i.

If your CD (hardware) is not bootable, then there are other methods to get
started. The smartbootmgr is one tool, as already suggested on this
thread. There are instructions for using USB or PXE in this directory:
http://tinyurl.com/2vs3q4

--
Douglas Mayne

--
Douglas Mayne

KarinV

unread,
Sep 14, 2007, 11:53:08 AM9/14/07
to

Douglas Mayne schrieb:

> If your CD (hardware) is not bootable, then there are other methods to get
> started. The smartbootmgr is one tool, as already suggested on this
> thread. There are instructions for using USB or PXE in this directory:
> http://tinyurl.com/2vs3q4

My CD-Harddrive is bootable, but the slackware12-CD not.
I tryed to make a bootfloppydisk by smatbootmgr. Something was written
on floppydisk, but it doesn't work.

Karin

Michael Black

unread,
Sep 14, 2007, 12:22:03 PM9/14/07
to
There is definitely a "problem" about booting more recent SLackwares.
I used the same drive as I'd used for Slack 10, and I coudln't boot
11 or 12. In both cases, making a smart boot manager as described
on the DVD enabled me to boot the DVD.

This "problem" has been described here in the past, and it amounts
to a change in how the DVD was put together after Slack 10.

The floppy disk you tried might be bad, I seem to notice that quite a bit
(though oddly, going through decade old Mac and MSDOS floppies recently,
they all seemed in fine shape). I do note that the instructions on
the DVD said to cat the image to the floppy device, and when that didn't
work, I used dd and the floppy did work. I can't recall whether I
used a different floppy for the second try or not.

Michael


Manuel Otto

unread,
Sep 15, 2007, 6:40:27 AM9/15/07
to
On 14 Sep 2007 16:22:03 GMT, et...@FreeNet.Carleton.CA (Michael Black)
wrote:

>KarinV (k.vo...@gmx.net) writes:
>> Douglas Mayne schrieb:
>>
>>> If your CD (hardware) is not bootable, then there are other methods to get
>>> started. The smartbootmgr is one tool, as already suggested on this
>>> thread. There are instructions for using USB or PXE in this directory:
>>> http://tinyurl.com/2vs3q4
>>
>> My CD-Harddrive is bootable, but the slackware12-CD not.
>> I tryed to make a bootfloppydisk by smatbootmgr. Something was written
>> on floppydisk, but it doesn't work.
>>
>> Karin
>>
>There is definitely a "problem" about booting more recent SLackwares.
>I used the same drive as I'd used for Slack 10, and I coudln't boot
>11 or 12. In both cases, making a smart boot manager as described
>on the DVD enabled me to boot the DVD.
>
>This "problem" has been described here in the past, and it amounts
>to a change in how the DVD was put together after Slack 10.

About this problem, you may also want to read this for an explanation:

http://groups.google.com/group/alt.os.linux.slackware/browse_thread/thread/94d077fa91131e1f/614ecfb11bb93080?lnk=st

>The floppy disk you tried might be bad, I seem to notice that quite a bit
>(though oddly, going through decade old Mac and MSDOS floppies recently,
>they all seemed in fine shape). I do note that the instructions on
>the DVD said to cat the image to the floppy device, and when that didn't
>work, I used dd and the floppy did work. I can't recall whether I
>used a different floppy for the second try or not.
>
> Michael
>

Manuel

Douglas Mayne

unread,
Sep 15, 2007, 10:10:12 AM9/15/07
to
On Sat, 15 Sep 2007 12:40:27 +0200, Manuel Otto wrote:
<snip>

>>There is definitely a "problem" about booting more recent SLackwares.
<snip>

>
> About this problem, you may also want to read this for an explanation:
>
> http://groups.google.com/group/alt.os.linux.slackware/browse_thread/thread/94d077fa91131e1f/614ecfb11bb93080?lnk=st
<snip>
>
>>
> Manuel
>
Thanks for that. I did skip reading a lot of the threads on this group
which document boot problems.

Come to think of it, I may have been hit by that bug myself. This was on
a bootable CD using the grub loader, not isolinux, too.
http://groups.google.com/group/alt.os.linux.slackware/msg/b86c62f7ed2e65a4

It's probably a miracle that the BIOS functions are as compatible as they
are across the entire PC landscape. Floppy discs are not obsolete quite
yet. They usually work as advertised (except when the media is bad, and
that happens a lot :( )

--
Douglas Mayne

Manuel Otto

unread,
Sep 15, 2007, 12:55:16 PM9/15/07
to
On Sat, 15 Sep 2007 08:10:12 -0600, Douglas Mayne <do...@sl12.localnet>
wrote:

>On Sat, 15 Sep 2007 12:40:27 +0200, Manuel Otto wrote:
><snip>
>>>There is definitely a "problem" about booting more recent SLackwares.
><snip>
>>
>> About this problem, you may also want to read this for an explanation:
>>
>> http://groups.google.com/group/alt.os.linux.slackware/browse_thread/thread/94d077fa91131e1f/614ecfb11bb93080?lnk=st
><snip>
>>
>>>
>> Manuel
>>
>Thanks for that. I did skip reading a lot of the threads on this group
>which document boot problems.

