Partitioning Horrors?

[straightnut]

I was looking to buy the best partitioning software available, and
assumed it was Partition Magic, but I've read some very unappetizing
reviews of this product. And after scouring the web to find
alternatives, I have yet to find one that gets any glowing reviews.
I'm now ever so cautiously leaning towards Acronis Disk Director, but
this seems to have its share of adamant panners as well. I get the
impression that partitioning is a difficult thing for software to do
consistently well.

Jeff

[Arny Krueger]

"straightnut" > wrote in message


> I was looking to buy the best partitioning software
> available, and assumed it was Partition Magic, but I've
> read some very unappetizing reviews of this product. And
> after scouring the web to find alternatives, I have yet
> to find one that gets any glowing reviews. I'm now ever
> so cautiously leaning towards Acronis Disk Director, but
> this seems to have its share of adamant panners as well.
> I get the impression that partitioning is a difficult
> thing for software to do consistently well.

The basic business of partitioning hard drives involves next to nothing in
the way of processing or intelligence. For example, XPs' disk manager does
the basic job pretty well. XP's installation program does some basic
partitioning tasks, too.

It is all the bells and whistles that can come with repartitioning hard
drives that has the great potential for error and even disaster.

Furthermore, complex partitioning, and repartitioning of hard drives seems
to be falling into disfavor.

So, why are you even worrying about this problem in ways that can't be
solved with any copy of XP? ;-)

[Peter Larsen[_2_]]

straightnut wrote:

> I was looking to buy the best partitioning software available, and
> assumed it was Partition Magic, but I've read some very unappetizing
> reviews of this product. And after scouring the web to find
> alternatives, I have yet to find one that gets any glowing reviews.
> I'm now ever so cautiously leaning towards Acronis Disk Director, but
> this seems to have its share of adamant panners as well. I get the
> impression that partitioning is a difficult thing for software to do
> consistently well.

Partion Magic is a re-partitioning product, it can save the day and do
resizing and combining, you should preferably empty the partitions or drives
you want to change. There must be other good products out there by now, it
is no longer "the choice". I find I prefer to do it "the os way" now, things
are less simple than they were in its heyday.

For partitioning: use the disk manager or whatever that comes with the
actual OS. Right click "My Computer", select "Manage".

> Jeff


Kind regards

Peter Larsen

[Richard Crowley]

"straightnut" wrote ...
>I was looking to buy the best partitioning software available,

Why?

Why do you think that partitioning will buy you anything?

Why do you think you need some fancy software application
for such a simple and basic function (which is built into the
operating system)?

Unless you are trying to re-juggle a hard drive that is in-use
(NOT recommended!) the functionality of the operating system
should be all you need.

And hard drive space is so dirt-cheap, why does anybody fool
around with partitioning anyway? Just avoid the horror altogether.

[Scott Dorsey]

straightnut wrote:
>
> I was looking to buy the best partitioning software available, and
> assumed it was Partition Magic, but I've read some very unappetizing
> reviews of this product. And after scouring the web to find
> alternatives, I have yet to find one that gets any glowing reviews.
> I'm now ever so cautiously leaning towards Acronis Disk Director, but
> this seems to have its share of adamant panners as well. I get the
> impression that partitioning is a difficult thing for software to do
> consistently well.

No, partitioning is easy to do well and you need no additional software to
do it.

REpartitioning is difficult to do well, and in fact the chances of something
going wrong with any of the repartitioning systems is alarmingly high. But
we live now in the 21st century where disk space is cheap. There is no
longer really any reason to repartition your drive... just set up the
partitions on a new one, copy the data to it, then replace the drive with
the new one. Keep the old one on your shelf as a safety backup.

Repartitioning in place is just a bad idea. There was a time when there
was often no alternative because disk space was expensive. That time is
over.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[straightnut]

The reason I'm considering partitioning is the following article that
I read that made sense, not only to possibly improve performance, but
to organize my drives in a way that makes defragging and backups
easier.

http://www.soundonsound.com/sos/may05/articles/pcmusician.htm

I've navigated to the Windows XP Disk Management Utility and I
couldn't find a way to divide the one giant partition that contains
the operating system and applications already installed on my new PC.
That's why I figured a third party product would offer this. But it's
true that I'm not looking to resize partitions after the initial
setup, so if I can somehow partition with this utility I will. After
all it must have this capability if the utility even exists in the OS,
right?
I have 2 SATA drives, one of which in completely unformatted I
believe. The other, OS and Apps, is formatted in NTFS. I assume I
should format the other in NTFS as well if I'll be swapping files back
and forth from partitions of the soon to be audio data disk to those
of the OS/App disk?

Thanks,
Jeff

[straightnut]

On Dec 10, 2:47 pm, (Scott Dorsey) wrote:
> straightnut wrote:
>
> > I was looking to buy the best partitioning software available, and
> > assumed it was Partition Magic, but I've read some very unappetizing
> > reviews of this product. And after scouring the web to find
> > alternatives, I have yet to find one that gets any glowing reviews.
> > I'm now ever so cautiously leaning towards Acronis Disk Director, but
> > this seems to have its share of adamant panners as well. I get the
> > impression that partitioning is a difficult thing for software to do
> > consistently well.
>
> No, partitioning is easy to do well and you need no additional software to
> do it.
>
> REpartitioning is difficult to do well, and in fact the chances of something
> going wrong with any of the repartitioning systems is alarmingly high. But
> we live now in the 21st century where disk space is cheap. There is no
> longer really any reason to repartition your drive... just set up the
> partitions on a new one, copy the data to it, then replace the drive with
> the new one. Keep the old one on your shelf as a safety backup.
>
> Repartitioning in place is just a bad idea. There was a time when there
> was often no alternative because disk space was expensive. That time is
> over.
> --scott
>
> --
> "C'est un Nagra. C'est suisse, et tres, tres precis."

