On Sat, 10 Jan 2004 01:17:01 +0000, William Hooper wrote:

> "P.T. Breuer" <ptb@oboe.it.uc3m.es> wrote in message
> news:roentb.8oe.ln@news.it.uc3m.es...
>>
>>
>> William Hooper <whooper@infinet.com_nospam> wrote:
>> > "P.T. Breuer" <ptb@oboe.it.uc3m.es> wrote in message
>> > news:4aultb.62e.ln@news.it.uc3m.es...
>> > >
>> > > Ask rather why a cdrw must be master to a cdrom.
>> > >
>> > > Cdroms don't have the electronics to be master to a real
>> > > no-holds-barred full-spectrum freeform high-profile multi-vitamin
>> > > ide read/write device. They can't write themselves, so they aren't
>> > > going to be very good at managing somebody elses writing.
>>
>> > I call BS. Cite your source.
>>
>> You cite yours. I'll vote for my memory (I think the last time I looked
>> this up, I was in the atapi howto, or something equally arcane). You
>> google. I don't feel a need to back up my memory with quotes! If you
>> want to try, my first lookup would be "linux cdrom as master on ide
>> bus".
>
> It is your claim to prove, not mine to disprove. Others have already
> disagreed with you. Disproving something that doesn't exist is next to
> impossible.

I faintly recall that some of the early CDRW drives could be quite picky
as to whether they were slave or master. However, it has been 5 years or
more since I recall bumping into that issue. And the issue wasn't limited
to CD drives, but included HDs, too.

I don't believe this is an issue in the least with any of the recent or
current CDRW units on the market. I've had several configurations where
the CDRW was slave to a regular CD with zero problem.

I think someone is just living in the past.

noone wrote:
> William Hooper wrote:
>
>> "noone" <noone@noone.org> wrote in message
>> news:btjmhb$kn3$1@lust.ihug.co.nz...
>>
>>> So it looks like a defective CD problem, despite the fact that the
>>> checkmedia process says it passed ... but there's more to this as you
>>> will see below.
>>
>>
>>
>>
>> Possibly a CD-ROM drive problem. Did you try:
>>
>> linux ide=nodma
>>
>> at the boot prompt?
>>
>
>
> .. hhmmm .... isn't that supposed to be ideX=nodma?
> Which in my case, is ide1=nodma ( since the CDROM drive is IDE channel 1
> ) ?
>
>

I tried at the boot prompt:

linux ide1=nodma


.... and from /tmp/syslog, I said BAD OPTION, but I was able to do:


hdparm -d0 /dev/hdc

.... from the shell available in Ctrl-Alt-F2, which then disabled DMA.
Able to pass the copying and installation of setupXXX.rpm, but failed
elsewhere.


Same thing with:

linux ide=nodma


Tried with 2 sets of Fedora discs.

Might be a problem with the CDR burning software.

Can you do an MD5 on a burned ISO image on CD to check it is exactly the
same as the ISO file ?

On Sat, 10 Jan 2004 14:46:19 +1100,
noone <noone@noone.org> wrote:
> P.T. Breuer wrote:
> > noone <noone@noone.org> wrote:
> >
> >>P.T. Breuer wrote:
> >>
> >>>noone <noone@noone.org> wrote:
> >>>
> >>>>P.T. Breuer wrote:
> >>>>
> >>>>>noone <noone@noone.org> wrote:
> >>>>
> >>>>There is the CDROM drive and a CDRW drive on the second bus.
> >>>
> >>>That's what's wrong then. A cdrom must be slave to a cdrw.
> >
> >
> >>OK ... call me naive ... but why does a CDROM must be a slave to a CDRW ?
> >
> >
> > Ask rather why a cdrw must be master to a cdrom.
> >
> > Cdroms don't have the electronics to be master to a real
> > no-holds-barred full-spectrum freeform high-profile multi-vitamin
> > ide read/write device. They can't write themselves, so they aren't
> > going to be very good at managing somebody elses writing.
> >
> > Peter
>
>
> Almost all of the retail PCs sold with CD-ROM and CD-RW have CD-ROM as
> master and CD-RW as slave on the second IDE channel.
>
> If CD-ROMS cannot be master to a CDRW ...
>
> 1) Why then does my existing RH 7.2 installation works well accessing
> the CD-ROM and burning CDs with the CD-RW with the current setup as it
> is ( CD-ROM as master, CD-RW as slave ) ?
>
> 2) Whey then does windoze ( I dual-boot ) also works well just like RH 7.2 ?

You make the fastest device master, it's as simple as that. These days
recent cdroms can still read faster than recent cdrw's can write. You'd
have complaints that the cdrom doesn't read at such and such a speed.