You're welcome, always a pleasure to be able to help. I also not read
anything (to much reading, and often to complicated for a newbie).

I didn't read the thread 'Unable to boot Slackware 12 disc' at first,
cause I was still using Slack 10 then.

Then problem occured, did a Google group search, and found the thread
again (in the group I daily read).

>Come to think of it, I may have been hit by that bug myself. This was on
>a bootable CD using the grub loader, not isolinux, too.
>http://groups.google.com/group/alt.os.linux.slackware/msg/b86c62f7ed2e65a4

Well, if it uses the same '-boot-load-size 32' option as the Slack
11/12 cd/dvd iso's, and you cann't boot those. It has nothing to do
with lilo or grub, as far as I understand.

>It's probably a miracle that the BIOS functions are as compatible as they
>are across the entire PC landscape. Floppy discs are not obsolete quite
>yet. They usually work as advertised (except when the media is bad, and
>that happens a lot :( )

Keep backups of (real important) floppy's as images:

dd if=/dev/fd0 of=/path_to_image_file/floppy_image.img bs=512

The backups will never go bad :-)

P.S. I still wonder wy Slackware has to obey the rule that
-boot-load-size should be 32 that strictly, considering that it also
works with a lower value, but with the value 32 it will not boot on
some (not even that much) older machienes.

Also, Slackware is really the only bootable cd/dvd where I had this
problem with, and I've tried quite a few live cd's the last 2 years
before Slackware became my OS of choice.

Few weeks ago, out of curiousity, I tried a Debian live cd, and it did
boot fine on this 'old' (2001) machiene. So the Debian cd for sure did
not use '-boot-load-size 32'.

It seems nobody does, except Slackware...

Can this not be changed in future please, if it serves no purpos at
all?

Manuel

KarinV

unread,
Sep 18, 2007, 5:52:35 PM9/18/07
to
Thanks all for your hints.
Today I wrote a bootfloppy with sbootmgr with the help of a DOS-
Partition under win98.
This floppy was really working and so I could install Slackware12 with
the not normal
booting installation-cd.
All other trying of writing bootfloppy under Slackware10 or winxp
failed.
My cd/dvd-drive is bootable with all other bootable cds or dvds, only
not with that
cd from Slackware12 with the very new loading size 32. My cd-/dvd-
drive was built in
2005 by Benq. I would say, it isn't old! I don't know why Slackware
makes such trouble,
using such an unnormally method. Normally Slackware is a very good OS
supporting a lot
of hardware.
I hope, the configuration of Slackware12 will not make other new
problems!

Karin

Henrik Carlqvist

unread,
Sep 19, 2007, 2:36:03 AM9/19/07
to
KarinV <k.vo...@gmx.net> wrote:
> My cd/dvd-drive is bootable with all other bootable cds or dvds, only
> not with that cd from Slackware12 with the very new loading size 32. My
> cd-/dvd- drive was built in 2005 by Benq. I would say, it isn't old!

The problem is probably not the age or the capabilities of the CD but the
the age and the functionality of your motherboard. My guess is that you
have a motherboard with a rather old BIOS? A long time ago, something like
a 100 MHz pentium motherboard was not able to boot from any CD at all.
This was not a limitation of the CD drives but a limitation in the
motherboard BIOS.

regards Henrik
--
The address in the header is only to prevent spam. My real address is:
hc1(at)poolhem.se Examples of addresses which go to spammers:
root@localhost postmaster@localhost

Manuel Otto

unread,
Sep 19, 2007, 8:29:50 AM9/19/07
to
On Wed, 19 Sep 2007 08:36:03 +0200, Henrik Carlqvist
<Henrik.C...@deadspam.com> wrote:

>KarinV <k.vo...@gmx.net> wrote:
>> My cd/dvd-drive is bootable with all other bootable cds or dvds, only
>> not with that cd from Slackware12 with the very new loading size 32. My
>> cd-/dvd- drive was built in 2005 by Benq. I would say, it isn't old!
>
>The problem is probably not the age or the capabilities of the CD but the
>the age and the functionality of your motherboard. My guess is that you
>have a motherboard with a rather old BIOS? A long time ago, something like
>a 100 MHz pentium motherboard was not able to boot from any CD at all.
>This was not a limitation of the CD drives but a limitation in the
>motherboard BIOS.
>
>regards Henrik

The way I see the problem is that while Slackware always was somewhat
conservative in supporting the very latest, it suddenly turned around
180 degrees, has now latest X, quite a new kernel, even Compiz (who
needs it).