Forunately I'm starting completely fresh. I have nothing to transfer.
I just have 2 250G SATA drives that I wanted to configure with
partitions.
Jeff

[Richard Crowley]

"straightnut" wrote ...
> The reason I'm considering partitioning is the following article that
> I read that made sense, not only to possibly improve performance, but
> to organize my drives in a way that makes defragging and backups
> easier.
>
> http://www.soundonsound.com/sos/may05/articles/pcmusician.htm

The article admits in the very first paragraph that is was
virtually out of date when it was written almost 3 years ago.

Do you subscribe to computer magazines for musical advice?

I can't think of ANY valid reason to partition discs. Particularly
ones that are currently in use. You're just asking for trouble
with NO known advantage. Dunno why anyone is still doing
this?

[Arny Krueger]

"straightnut" > wrote in message


> Forunately I'm starting completely fresh. I have nothing
> to transfer. I just have 2 250G SATA drives that I wanted
> to configure with partitions.

If you are starting out fresh, then you have no need for repartitioning
software, right?

[Richard Crowley]

"straightnut" wrote ...
> Forunately I'm starting completely fresh. I have nothing to transfer.
> I just have 2 250G SATA drives that I wanted to configure with
> partitions.

If re starting with empty drives, then just use the functionality
you already have. Although I still see no valid advantage.

[Keith W. Blackwell]

straightnut > wrote:
> I was looking to buy the best partitioning software available, and
> assumed it was Partition Magic, but I've read some very unappetizing
> reviews of this product. And after scouring the web to find
> alternatives, I have yet to find one that gets any glowing reviews.
> I'm now ever so cautiously leaning towards Acronis Disk Director, but
> this seems to have its share of adamant panners as well. I get the
> impression that partitioning is a difficult thing for software to do
> consistently well.

Sorry, I've never used any products that cost money. The
last few times I've done repartitioning, GPartEd LiveCD gave
me what I needed. It's free: http://gparted-livecd.tuxfamily.org/
But it helps to know a little about what you're doing with it.
OK, to know a LOT. :-)

--
Keith W. Blackwell
(I do not speak for my employer or anyone else)

[Mike Rivers]

On Dec 10, 2:56 pm, straightnut > wrote:

> Forunately I'm starting completely fresh. I have nothing to transfer.
> I just have 2 250G SATA drives that I wanted to configure with
> partitions.

In that case, FDISK is the tool to use. I assume you're going to leave
your boot drive alone and partition your data drive. That would be the
smart thing to do.

[straightnut]

On Dec 10, 3:08 pm, "Arny Krueger" > wrote:
> "straightnut" > wrote in message
>
>
>
> > Forunately I'm starting completely fresh. I have nothing
> > to transfer. I just have 2 250G SATA drives that I wanted
> > to configure with partitions.
>
> If you are starting out fresh, then you have no need for repartitioning
> software, right?

You've all convinced me of that now, yes.
Jeff

[Laurence Payne]

On Mon, 10 Dec 2007 11:56:54 -0800 (PST), straightnut
> wrote:

>Forunately I'm starting completely fresh. I have nothing to transfer.
>I just have 2 250G SATA drives that I wanted to configure with
>partitions.
>Jeff

In which case you just need fdisk (or whatever Windows calls it now).
A simple Windows utility.

[straightnut]

On Dec 10, 3:12 pm, Mike Rivers > wrote:
> On Dec 10, 2:56 pm, straightnut > wrote:
>
> > Forunately I'm starting completely fresh. I have nothing to transfer.
> > I just have 2 250G SATA drives that I wanted to configure with
> > partitions.
>
> In that case, FDISK is the tool to use. I assume you're going to leave
> your boot drive alone and partition your data drive. That would be the
> smart thing to do.

I could do that. The article gives different schemes to take advantage
of the outer edge speed of the drive for streaming samples and for
audio data. I don't know if I'll be using streaming samples, but I was
going to leave a partition somewhere towards the outside for this
potentiality on one of the two disks. But now that you guys are all
yelling at me I would certainly reconsider.
With FDISK I assume I would wipe everything clean and have to
reinstall the operating system? If this is necessary I'll do it, if
not I would like to use the XP Disk Management tool if I can figure
out how. Maybe there's a help file somewhere.
Jeff

[Scott Dorsey]

Richard Crowley > wrote:
>"straightnut" wrote ...
>> The reason I'm considering partitioning is the following article that
>> I read that made sense, not only to possibly improve performance, but
>> to organize my drives in a way that makes defragging and backups
>> easier.
>>
>> http://www.soundonsound.com/sos/may05/articles/pcmusician.htm
>
>The article admits in the very first paragraph that is was
>virtually out of date when it was written almost 3 years ago.
>
>Do you subscribe to computer magazines for musical advice?
>
>I can't think of ANY valid reason to partition discs. Particularly
>ones that are currently in use. You're just asking for trouble
>with NO known advantage. Dunno why anyone is still doing
>this?

So if you overflow one fileystem it doesn't interfere with everyone else's
disk or with the OS disk in a shared system?
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Laurence Payne]

On Mon, 10 Dec 2007 12:25:00 -0800 (PST), straightnut
> wrote:

>I could do that. The article gives different schemes to take advantage
>of the outer edge speed of the drive for streaming samples and for
>audio data.

The numbers are doubtless correct. But I think you'll find little
practical difference. There's a lot more to disk performance than raw
reading rates. And modern mainstream drives are rarely the weak link
in a DAW system.

[Peter Larsen[_2_]]

straightnut wrote:

> The reason I'm considering partitioning is the following article that
> I read that made sense, not only to possibly improve performance, but
> to organize my drives in a way that makes defragging and backups
> easier.

Has some merit. Now re-read my first follow up, and do try to understand the
difference between partitioning and re-partitioning.