If I got a new cdrom, I'd make it master and my 4x cdrw slave, otherwise
I'd have wasted my money on a modern cdrom.

Michael C.
--
mcsuper5@usol.com http://mcsuper5.freeshell.org/
Registered Linux User #303915 http://counter.li.org/

"noone" <noone@noone.org> wrote in message
news:btnt92$ii0$1@lust.ihug.co.nz...
>
>
> William Hooper wrote:
> > "noone" <noone@noone.org> wrote in message
> > news:btjmhb$kn3$1@lust.ihug.co.nz...
> >
> >>So it looks like a defective CD problem, despite the fact that the
> >>checkmedia process says it passed ... but there's more to this as you
> >>will see below.
> >
> >
> >
> > Possibly a CD-ROM drive problem. Did you try:
> >
> > linux ide=nodma
> >
> > at the boot prompt?
> >
> .. hhmmm .... isn't that supposed to be ideX=nodma?
> Which in my case, is ide1=nodma ( since the CDROM drive is IDE channel 1 )
?


From what I'm seeing in bugzilla ide=nodma is correct. I believe the
ideX=nodma is for the kernel after install, but I'm not sure.

Did I miss what kind of motherboard you have?

Do you have another machine or broadband? At this point a network install
might be easier.

--
William Hooper

Make it idiot-proof and someone will make a better idiot

William Hooper <whooper@infinet.com_nospam> wrote:
> "P.T. Breuer" <ptb@oboe.it.uc3m.es> wrote in message
> news:7pmntb.hda.ln@news.it.uc3m.es...

> > > It is your claim to prove, not mine to disprove.
> >
> > No, it's your claim that cdroms can be masters to normal ide devices!
> > You "prove" it.

> How is a CD-RW a "normal" ide device? Isn't it just as ATAPI as a CD-ROM
> drive?


It's even more atapi. But I said ide. IDE means "integrated drive
electronics".


> > > Others have already
> > > disagreed with you.
> >
> > No they haven't. I only recall one person, and I'm not sure that wasn't
> > you. And if they do, so what?

> "No they haven't... well, that one, but nobody else..."


Your problem with singular/plurals is ...?


> > > Disproving something that doesn't exist is next to impossible.
> >
> > There you are then - you prove it, since you claim you have a hold of
> > the fact. You find the doc that says that cdroms can master ide disks.
> >
> > Maybe it exists, maybe it doesn't!


> Following some links from "How stuff works" (you should try it sometime, you
> might learn something).
> http://www.pcguide.com/ref/hdd/if/ide/conf.htm
> "Note that despite the hierarchical-sounding names of "master" and "slave",
> the master drive does not have any special status compared to the slave one;
> they are really equals in most respects. The slave drive doesn't rely on the
> master drive for its operation or anything like that..."


That's interesting! Well done! But I think it does depend for some
important things, like arbitrating the bus and startup.


> You see, IDE (Integrated Drive Electronics) means that all the electronics
> to operate the drive are (wait for it) integrated with the drive...



Yep. And that is what I believe is not the case for cdroms.


> > > > Oh, a random scan through the atapi howto (WHY do I do this for you?)
> > > > stumps up the following kind of burble:
> > > >
> > > > Q5: I've installed the drive as the only device on the secondary
> > > ^^^^^^^^
> > > > interface, with the CDROM drive jumpered as SLAVE/SINGLE, and Linux
> > > > won't see it. What am I missing?
> >
> > > That doesn't say anything about using a CD-RW as a slave with a CD-Rom
> as
> > > master.
> >
> > Nor do I much care about cdrws as slaves in particular - cdrws aren't
> > partcularly generic ide devices either.

> I would rely on that memory of yours too much:


Thank you. (I know what I said!).


> > > > > > Ask rather why a cdrw must be master to a cdrom.

> > No, it's your claim that cdroms can be masters to normal ide devices!

> You just forgot what you were talking about twice in one sentence.


No I didn't. I know. If you wish to think so, that's your problem.
I've been talking about the consequences of cdroms not having the
electronics on them necessary to operate as master on an ide bus to an
ide slave. What you *think* I have been talking about is a product of
your perception and my expression, and is not the same.


Peter

Michael C. <mcsuper5@usol.com> wrote:
> > If CD-ROMS cannot be master to a CDRW ...
> >
> You make the fastest device master, it's as simple as that. These days

Yes, that's true. The master is the one that's talked to and which
mediates the bus.

> recent cdroms can still read faster than recent cdrw's can write. You'd
> have complaints that the cdrom doesn't read at such and such a speed.

> If I got a new cdrom, I'd make it master and my 4x cdrw slave, otherwise
> I'd have wasted my money on a modern cdrom.

And how fast does your cdrom write?