The downside is maybe that hardware just a little older not works,
like booting from dvd or cd.

Apart from the sudden 180 degree turn from conservative to very modern
(weird and unexpected), I still don't see a point in why
'-boot-load-size' should be 32, while nobody else seem to use that!

And it doesn't seem to be necessary either: ('-boot-load-size 32').

Am I right, or am I wrong?

Manuel

Henrik Carlqvist

unread,
Sep 19, 2007, 5:39:04 PM9/19/07
to
Manuel Otto <not-fo...@adress.is.invalid> wrote:
> Apart from the sudden 180 degree turn from conservative to very modern
> (weird and unexpected), I still don't see a point in why
> '-boot-load-size' should be 32, while nobody else seem to use that!
>
> And it doesn't seem to be necessary either: ('-boot-load-size 32').
>
> Am I right, or am I wrong?

From the README.TXT which came with Slackware 10.1 which used
-boot-load-size 4 :

-8<----------------------------------------
Techincally the --boot-load-size should be a lot bigger, like 20 or so
in order to hold the isolinux.bin boot block. However, setting it to
4 causes it to load on more BIOSes. I don't know why, but I've had so
many people report this to me that I'm inclined to believe it. But, if
the resulting discs don't boot in your machine and you find that using
a more correct value here fixes it, please let me know! If it's going
to be broken for some BIOSes either way, I'd rather be correct.
-8<----------------------------------------

So yes, it seems as if having -boot-load-size 4 really is wrong even
though some older hardware prefers to have it that way. Once that old
hardware gets obsolete there will no longer be any point in keeping the
wrong value. For old and obsolete hardware unable too boot from CDs (maybe
only some CDs with special values for boot-load size, maybe all CDs) at
least there is still the option to boot from a floppy. For newer hardware
booting from a floppy might not be an alternative as many machines today
are sold without floppy drives.

Manuel Otto

unread,
Sep 19, 2007, 8:08:00 PM9/19/07
to
On Wed, 19 Sep 2007 23:39:04 +0200, Henrik Carlqvist
<Henrik.C...@deadspam.com> wrote:

>Manuel Otto <not-fo...@adress.is.invalid> wrote:
>> Apart from the sudden 180 degree turn from conservative to very modern
>> (weird and unexpected), I still don't see a point in why
>> '-boot-load-size' should be 32, while nobody else seem to use that!
>>
>> And it doesn't seem to be necessary either: ('-boot-load-size 32').
>>
>> Am I right, or am I wrong?
>
>From the README.TXT which came with Slackware 10.1 which used
>-boot-load-size 4 :
>
>-8<----------------------------------------
> Techincally the --boot-load-size should be a lot bigger, like 20 or so
> in order to hold the isolinux.bin boot block. However, setting it to
> 4 causes it to load on more BIOSes. I don't know why, but I've had so
> many people report this to me that I'm inclined to believe it. But, if
> the resulting discs don't boot in your machine and you find that using
> a more correct value here fixes it, please let me know! If it's going
> to be broken for some BIOSes either way, I'd rather be correct.
>-8<----------------------------------------

I've read that, yes. That's how I was able to make my Salckware dvd
bootable.

Again, what I di don't understand that the other bootable cd's/dvd's I
tried lately all booted well, so they should be using teh technically
wrong value, seems to work on all machines: old or new...

You don't get my point?

>So yes, it seems as if having -boot-load-size 4 really is wrong even
>though some older hardware prefers to have it that way. Once that old
>hardware gets obsolete there will no longer be any point in keeping the
>wrong value. For old and obsolete hardware unable too boot from CDs (maybe
>only some CDs with special values for boot-load size, maybe all CDs) at
>least there is still the option to boot from a floppy. For newer hardware
>booting from a floppy might not be an alternative as many machines today
>are sold without floppy drives.

Well, having a floppy drive is always a good idea.

And what do you consider old hardware? My Slackware machine is a
Pentium 3, 1000MHz, 512 MB RAM, hda 40 GB, hdb 250 GB.

I'm very happy with how it performs, and don't consider it old.

Before Slackware 11, Slackware used to support *much* older hardware,
but now everything must be latest of the latest, it seems.

Considering that '-boot-load-size 4' seem to work on more more
machines, I don't understand why Slackware uses '-boot-load-size 32'.