> http://www.soundonsound.com/sos/may05/articles/pcmusician.htm

> I've navigated to the Windows XP Disk Management Utility and I
> couldn't find a way to divide the one giant partition that contains
> the operating system and applications already installed on my new PC.

What you ask about is re-partitioning. To do it you have to empty the drive
first. Don't do it. My advice to you: use the bathtub principle, just format
the drives NTFS and be happy.


Kind regards

Peter Larsen

[Peter Larsen[_2_]]

Richard Crowley wrote:

> I can't think of ANY valid reason to partition discs. Particularly
> ones that are currently in use. You're just asking for trouble
> with NO known advantage. Dunno why anyone is still doing
> this?

It is not my advice to the OP, because the situation is different, but
fairly small partitions outermost on drives work well for me, because my
daw'ing is mostly done in the edit view of Audition and because I like to
protect my os and software installation via Drive Image 7. The theoretical
speed advantage is getting irrelevant, but the small OS and software
partition has some merit in case imaging is used as backup.


Kind regards

Peter Larsen

[Arny Krueger]

"Scott Dorsey" > wrote in message

> Richard Crowley > wrote:
>> "straightnut" wrote ...
>>> The reason I'm considering partitioning is the
>>> following article that I read that made sense, not only
>>> to possibly improve performance, but to organize my
>>> drives in a way that makes defragging and backups
>>> easier.
>>>
>>> http://www.soundonsound.com/sos/may05/articles/pcmusician.htm
>>
>> The article admits in the very first paragraph that is
>> was virtually out of date when it was written almost 3
>> years ago.
>>
>> Do you subscribe to computer magazines for musical
>> advice?
>>
>> I can't think of ANY valid reason to partition discs.
>> Particularly ones that are currently in use. You're just
>> asking for trouble with NO known advantage. Dunno why
>> anyone is still doing
>> this?
>
> So if you overflow one fileystem it doesn't interfere
> with everyone else's disk or with the OS disk in a shared
> system? --scott

By overflow do you mean fill the disk, or do you mean exhaust the bounds of
the file system?

No matter which, XP has a quota system that you can use to keep from
overflowing the bounds.

[Mike Rivers]

On Dec 10, 3:25 pm, straightnut > wrote:

> With FDISK I assume I would wipe everything clean and have to
> reinstall the operating system? If this is necessary I'll do it,

That's why I suggested leaving your boot drive alone and just
partition your data drive, which I assume is empty now. When you run
FDISK, you definitely wipe out anything on the drive and you'll have
to re-install Windows and anything else loaded on there. When you use
a partition manager you MIGHT wipe out what's on there, or rather,
make it unaccessible from the operating system.

> if
> not I would like to use the XP Disk Management tool if I can figure
> out how. Maybe there's a help file somewhere.

I've never found that useful for anything other than information.

[Arny Krueger]

"straightnut" > wrote in message

> On Dec 10, 3:08 pm, "Arny Krueger" >
> wrote:
>> "straightnut" > wrote in message
>>
>>
>>
>>> Forunately I'm starting completely fresh. I have nothing
>>> to transfer. I just have 2 250G SATA drives that I
>>> wanted to configure with partitions.
>>
>> If you are starting out fresh, then you have no need for
>> repartitioning software, right?
>
> You've all convinced me of that now, yes.

The other pro-partitioning idea I can give some credibility to, is the idea
of partitioning drives into high-performance (outer) and low-performance
(inner) partitions. Archived data can be put into the low-performance
partition.

[Mike Rivers]

On Dec 10, 4:19 pm, Chel van Gennip > wrote:

> If you look in the supplied graphs you see even 5400rpm 40GB disk will
> do better than 20MByte/sec on the wost part of the drive. In audio terms
> that is over 200 channels 16/44k1 audio.

Even if you figure that reality is 75% of theory, that's 150 channels,
or more than 100 channels at 24 bits. At 96 kHz, that's 46 channels.
Considering that when doing a punch-in you need to be playing each
channel while you're recording it, at 24/96 you can punch in on 23
channels at once. Geez, that isn't much. <G>

But then everybody needs something to write an article about.

[Richard Crowley]

"Scott Dorsey" wrote...
> Richard Crowley wrote:
>>I can't think of ANY valid reason to partition discs. Particularly
>>ones that are currently in use. You're just asking for trouble
>>with NO known advantage. Dunno why anyone is still doing
>>this?
>
> So if you overflow one fileystem it doesn't interfere with everyone else's
> disk or with the OS disk in a shared system?

Right. I was speaking from the context of "straightnut"s single-
OS, modern machine, though.

[straightnut]

On Dec 10, 4:19 pm, Chel van Gennip > wrote:

>
> Partitioning could introduce extra seek problems and only solves
> imaginary streaming problems.

What seek problems could be introduced?
Thanks,
Jeff

[straightnut]

On Dec 10, 5:35 pm, Mike Rivers > wrote:
>
> > if
> > not I would like to use the XP Disk Management tool if I can figure
> > out how. Maybe there's a help file somewhere.
>
> I've never found that useful for anything other than information.

I didn't look hard enough for the help file. I found it. It's vague,
but perhaps enough.
Jeff

[straightnut]

On Dec 10, 5:42 pm, Mike Rivers > wrote:
> On Dec 10, 4:19 pm, Chel van Gennip > wrote:
>
> > If you look in the supplied graphs you see even 5400rpm 40GB disk will
> > do better than 20MByte/sec on the wost part of the drive. In audio terms
> > that is over 200 channels 16/44k1 audio.
>
> Even if you figure that reality is 75% of theory, that's 150 channels,
> or more than 100 channels at 24 bits. At 96 kHz, that's 46 channels.
> Considering that when doing a punch-in you need to be playing each
> channel while you're recording it, at 24/96 you can punch in on 23
> channels at once. Geez, that isn't much. <G>
>
> But then everybody needs something to write an article about.