Peter

noone <noone@noone.org> wrote:
> P.T. Breuer wrote:
> > noone <noone@noone.org> wrote:
> >>P.T. Breuer wrote:
> >>>noone <noone@noone.org> wrote:
> >>>>P.T. Breuer wrote:
> >>>>>noone <noone@noone.org> wrote:
> >>>>
> >>>>There is the CDROM drive and a CDRW drive on the second bus.
> >>>
> >>>That's what's wrong then. A cdrom must be slave to a cdrw.
> >
> >>OK ... call me naive ... but why does a CDROM must be a slave to a CDRW ?
> >
> > Ask rather why a cdrw must be master to a cdrom.
> >
> > Cdroms don't have the electronics to be master to a real
> > no-holds-barred full-spectrum freeform high-profile multi-vitamin
> > ide read/write device. They can't write themselves, so they aren't
> > going to be very good at managing somebody elses writing.


> Almost all of the retail PCs sold with CD-ROM and CD-RW have CD-ROM as
> master and CD-RW as slave on the second IDE channel.


Then they are jumpered "wrong". All cdroms produced by manufacturers come
with the cdrom jumpered as slave, and it takes skill to invert that in
the wrong context!

Mind you, consistently getting things wrong would not be new for retail
shops. I recall them always forgetting to terminate scsi busses, 100%
of the time.


> If CD-ROMS cannot be master to a CDRW ...

Why are you focussing on that, when I said: "cdroms don't have the
electronics to be master to a real no-holds-barred full-spectrum
freeform high-profile multi-vitamin ide read/write device. They can't
write themselves, so they aren't going to be very good at managing
somebody elses writing".

I can't put much more emphasis on it!

> 1) Why then does my existing RH 7.2 installation works well accessing
> the CD-ROM and burning CDs with the CD-RW with the current setup as it
> is ( CD-ROM as master, CD-RW as slave ) ?

Luck. Green fingers. Never tried using them both at the same time. Etc.

> 2) Whey then does windoze ( I dual-boot ) also works well just like RH 7.2 ?

Because it doesn't.


Peter

"P.T. Breuer" <ptb@oboe.it.uc3m.es> wrote in message
news:tdfotb.6ub.ln@news.it.uc3m.es...
>
> No I didn't. I know. If you wish to think so, that's your problem.
> I've been talking about the consequences of cdroms not having the
> electronics on them necessary to operate as master on an ide bus to an
> ide slave.

This is false, and has been disputed now by multiple people with working
configurations. I know, they have just been "lucky". Yeah, that's it.

Until you prove your statement, I stand by my comment that it is bull shit.

--
William Hooper

Finally, a good use for Clippy!
"I see you're trying to email a program to every member of your
Outlook Address book. DON'T DO THAT YOU FUCKING MORON!"
-- Seen on Slashdot.org

"noone" <noone@noone.org> wrote in message
news:bto05v$ksc$1@lust.ihug.co.nz...
> Tried with 2 sets of Fedora discs.
>
> Might be a problem with the CDR burning software.
>
> Can you do an MD5 on a burned ISO image on CD to check it is exactly the
> same as the ISO file ?

This is in essence what the "mediacheck" at the beginning of the install
does.

I have had a similar situation once where the media check passed, but the
install didn't. After messing with it for a while I just swapped a
different CD-ROM drive in and it worked.

--
William Hooper

Finally, a good use for Clippy!
"I see you're trying to email a program to every member of your
Outlook Address book. DON'T DO THAT YOU FUCKING MORON!"
-- Seen on Slashdot.org

William Hooper <whooper@infinet.com_nospam> wrote:
> "P.T. Breuer" <ptb@oboe.it.uc3m.es> wrote in message
> news:tdfotb.6ub.ln@news.it.uc3m.es...
> >
> > No I didn't. I know. If you wish to think so, that's your problem.
> > I've been talking about the consequences of cdroms not having the
> > electronics on them necessary to operate as master on an ide bus to an
> > ide slave.

> This is false, and has been disputed now by multiple people with working
> configurations.

Stop making claims based on other peoples claims, and I have seen no
"multiple people". Maybe your idea of a crowd is different. Perhaps
their notions of working are different to yours. Perhaps they don't
notice dropouts, inability to read while writing, etc. etc. Perhaps
they have hardwired the device "recognition" into their kernel boot
parameters. Who knows!

> I know, they have just been "lucky". Yeah, that's it.

Possibly.

> Until you prove your statement, I stand by my comment that it is bull shit.

And I "stand by" my noncomment that yours is ape shit, so nurr-nurr.
/You/ can prove it. I don't see why /I/ should. You haven't come
up with a document that says "a cdrom may master an ide drive". If you
want swords at dawn, at least flash a butter knife.


Peter