I don't understand why Slackware from a conservative, very reliable,
'working even with quite old hardware'-approach suddenly changed 180
degrees, and got very modern, bleeding-edge.

That's my point.

Happy with latest Slackware, got it al working, solved all problems,
but wonder where this modernness will lead to with upcoming
versions...

Wonder if Slackware will remain Slackware as I knew it it, and like
it.

>regards Henrik

Have a nice day,
Manuel

Kees Theunissen

unread,
Sep 20, 2007, 1:00:08 AM9/20/07
to
Manuel Otto wrote:
> On Wed, 19 Sep 2007 23:39:04 +0200, Henrik Carlqvist
> <Henrik.C...@deadspam.com> wrote:

>>From the README.TXT which came with Slackware 10.1 which used
>> -boot-load-size 4 :
>>
>> -8<----------------------------------------
>> Techincally the --boot-load-size should be a lot bigger, like 20 or so
>> in order to hold the isolinux.bin boot block. However, setting it to
>> 4 causes it to load on more BIOSes. I don't know why, but I've had so
>> many people report this to me that I'm inclined to believe it. But, if
>> the resulting discs don't boot in your machine and you find that using
>> a more correct value here fixes it, please let me know! If it's going
>> to be broken for some BIOSes either way, I'd rather be correct.
>> -8<----------------------------------------
>
> I've read that, yes. That's how I was able to make my Salckware dvd
> bootable.
>
> Again, what I di don't understand that the other bootable cd's/dvd's I
> tried lately all booted well, so they should be using teh technically
> wrong value, seems to work on all machines: old or new...
>
> You don't get my point?

So you proved that all supposedly technically wrong CD's _you_own_
seem to work on all _your_ computers.
I really don't know why Patrick moved away from "--boot-load-size=4"
but he might have received reports that it doesn't work on some
modern systems.

[ snip ]

> I don't understand why Slackware from a conservative, very reliable,
> 'working even with quite old hardware'-approach suddenly changed 180
> degrees, and got very modern, bleeding-edge.
>
> That's my point.
>
> Happy with latest Slackware, got it al working, solved all problems,
> but wonder where this modernness will lead to with upcoming
> versions...

Is Slackware 12.0 really bleeding-edge?
It should be clear that a new OS release should support all modern
hardware. And that involves running a recent (the latest stable)
kernel that includes drivers for this modern hardware.
There shouldn't be any discussion at all about this.
People using old hardware, not supported by modern kernels anymore,
should simply use a previous Slackware release. Several previous
releases are still supported with security updates.

And yes, running the latest kernel involves several incompatibilities
with previous Slackware releases. That is why Slackware 12.0 is so
different from 11.0. I personally think that Pat delayed the Big Switch
(TM) too long.

Coming back to the bootable CD issue: all this "modern OS" stuff has
nothing to do with the "--boot-load-size=4" parameter. Those issues
just happen to coincide.

Finally I like to mention a work-around for the bootable CD issue
not mentioned before. If you want to install Slackware 12.0 on a
system that is current running any other Linux release, and if you
can't boot form the CD, you could try the following:
-- copy kernels/huge.s/bzImage and isolinux/initrd.img from the
installation CD/DVD somewhere to your hard disk (for instance
to /tmp).
-- add a stanza to your Lilo or Grub configuration referencing
these initrd and kernel.
For instance your lilo stanza could read:
image = /tmp/bzImage
initrd = /tmp/initrd.img
append = "load_ramdisk=1 prompt_ramdisk=0 rw SLACK_KERNEL=huge.s"
label = slack12_install
-- Reboot, using this stanza, and install Slackware 12.0.
If the installation of Slackware 12.0 involves erasing or overwriting
the boot loader and/or the kernel/ramdisk you used to boot, you would
better make sure that the new installation would boot the first time
after installation though.


Regards,

Kees.

--
Kees Theunissen.

Henrik Carlqvist

unread,
Sep 20, 2007, 2:28:33 AM9/20/07
to
Manuel Otto <not-fo...@adress.is.invalid> wrote:
> I've read that, yes. That's how I was able to make my Salckware dvd
> bootable.

So then we both know that the value has been 4 before and now the value
has changed to 32. As the value has changed we can assume there was a
reason for that change.



> Again, what I di don't understand that the other bootable cd's/dvd's I
> tried lately all booted well, so they should be using teh technically
> wrong value, seems to work on all machines: old or new...

The fact that you on your machines don't see any reason to change the
value doesn't mean that the reason doesn't exist. In this thread you ask
for the reason, but unfortunately I dont know why he finally decided to
change.

> You don't get my point?