So channels are covered, but what about streaming samples? Is this a
non-issue as well?
Thanks,
Jeff

[Frank Stearns]

(Scott Dorsey) writes:

>straightnut wrote:
>>
>> I was looking to buy the best partitioning software available, and
>> assumed it was Partition Magic, but I've read some very unappetizing
>> reviews of this product. And after scouring the web to find
>> alternatives, I have yet to find one that gets any glowing reviews.
>> I'm now ever so cautiously leaning towards Acronis Disk Director, but
>> this seems to have its share of adamant panners as well. I get the
>> impression that partitioning is a difficult thing for software to do
>> consistently well.

>No, partitioning is easy to do well and you need no additional software to
>do it.

snips

>Repartitioning in place is just a bad idea. There was a time when there
>was often no alternative because disk space was expensive. That time is
>over.


To both Scott and Richard -

I must disagree a bit on this one... I partition the main system drive for two main
reasons:

- much easier and faster daily delta backups over the LAN to my "standby" image
machine (the old clunker that can take over in a matter of minutes should the main
machine croak.

- one more layer of simple, high-level organization: OS on the first partition, Apps
on the next, data files for those apps on the next, then, indeed, single drives as
one big partition for data (just the other day added another 1/2 terabyte drive for
audio projects).

Longer-term advantages include:

- I take "image" backups of partitions, such as the OS or selected Data partitions.
Again, hugely faster to do these in smaller chunks rather than imaging an entire
disk.

- in the case of a catastrophic OS hiccup (they don't happen often, but when
they do, yikes) damage is usually limited to just that partition. Again, from these
"smaller containers" of image backups, I can get back up and running way, way
faster, and I usually don't need to reinstall anything.

- easier migration to new hw.


Indeed, partition SW can muck things up if you don't think it through, but over the
past 22 years I've been on PCs (and even on coal-fired systems for 7-8 years before
that, and also in parallel on the Sun workstation I owned for many years) I can't
imagine surviving without partitions for the reasons noted.

But, as always, YMMV.

Frank Stearns
Mobile Audio

--

[Richard Crowley]

"straightnut" wrote ...
> Chel van Gennip wrote:
>> Partitioning could introduce extra seek problems and only solves
>> imaginary streaming problems.
>
> What seek problems could be introduced?

You lose contact with a "drive" while the (shared) head goes
away seeking, reading, writing content on one of the other
"drives" on the same spindle.

[Richard Crowley]

"Frank Stearns" wrote ...
> To both Scott and Richard -
> I must disagree a bit on this one... I partition the main
> system drive for two main reasons:
>
> - much easier and faster daily delta backups over the LAN...
> - one more layer of simple, high-level organization: ....
>
> Longer-term advantages include:
> - I take "image" backups of partitions, ....
> - damage is usually limited to just that partition. ....
> - easier migration to new hw.

I guess I rather do that myself for most of the same reasons.
But not with partitions. I generally use a small (~80GB) hard
drive dedicated as the boot/system/cache/program drive.

Then I store actual data on other drives (typically removable
ones organized by project, etc.) Same technique for both
audio and video projects.

[Peter Larsen[_2_]]

Arny Krueger wrote:

>> So if you overflow one fileystem it doesn't interfere
>> with everyone else's disk or with the OS disk in a shared
>> system? --scott

> By overflow do you mean fill the disk, or do you mean exhaust the
> bounds of the file system?

I think he means overwrite the disk. I've seen an NT4 box do it when 4
concurrent zip streams ran out of temp space.

> No matter which, XP has a quota system that you can use to keep from
> overflowing the bounds.

Yeah.


Kind regards

Peter Larsen

[Laurence Payne]

On Mon, 10 Dec 2007 18:34:15 -0800 (PST), straightnut
> wrote:

>So channels are covered, but what about streaming samples? Is this a
>non-issue as well?

Same difference really. Just as a standard disk can stream lots of
audio tracks, it can stream lots of samples. If you're going to be a
heavy sample user, it's a good idea of the samples are coming from a
different disk to the audio (not just a different partition of the
same disk. But, with adequate RAM installed, you might be surprised
just how much sample data is cached in RAM and doesn't NEED to stream.

[David Morgan \(MAMS\)]

"straightnut" > wrote in message ...
> I was looking to buy the best partitioning software available, and
> assumed it was Partition Magic, but I've read some very unappetizing
> reviews of this product. And after scouring the web to find
> alternatives, I have yet to find one that gets any glowing reviews.
> I'm now ever so cautiously leaning towards Acronis Disk Director, but
> this seems to have its share of adamant panners as well. I get the
> impression that partitioning is a difficult thing for software to do
> consistently well.
>
> Jeff

Partitioning is such a simple matter that it's built in to Windows installation
and it also comes with a number of odd-ball 'disk tool' sets.

For Windows XP, Partition Magic 8 is perfect. Forget all the bull**** extra
tools, you want a partition or two with names, sizes and file formats - PM8
will do this excellently and with ultimate simplicity and a massively friendly
user interface. The 'New Partition Wizard' literally kisses your ass. Never
had *any* issues with PM, since PowerQuest's version three and forward.

There.... have another "review". ;-) Not worth much, eh?

But I'm not on a blog or newsgroup trying to diss anyone's product or
sell my own, either. PM works and it works well.


--
David B. Morgan (MAMS)
Morgan Audio Media Service
http://www.m-a-m-s DOT com
Dallas, Texas (214) 662-9901
_____________________________
http://www.januarysound.com

[Nick Brown]

I started reading this thread last night, and was intending to post
here this morning to mention the advantages as I see them of
partitioning, but Frank's saved me the effort.

For anyone not going down the image backup route, I'd agree with other
folks here that partitioning is of limited value.