I get your point that the old value used to work before and that changing
it breaks things for old hardware. However, I still tink there was a
reason that the value was changed even though I don't know which reason.

> And what do you consider old hardware?

Old could mean different things, like:

1) Not possible to buy anymore.

2) Not any longer having any economical value for book keeping purposes.
At least here in Sweden computers in a companies inventory usually have
no economical value after 3 years.

3) Technically broken because of age and wear.

> My Slackware machine is a Pentium 3, 1000MHz, 512 MB RAM, hda 40 GB, hdb
> 250 GB.

As that machine still is working it doesn't fulfill requirement 3 above,
however I don't think you can buy any such machine any longer and without
any knowledge of your book keeping rules my guess is that it as an
inventory doesn't have any economical value.

Even though this machine doesn't fulfill requirement 3 above it probably
would have done that if it would have been about 1 year newer. Many early
P4 motherboards had bad capacitors which made the motherboards unusable
after about 5 years.

On the other hand, if your machine would have been only 1-2 years older it
would probably not be able to boot from any CD at all.

> I'm very happy with how it performs, and don't consider it old.

In my network at home I still have 4 i486 machines up and running.
However, I would not even try to install Slackware 12 on any of them.

> Before Slackware 11, Slackware used to support *much* older hardware,
> but now everything must be latest of the latest, it seems.

Today your P3 machine is about 7 year old. When Slackware 1.1.2 was
released 1994 it would require at least a 386 system and would not run on
a 286 system sold 1988 6 years earlier. From
http://pycckuu.blogspot.com/2006/07/intel-desktop-processor-timeline-in-25.html

-8<--------------------------------------------
1982: 286 Microprocessor
The 286, also known as the 80286, was the first Intel processor that could
run all the software written for its predecessor. This software
compatibility remains a hallmark of Intel's family of microprocessors.
Within 6 years of it release, there were an estimated 15 million 286-based
personal computers installed around the world. The 286 processors contain
134,000 transistors.
-8<--------------------------------------------

> I don't understand why Slackware from a conservative, very reliable,
> 'working even with quite old hardware'-approach suddenly changed 180
> degrees, and got very modern, bleeding-edge.
>
> That's my point.
>
> Happy with latest Slackware, got it al working, solved all problems, but
> wonder where this modernness will lead to with upcoming versions...
>
> Wonder if Slackware will remain Slackware as I knew it it, and like it.

Slackware has allways included the latest stable version of most software.
Stability has been more important than latest, that is why glibc wasn't
upgraded until long after other distributions and that is why kernel 2.6
didn't become the default kernel until recently. I wouldn't say that
Slackware 12.0 is any more bleeding edge than any other *.0 version of
Slackware during the years.

I would say that the fact that -boot-load-size has changed so that a few
still old but working machines are unable to boot from this CD, even thogh
they barely are new enough to boot from other CDs, is a minor problem that
not many people will notice. Once noticed, there are known workarounds for
the problem. People with really old hardware have seen worse problems
during the years, like no longer being able to install from floppies only...

Some posts in this thread has had another point, why run the latest
Slackware on old hardware? Slackware releases back to 9.1 are still
maintained with security fixes. Slackware 9.1 is still a few years newer
than your hardware, but it will run fine on your machine. By this I don't
mean that it is the latest version that will work fine, but it is the
oldest version that still is maintained.

Manuel Otto

unread,
Sep 21, 2007, 12:31:35 PM9/21/07
to
On Thu, 20 Sep 2007 08:28:33 +0200, Henrik Carlqvist
<Henrik.C...@deadspam.com> wrote:

>Manuel Otto <not-fo...@adress.is.invalid> wrote:
>> I've read that, yes. That's how I was able to make my Salckware dvd
>> bootable.
>
>So then we both know that the value has been 4 before and now the value
>has changed to 32. As the value has changed we can assume there was a
>reason for that change.

Of course. But since I don't the reason, the reason *might* be as well
just that it is the technically correct value. If that's the only
reason, I'd like it more if it would be changed back to value 4 again.

>> Again, what I di don't understand that the other bootable cd's/dvd's I
>> tried lately all booted well, so they should be using teh technically
>> wrong value, seems to work on all machines: old or new...
>
>The fact that you on your machines don't see any reason to change the
>value doesn't mean that the reason doesn't exist. In this thread you ask
>for the reason, but unfortunately I dont know why he finally decided to
>change.
>
>> You don't get my point?
>
>I get your point that the old value used to work before and that changing
>it breaks things for old hardware. However, I still tink there was a
>reason that the value was changed even though I don't know which reason.

If anybody knows the reason, I would really like to read it.

>> And what do you consider old hardware?
>
>Old could mean different things, like:
>
>1) Not possible to buy anymore.

Bought my last machine second hand on Web, have no many for new
machines. But no, can not be bought in shop anymore.

>2) Not any longer having any economical value for book keeping purposes.
> At least here in Sweden computers in a companies inventory usually have
> no economical value after 3 years.

I'm not a company, I'm a single user in this home. I use computers
much longer then just 3 years.

>3) Technically broken because of age and wear.

This machien defenitly not yet.

>> My Slackware machine is a Pentium 3, 1000MHz, 512 MB RAM, hda 40 GB, hdb
>> 250 GB.
>
>As that machine still is working it doesn't fulfill requirement 3 above,
>however I don't think you can buy any such machine any longer and without
>any knowledge of your book keeping rules my guess is that it as an
>inventory doesn't have any economical value.

You are aright, but for me it has great economic value (beeing my
newest pc, with Linux on it, and cann't affort newer machine yet).

>Even though this machine doesn't fulfill requirement 3 above it probably
>would have done that if it would have been about 1 year newer. Many early
>P4 motherboards had bad capacitors which made the motherboards unusable
>after about 5 years.

Thanks for that info! Will avoid buying that hardware as my next
machine.

>On the other hand, if your machine would have been only 1-2 years older it
>would probably not be able to boot from any CD at all.

Hmmm..., just didn't think of that.

>> I'm very happy with how it performs, and don't consider it old.
>
>In my network at home I still have 4 i486 machines up and running.
>However, I would not even try to install Slackware 12 on any of them.
>
>> Before Slackware 11, Slackware used to support *much* older hardware,
>> but now everything must be latest of the latest, it seems.
>
>Today your P3 machine is about 7 year old. When Slackware 1.1.2 was
>released 1994 it would require at least a 386 system and would not run on
>a 286 system sold 1988 6 years earlier. From
>http://pycckuu.blogspot.com/2006/07/intel-desktop-processor-timeline-in-25.html
>
>-8<--------------------------------------------
>1982: 286 Microprocessor
>The 286, also known as the 80286, was the first Intel processor that could
>run all the software written for its predecessor. This software
>compatibility remains a hallmark of Intel's family of microprocessors.
>Within 6 years of it release, there were an estimated 15 million 286-based
>personal computers installed around the world. The 286 processors contain
>134,000 transistors.
>-8<--------------------------------------------

I never thought of my machine as an old machine, but now I see it is
really not that young anymore...

Still very happy with it though.

Thanks a lot for taking the time to explain this to me! Your point is
clear, I think I agree with you, and I was just overreacting, as I do
sometimes.

>regards Henrik

Appreciate,
Manuel

Manuel Otto

unread,
Sep 21, 2007, 12:31:34 PM9/21/07
to
On Thu, 20 Sep 2007 07:00:08 +0200, Kees Theunissen
<theu...@rijnh.nl> wrote:

>Manuel Otto wrote:
>> On Wed, 19 Sep 2007 23:39:04 +0200, Henrik Carlqvist
>> <Henrik.C...@deadspam.com> wrote:
>
>>>From the README.TXT which came with Slackware 10.1 which used
>>> -boot-load-size 4 :
>>>
>>> -8<----------------------------------------
>>> Techincally the --boot-load-size should be a lot bigger, like 20 or so
>>> in order to hold the isolinux.bin boot block. However, setting it to
>>> 4 causes it to load on more BIOSes. I don't know why, but I've had so
>>> many people report this to me that I'm inclined to believe it. But, if
>>> the resulting discs don't boot in your machine and you find that using
>>> a more correct value here fixes it, please let me know! If it's going
>>> to be broken for some BIOSes either way, I'd rather be correct.
>>> -8<----------------------------------------
>>
>> I've read that, yes. That's how I was able to make my Salckware dvd
>> bootable.
>>
>> Again, what I di don't understand that the other bootable cd's/dvd's I
>> tried lately all booted well, so they should be using teh technically
>> wrong value, seems to work on all machines: old or new...
>>
>> You don't get my point?
>
>So you proved that all supposedly technically wrong CD's _you_own_
>seem to work on all _your_ computers.

Yes...

>I really don't know why Patrick moved away from "--boot-load-size=4"
>but he might have received reports that it doesn't work on some
>modern systems.

I guess so, but the I don't understand why other (live) cd's don't
give a problem on my machines. If it doesn't give aproblem on my older
machines, it should give problems on newer machines. Still Debian live
cd, and Zenwalk live cd seem to use to the technically wrong value for
--boot-load-size.

That made me think why it would be needed to use the 32 value, but
yes, of course there will be a reason, which I don't now either.