Even if the OP does decide to create separate partitions, that still
doesn't require third party-partitioning software, so long as he
doesn't subsequently change his mind about what partition size he
wants. (The C partition can be created at a user-specified size by the
Windows installer, and once Windows is running the remaining disk
space can be partitioned, formated and have a drive letter assigned by
the disk management tool - Start menu, run, "diskmgmt.msc" - without
even a reboot required.)

FWIW I've used Partition Magic on probably a hundred computers, and
I've always found it to behave itself. I don't know the Acronis
product, but I do use their TrueImage software (in preference to
Ghost), both the regular and server versions, and on account of it I
have a pretty high opinion of Acronis, so I'd guess their partitioning
software is worth a look, if you want partitioning software.

-Nick

[Mike Rivers]

On Dec 10, 9:34 pm, straightnut > wrote:

> So channels are covered, but what about streaming samples? Is this a
> non-issue as well?

Samples (I assume you're thinking about virtual instruments) unless
you're doing something really extreme probably uses less disk access
since samples are usually loaded into memory rather than played
directly off the disk.

[straightnut]

On Dec 10, 11:57 pm, "Richard Crowley" > wrote:

>
> > What seek problems could be introduced?
>
> You lose contact with a "drive" while the (shared) head goes
> away seeking, reading, writing content on one of the other
> "drives" on the same spindle.

I see. But why would the spindle care what part of the physical drive
it's been told to access? Is there extra instruction that delays the
spindle when going across partitions?

Jeff

[straightnut]

On Dec 11, 5:02 am, Laurence Payne <NOSPAMlpayne1ATdsl.pipex.com>
wrote:
> On Mon, 10 Dec 2007 18:34:15 -0800 (PST), straightnut
>
> > wrote:
> >So channels are covered, but what about streaming samples? Is this a
> >non-issue as well?
>
> Same difference really. Just as a standard disk can stream lots of
> audio tracks, it can stream lots of samples. If you're going to be a
> heavy sample user, it's a good idea of the samples are coming from a
> different disk to the audio (not just a different partition of the
> same disk.

That's why I was figuring on putting the sampler partition on the same
physical drive as the OS.

But, with adequate RAM installed, you might be surprised
> just how much sample data is cached in RAM and doesn't NEED to stream.

I was under the impression that Gigastudio, HALion, Kontakt were
designed to stream from the disk as opposed to RAM. Or do samplers
like these only use disk streaming if RAM is not available?

Thanks,
Jeff

[straightnut]

On Dec 11, 5:38 am, "David Morgan \(MAMS\)" /Odm>
wrote:
>
> Partitioning is such a simple matter that it's built in to Windows installation
> and it also comes with a number of odd-ball 'disk tool' sets.
>
> For Windows XP, Partition Magic 8 is perfect. Forget all the bull**** extra
> tools, you want a partition or two with names, sizes and file formats - PM8
> will do this excellently and with ultimate simplicity and a massively friendly
> user interface. The 'New Partition Wizard' literally kisses your ass. Never
> had *any* issues with PM, since PowerQuest's version three and forward.
>
> There.... have another "review". ;-) Not worth much, eh?
>
> But I'm not on a blog or newsgroup trying to diss anyone's product or
> sell my own, either. PM works and it works well.

You're right, there's just no way to trust the user reviews really.
But it's something when I have nothing else.
Thanks,
Jeff

[straightnut]

On Dec 11, 7:04 am, Mike Rivers > wrote:
> On Dec 10, 9:34 pm, straightnut > wrote:
>
> > So channels are covered, but what about streaming samples? Is this a
> > non-issue as well?
>
> Samples (I assume you're thinking about virtual instruments) unless
> you're doing something really extreme probably uses less disk access
> since samples are usually loaded into memory rather than played
> directly off the disk.

It was suggested in that article that Gigastudio, HALion, Kontakt and
the like are disk-streaming type samplers as opposed to those that run
from memory.
Jeff

[John Williamson]

straightnut wrote:
> On Dec 10, 11:57 pm, "Richard Crowley" > wrote:
>
>>> What seek problems could be introduced?
>> You lose contact with a "drive" while the (shared) head goes
>> away seeking, reading, writing content on one of the other
>> "drives" on the same spindle.
>
> I see. But why would the spindle care what part of the physical drive
> it's been told to access? Is there extra instruction that delays the
> spindle when going across partitions?
>
No there aren't, but each "drive" is held on a different physical area
of the platter, (The first drive of a partitioned HD would normally be
on the inner tracks, & the second drive on the outer tracks)& where with
2 real drives, data can be simultaneously accessed on both of them, with
a partitioned HD only one can be accessed at a time. This causes delays
due to the head traversing from one end of its positioning range to
the other, which is avoided if you use 2 HD units.

It also means that the head does a lot more moving around, which *may*
have an effect on the disc life, but I've not noticed this in practice.
The ideal situation for absolute maximum performance & reliability would
be a RAID 5 array for data, a boot drive which only held executables &
the operating system, & a separate drive purely for the swapfile. But
that's getting somewhat over the top....

Watching this thread, & with my experience, I'd only partition a Hard
Disc if I had no choice, with a rider that I always hold data & programs
in a different partition to the data, whether that other partition is
on a different HD or not. I also leave the OS & program install files &
things like install keys in the data partition. It saves a lot of
problems if you need to reinstall your OS for any reason, as the data is
(unless the HD fails) normally left untouched on the data partition, &
once you've booted the system from an external boot disk, installation
goes a fair bit faster.

--
Tciao for Now!

John.

[Richard Crowley]

"straightnut" wrote ...
> "Richard Crowley" wrote:
>> > What seek problems could be introduced?
>>
>> You lose contact with a "drive" while the (shared) head goes
>> away seeking, reading, writing content on one of the other
>> "drives" on the same spindle.
>
> I see. But why would the spindle care what part of the
> physical drive it's been told to access?

Because physically moving the head takes a L-O-N-G
time compared to straight reading or writing. More time
the farther it has to move. And sharing the same head
and spindle between two virtual drives means that the
head must move an average of halfway across the
disc to get from one "virtual drive" to the other. (And then
back again)

> Is there extra instruction that delays the
> spindle when going across partitions?

No. It is just the amount of time it takes to physically move
the head from the first "virtual drive" to the other one, and
then back again. The more compactly data is stored (as
with larger and larger hard drives), the more bandwidth
is lost when you have to physically move the head.
Because the head movement speed has not scaled to
the same extent as the data density of the platters.

Partitioning is much like RAID. They were both quite valid
solutions back when hard drives were much more
expensive and much smaller than they are now. But
they are now more often implemented by people who
don't make proper cost/benefit analysis of their situation.
Or even worse, just because it seems "cool". (Or whatever
word the kids are using these days :-)

Just because they were good ideas 5-10 years ago
doesn't make them desirable today. And I'm particularly
dubious of advice from magazines unless the author is
a known quantity (and the editor doesn't screw up the
content.)

[straightnut]

On Dec 11, 12:03 pm, straightnut > wrote:
> On Dec 10, 11:57 pm, "Richard Crowley" > wrote:
>
>
>
> > > What seek problems could be introduced?
>
> > You lose contact with a "drive" while the (shared) head goes
> > away seeking, reading, writing content on one of the other
> > "drives" on the same spindle.
>
> I see. But why would the spindle care what part of the physical drive
> it's been told to access? Is there extra instruction that delays the
> spindle when going across partitions?
>
> Jeff

I misused the term "spindle." I suppose I'm referring to "head."
Jeff

[straightnut]

After this thread, I'm fine with not partitioning the drives, but I
visualize data being spread out all over the drive if I don't
partition. So what tells the drive where to write data, and over time
will the data be written in a logical way for the drive to be able to
read the data without excessive head movement?
As new data comes in and has to be written on the drive, does it go
from outside to inside until the drive is full?
Does the frequency of doing defragmenting determine where this new
data will be written?
Thanks,
Jeff

[Frank Stearns]

"Richard Crowley" > writes:

>"straightnut" wrote ...
>> "Richard Crowley" wrote:
>>> > What seek problems could be introduced?
>>>
>>> You lose contact with a "drive" while the (shared) head goes
>>> away seeking, reading, writing content on one of the other
>>> "drives" on the same spindle.
>>
>> I see. But why would the spindle care what part of the
>> physical drive it's been told to access?

>Because physically moving the head takes a L-O-N-G
>time compared to straight reading or writing. More time
>the farther it has to move. And sharing the same head
>and spindle between two virtual drives means that the
>head must move an average of halfway across the
>disc to get from one "virtual drive" to the other. (And then
>back again)

Good point, but in fact I've found that partitioning seems to help performance
somewhat (not as much as it did at one time, though; but still enough to do it for
this plus the other reasons I've noted previously).

I attribute this to the way M$ file systems scatter data (and pieces of files) all
over the place, seemingly for no apparent reason. At least with a partition there is
a probability of a more constrained geometry and therefore a chance of less seeking.
Admittedly, a bit of a crap shoot. But I have noticed less "sounds of seeking" from
my systems than when I've been around other folks' doing roughly the same kinds of
things. Some of them seem to go crazy; mine are generally calm.

Of course, I DON'T apply this to big audio/video files and those Big File drives do
not get partitioned.

If there is method to M$ file system madness (NTFS, FAT, etc), I'd love to hear
about it. But based on other design "features" of M$ operating system, I really
don't trust them to do file systems optimally, either.

Again, YMMV. And, if this is getting too far a field from audio, that's fine too.
Just curious.

Frank Stearns
Mobile Audio

--

[Laurence Payne]

On Tue, 11 Dec 2007 09:13:39 -0800 (PST), straightnut
> wrote:

>> Samples (I assume you're thinking about virtual instruments) unless
>> you're doing something really extreme probably uses less disk access
>> since samples are usually loaded into memory rather than played
>> directly off the disk.
>
>It was suggested in that article that Gigastudio, HALion, Kontakt and
>the like are disk-streaming type samplers as opposed to those that run
>from memory.

They'll stream if they have to. But if there's ample RAM they would
be stupid to INSIST on streaming.

[Arny Krueger]

"straightnut" > wrote in message


> After this thread, I'm fine with not partitioning the
> drives, but I visualize data being spread out all over
> the drive if I don't partition.

Worse things could happen.

> So what tells the drive
> where to write data,

The NT file system.

> and over time will the data be
> written in a logical way for the drive to be able to read
> the data without excessive head movement?

The defrag program is supposed to keep an eye on data access patterns, and
try to put the right things in the right places.

> As new data comes in and has to be written on the drive,
> does it go from outside to inside until the drive is full?

Not necessarily. In fact, NTFS does limited scattering of data when it
writes it.


> Does the frequency of doing defragmenting determine where
> this new data will be written?

Only in the sense that defragging will free up some space where data used to
be.

[Mike Rivers]

On Dec 11, 12:13 pm, straightnut > wrote:

> It was suggested in that article that Gigastudio, HALion, Kontakt and
> the like are disk-streaming type samplers as opposed to those that run
> from memory.

Look, I know you're going to do what you already decided to do. Do it.
Get Partition Magic if you don't want to re-install Windows. Since you
don't have much installed in it yet other than what came with it, if
it screws up, you won't lose much except time. But take the advice of
most everyone here and don't juggle your partitions around once you
start working. Think it through, decide what you need, partition your
drives that way, and then leave them alone.