>[ snip ]
>
>> I don't understand why Slackware from a conservative, very reliable,
>> 'working even with quite old hardware'-approach suddenly changed 180
>> degrees, and got very modern, bleeding-edge.
>>
>> That's my point.
>>
>> Happy with latest Slackware, got it al working, solved all problems,
>> but wonder where this modernness will lead to with upcoming
>> versions...
>
>Is Slackware 12.0 really bleeding-edge?

I switched from 10.2 to 12, and I found it bleeding edge, compared
with 10.2, but what I found is maybe not tha objective, you are
right...

>It should be clear that a new OS release should support all modern
>hardware. And that involves running a recent (the latest stable)
>kernel that includes drivers for this modern hardware.
>There shouldn't be any discussion at all about this.

Much agree, specially about 'latest *stable*'.

>People using old hardware, not supported by modern kernels anymore,
>should simply use a previous Slackware release. Several previous
>releases are still supported with security updates.

It was just that I thought that (but I better, in the future, know,
instead of 'think that') Slackware was well-known for it's
compatebility with (much) older hardware, and the suddenly there's
such a big jump to quite new kernel and other stuff, was but confusing
for me.

>And yes, running the latest kernel involves several incompatibilities
>with previous Slackware releases. That is why Slackware 12.0 is so
>different from 11.0. I personally think that Pat delayed the Big Switch
>(TM) too long.

= :-)

>Coming back to the bootable CD issue: all this "modern OS" stuff has
>nothing to do with the "--boot-load-size=4" parameter. Those issues
>just happen to coincide.

Okay, thanks for information.

I tend to cry out my dislikes to loud, and to quickly. It's
aprreciated that even if I do so, some people still take the time and
effort to react and explain!

>Finally I like to mention a work-around for the bootable CD issue
>not mentioned before. If you want to install Slackware 12.0 on a
>system that is current running any other Linux release, and if you
>can't boot form the CD, you could try the following:
>-- copy kernels/huge.s/bzImage and isolinux/initrd.img from the
> installation CD/DVD somewhere to your hard disk (for instance
> to /tmp).
>-- add a stanza to your Lilo or Grub configuration referencing
> these initrd and kernel.
> For instance your lilo stanza could read:
> image = /tmp/bzImage
> initrd = /tmp/initrd.img
> append = "load_ramdisk=1 prompt_ramdisk=0 rw SLACK_KERNEL=huge.s"
> label = slack12_install
>-- Reboot, using this stanza, and install Slackware 12.0.

Thank you! I learn there are many ways of solving a problem (once one
understands what the problem is, how it works, and why it works, or
not works).

> If the installation of Slackware 12.0 involves erasing or overwriting
> the boot loader and/or the kernel/ramdisk you used to boot, you would
> better make sure that the new installation would boot the first time
> after installation though.
>

So, kind of tricky...
>
>Regards,
>
>Kees.

Regards and thanks,
Manuel

Henrik Carlqvist

unread,
Sep 21, 2007, 2:41:57 PM9/21/07
to
Manuel Otto <not-fo...@adress.is.invalid> wrote:
> I never thought of my machine as an old machine, but now I see it is
> really not that young anymore...
>
> Still very happy with it though.

Even though the machine is a few years old it isn't that bad when it comes
to performance. When your machine was new the CPU performance
approximately doubled in about 18 months. So a machine that is 1.5 years
older than your machine would probably be a 500 MHz PIII. However, this
development pace has decreased alot. A P4 CPU is not as efficient per
clockcycle as a PIII. As a rule of thumb you can say that a 1 GHz PIII is
about as fast as a 1.5 GHz P4. Even though your machine is a lot older
than 18 months the fastest CPUs today are still about as fast as a 3 GHz
P4 which is only about twice as fast as your machine.

By the way. Some month ago I installed Slackware 12 on a 1 GHz PIII at
home. Because of this thread I started wondering how come that I didn't
get any problems with that installation as that machine very much
resembles your machine. As I have Slackware 12 on DVD and the machine only
has a CDROM I did an NFS installation. For that installation I used a
custom boot CD where most installation questions have been replaced with
my prefered values as default. That installation CD was created by a
script which I once wrote for Slackware 9.1 and now have modified for
Slackware 12. I hadn't noticed until now that my script still uses
"-boot-load-size 4". This kind of CD have also been used on 4 modern
machines without any problem. So, like you, I can say that I so far
haven't seen any problem with the value 4.