Re-partitioning tools were a godsend in the days of 20 MB (yes, MB,
not GB) drives where we thought we were being smart and partitioned
them when setting up the computer, then finding that our data was in
danger of overflowing our carefully chosen data partition size. Even
though data size has grown hugely since then, disk drive size has
grown faster. Be liberal when assigning your partition sizes so you
won't have to change them later on.

[Mike Rivers]

On Dec 11, 12:37 pm, John Williamson >
wrote:

> each "drive" is held on a different physical area
> of the platter, (The first drive of a partitioned HD would normally be
> on the inner tracks, & the second drive on the outer tracks)& where with
> 2 real drives, data can be simultaneously accessed on both of them, with
> a partitioned HD only one can be accessed at a time. This causes delays
> due to the head traversing from one end of its positioning range to
> the other

This is why you need to think about how your drive is partitioned. If
you put the operating system and applications (which normally go into
memory and don't need to access the disk on a regular basis) then the
heads can stay in the data area where things need to be accessed more
often.

But what happens when you have one partition for virtual instrument
sounds (assuming they need to stream off the disk) and another
partition for recorded audio? The head will be jumping back and forth
to stream both when you're playing your song. That could be hard on
the drive. It would be an advantage to have these two forms of data on
separate drives, but pretty soon you'll want a drive for everything
and you can't have that.

[Richard Crowley]

"straightnut" wrote ...
> After this thread, I'm fine with not partitioning the drives, but I
> visualize data being spread out all over the drive if I don't
> partition.

Partitioning doesn't prevent fragmentation. It just
fragments the fragmentation! :-)

De-fragging is always advisable on large drives with
large media (audio, video) files.

> So what tells the drive where to write data, and over time

A rather simplistic algorithm built into the file system.

> will the data be written in a logical way for the drive to be able to
> read the data without excessive head movement?

Not without defragging.

> As new data comes in and has to be written on the drive, does it go
> from outside to inside until the drive is full?

More or less. The file system seems to break up large files,
apparently with some notion that it will help it cope with files
of various sizes. It rarely keeps big files all together by itself.
It takes defragging to optimize the file packing for large files.

> Does the frequency of doing defragmenting determine where this new
> data will be written?

Yes.

Sometimes it is faster to copy the files to a new (empty) drive
than to run the defragger.

[straightnut]

On Dec 11, 3:51 pm, Mike Rivers > wrote:
> On Dec 11, 12:13 pm, straightnut > wrote:
>
> > It was suggested in that article that Gigastudio, HALion, Kontakt and
> > the like are disk-streaming type samplers as opposed to those that run
> > from memory.
>
> Look, I know you're going to do what you already decided to do. Do it.

I hadn't decided on anything before talking to you guys. I try to do
at least a little research before bugging you all here. I'm
continually amazed at how accessable you regulars are and all that you
offer us, so I don't want to bog you down with nonsense. Sometimes I'm
a little fried from the stress of getting my setup set up and overlook
things that are plain to see, like, as you pointed out, help files.
Your advice has saved me a lot of money and aggravation over the years
and I don't take it lightly.

At this point, I see no need for repartitioning, and possibly no need
to partition at all except to make my second disk drive usable from
its unallocated state. If I do want to use one of those samplers I'll
just be sure to have it installed on the unpartitioned OS drive.

I may still want to add a partition or two to the audio data drive,
but I don't know how to make sure that when I create more than one
partition on a drive using the built in system tools(FDisk or XP's
Disk Management) that the partitions will be where I want them to
be(outside, inside, middle).

I also just noticed that my two physical drives may have been assigned
in the wrong order. The unallocated disk is listed as Disk 0, and the
OS disk is Disk 1.
Will this be a problem when partitioning the other disk?

Also, am I correct that when I create partitions on the audio data
disk that no primary partitions should be created? All(if no more than
4) should be logical within one full-disk-sized extended partition?

Thanks again,
Jeff

[Nick Brown]

On 11 Dec, 23:10, straightnut > wrote:

> I may still want to add a partition or two to the audio data drive,
> but I don't know how to make sure that when I create more than one
> partition on a drive using the built in system tools(FDisk or XP's
> Disk Management) that the partitions will be where I want them to
> be(outside, inside, middle).

I don't think you get to explicitly choose with either fdisk or XP
disk management. When you create a partition it begins at the start of
the region of free space you've asked for it to be created in. So for
example the first partition you create on a new drive would be placed
at the start (i.e. the outside) of the drive.

>
> I also just noticed that my two physical drives may have been assigned
> in the wrong order. The unallocated disk is listed as Disk 0, and the
> OS disk is Disk 1.
> Will this be a problem when partitioning the other disk?

I don't think that's a problem.

> Also, am I correct that when I create partitions on the audio data
> disk that no primary partitions should be created? All(if no more than
> 4) should be logical within one full-disk-sized extended partition?

That's not necessary any more. Under DOS, and the versions of Windows
that ran on top of it (3, 95, 98, ME), if there was more than one
primary partition on a drive, the OS could only see whichever one was
set as being the active primary partition. The business of extended
and logical partitions was basically a work-around for that.

Windows NT and its descendants (2000, XP, presumably Vista) don't have
that limitation, and can happy access multiple primary partitions.
There's still a limit of a maximum of four primary partitions per
drive, so if for some reason you wanted more than that then yes, you'd
have to create an extended partition and put logical partitions inside
that. (The extended partition counts as the fourth of the
maximum four primary partitions, so you can have up to three
primaries, followed by the extended partition.)

Hope that makes sense.

-Nick

[straightnut]

On Dec 11, 11:19 pm, Nick Brown > wrote:
> On 11 Dec, 23:10, straightnut > wrote:
>
> > I may still want to add a partition or two to the audio data drive,
> > but I don't know how to make sure that when I create more than one
> > partition on a drive using the built in system tools(FDisk or XP's
> > Disk Management) that the partitions will be where I want them to
> > be(outside, inside, middle).
>
> I don't think you get to explicitly choose with either fdisk or XP
> disk management. When you create a partition it begins at the start of
> the region of free space you've asked for it to be created in. So for
> example the first partition you create on a new drive would be placed
> at the start (i.e. the outside) of the drive.