Manuel Otto

unread,
Sep 21, 2007, 5:31:18 PM9/21/07
to
On Fri, 21 Sep 2007 20:41:57 +0200, Henrik Carlqvist
<Henrik.C...@deadspam.com> wrote:

>Manuel Otto <not-fo...@adress.is.invalid> wrote:
>> I never thought of my machine as an old machine, but now I see it is
>> really not that young anymore...
>>
>> Still very happy with it though.
>
>Even though the machine is a few years old it isn't that bad when it comes
>to performance. When your machine was new the CPU performance
>approximately doubled in about 18 months. So a machine that is 1.5 years
>older than your machine would probably be a 500 MHz PIII. However, this
>development pace has decreased alot. A P4 CPU is not as efficient per
>clockcycle as a PIII. As a rule of thumb you can say that a 1 GHz PIII is
>about as fast as a 1.5 GHz P4. Even though your machine is a lot older
>than 18 months the fastest CPUs today are still about as fast as a 3 GHz
>P4 which is only about twice as fast as your machine.

Amazing! I always thought, or understoud, that for the same number of
MHz's, a newer version of a CPU would be more effecient, not less?

Quite confusing. But interesting. Have you any idea where my false
information came from, this exactly contrary to what I always did
read, like processors becoming more efficient because of new
instructions, pre-empting, and whatever more there is (don't now much
about it, am not really deep into hardware).

>By the way. Some month ago I installed Slackware 12 on a 1 GHz PIII at
>home. Because of this thread I started wondering how come that I didn't
>get any problems with that installation as that machine very much
>resembles your machine. As I have Slackware 12 on DVD and the machine only
>has a CDROM I did an NFS installation. For that installation I used a
>custom boot CD where most installation questions have been replaced with
>my prefered values as default. That installation CD was created by a
>script which I once wrote for Slackware 9.1 and now have modified for
>Slackware 12. I hadn't noticed until now that my script still uses
>"-boot-load-size 4". This kind of CD have also been used on 4 modern
>machines without any problem. So, like you, I can say that I so far
>haven't seen any problem with the value 4.

I hope there is someone who can explain why the value is set to 32.

If it's only for the sake of beeing technically correct, it might not
be needed to be 32 at all, with the consequence that older machines
won't boot with it, while among Slackware users there might be quite a
few who use older hardware (not only because of lack of money, but
maybe also because of the idea that it is madness to buy new every few
years, if old still works very fine. Waste of money and energy too,
polution, environment).

One could fix the problem by creating own iso with "-boot-load-size
4", but not everybody will find that solution.

One could choose not to use latest Slackware, but after some initial
doubts, I now like latest! :-) (Everything just works on this old
machine: HAL, Wine with programs I use it with, DVD-burning,
multimedia, no problem viewing DVD full screen, no lack of power or
speed).

Doing a search (didn't went to any of the search results, though,
search is just an example) for "-boot-load-size 4" returns 13,200
results: http://www.google.com/linux?q=%22-boot-load-size+4%22

Doing a search for "-boot-load-size 32" gives just 257 results, and
most results on the first page are somehow related to Slackware:
http://www.google.com/linux?q=%22-boot-load-size+32%22

Which gives the impression that Slackware is among the few using
"-boot-load-size 32".

>regards Henrik

Regards,
Manuel

Henrik Carlqvist

unread,
Sep 22, 2007, 2:54:51 AM9/22/07
to
Manuel Otto <not-fo...@adress.is.invalid> wrote:
>> A P4 CPU is not as efficient per clockcycle as a PIII. As a rule of
>> thumb you can say that a 1 GHz PIII is about as fast as a 1.5 GHz P4.

> Amazing! I always thought, or understoud, that for the same number of


> MHz's, a newer version of a CPU would be more effecient, not less?

As you say, this is usually true because of new instructions or other
improvements. It is also true for the newer core architecture from intel
which is faster than P4 per MHz, but also not running at very high clock
frequencies.

If I understand things right about the P4 it is slow because or the many
pipeline steps. Those pipeline steps means that each instruction take x
number of clock cycles, but also, x number of instructions at different
stages in the pipeline can be processed at the same time. The increased
number of pipeline steps was a design decision that was done to make it
possible to increase the clock frequency. The problem with many pipeline
steps comes when one instruction depends on the result of a previous
instruction.

AMD stuck with the old CPU architecture more resembling a PIII when it
came to performance per clock cycle. For this reason they did choose to
name their CPUs not from their frequency but from the frequency of an
equivalent performing P4. For this reason they called their CPUs something
like 2800+ when the frequency only was something like 2000 MHz.

Manuel Otto

unread,
Sep 22, 2007, 11:34:06 AM9/22/07
to

Thanks Henrik, very informative.

My conclusion: for my next new (second-hand) pc I'll wait till
something newer then P4 get's cheap enough for me to buy.

Have avery nice day!
Manuel

0 new messages