Good to know. Thanks.


> That's not necessary any more. Under DOS, and the versions of Windows
> that ran on top of it (3, 95, 98, ME), if there was more than one
> primary partition on a drive, the OS could only see whichever one was
> set as being the active primary partition. The business of extended
> and logical partitions was basically a work-around for that.
>
> Windows NT and its descendants (2000, XP, presumably Vista) don't have
> that limitation, and can happy access multiple primary partitions.
> There's still a limit of a maximum of four primary partitions per
> drive, so if for some reason you wanted more than that then yes, you'd
> have to create an extended partition and put logical partitions inside
> that. (The extended partition counts as the fourth of the
> maximum four primary partitions, so you can have up to three
> primaries, followed by the extended partition.)

So you can't do only extended partitions on a physical drive, right?
You need to first create 3 primary partitions before you can create
the one and only extended?
Thanks,
Jeff

[Peter Larsen[_2_]]

Frank Stearns wrote:

>>> "Richard Crowley" wrote:

>> Because physically moving the head takes a L-O-N-G
>> time compared to straight reading or writing. More time
>> the farther it has to move. And sharing the same head
>> and spindle between two virtual drives means that the
>> head must move an average of halfway across the
>> disc to get from one "virtual drive" to the other. (And then
>> back again)


Concurrent disk threads on two partitions on the same drive is indeed to be
avoided.

> Good point, but in fact I've found that partitioning seems to help
> performance somewhat (not as much as it did at one time, though; but
> still enough to do it for this plus the other reasons I've noted
> previously).

The (additional) advantage of the small os + software partitition (16679
megabytes if FAT32 for block size reasons, those were why one partioned so
carefully in the days of old!) is that things are never that far distant
from each other and it is fast to defrag.

Defragging a drive with interleaved multitrack audio files may be very
unwise ....

> I attribute this to the way M$ file systems scatter data (and pieces
> of files) all over the place, seemingly for no apparent reason.

Search microsoft.com for sysinternals, it seems they bought Russinovich out
of business, pagedefrag is a free download now.

> If there is method to M$ file system madness (NTFS, FAT, etc), I'd
> love to hear about it.

NTFS is designed - at least that is how it seems to me - with a raid 5 array
in mind, it is very tolerant about fragmentation and has a central database
and uses transaction logging. The central database makes it very good at
finding things on very large drives - up to 64 terabytes - very fast, but
the transaction logging causes it to have more overhead than FAT32 on a
linear operation. It is getting irrelevant now, but some years ago the
unofficial recommendation over in the microsoft newstree was to put the
pagefile on fat32 partitions with a 4 kbyte block size for minimum overhead,
the block size specifically recommended because it fitst the memory map
block sizes.

> But based on other design "features" of M$
> operating system, I really don't trust them to do file systems
> optimally, either.

With all file systems: do not fill them up, things may get crumpled just as
in a real filing cabinet, experimentally determined.

> Again, YMMV. And, if this is getting too far a field from audio,
> that's fine too. Just curious.

These days it is closer to things audio than tape recorders are, and they
are on topic.

> Frank Stearns
> Mobile Audio


Kind regards

Peter Larsen

[Peter Larsen[_2_]]

straightnut wrote:

> So you can't do only extended partitions on a physical drive, right?

The OS will not let you. Third party tools will. I remember of hearing of
someone doing it for some drive letter ordering practicalities.

> You need to first create 3 primary partitions before you can create
> the one and only extended?

The difference between a primary partition and an extended partition is that
the extended partition is a oontainer that can contain logical parititons.

The classic setup is a primary and an extended with logical paritions. I
still use it because I like to be win9x compatible because I am yet another
luddist.

> Thanks,
> Jeff


Kind regards

Peter Larsen

[Arny Krueger]

"Richard Crowley" > wrote in message


> Sometimes it is faster to copy the files to a new (empty)
> drive than to run the defragger.

I do a fair number of upgrades that involve copying the entire contents of a
bootable drive to a larger disk. The next defrag is never very short.

OTOH, it may be shorter than a defrag of a disk that has been in a computer
that was used by a teenager for a year or more, and never defragged. I've
seen those run for more than a day.

When you defrag by copying files on a mass basis, all information about file
position optimization will be lost.

[Nick Brown]

On 12 Dec, 06:14, straightnut > wrote:
> On Dec 11, 11:19 pm, Nick Brown > wrote:
> > Windows NT and its descendants (2000, XP, presumably Vista) don't
have
> > that limitation, and can happy access multiple primary partitions.
> > There's still a limit of a maximum of four primary partitions per
> > drive, so if for some reason you wanted more than that then yes, you'd
> > have to create an extended partition and put logical partitions inside
> > that. (The extended partition counts as the fourth of the
> > maximum four primary partitions, so you can have up to three
> > primaries, followed by the extended partition.)
>
> So you can't do only extended partitions on a physical drive, right?
> You need to first create 3 primary partitions before you can create
> the one and only extended?
> Thanks,
> Jeff

The possibilities are:

- Between 1 and 4 primary partitions, and no extended partition.
- Between 0 and 3 primary partitions, followed by an extended
partition (which itself needs to contain one or more logical
partitions).

So you can create an extended partition over the whole of a physical
drive, if you're intending to use it solely for data, as only primary
partitions are bootable. Windows will insist upon installing itself
onto a primary partition.

-Nick

[straightnut]

On Dec 12, 6:21 am, "Peter Larsen" > wrote:
> straightnut wrote:
> > So you can't do only extended partitions on a physical drive, right?
>
> The OS will not let you. Third party tools will. I remember of hearing of
> someone doing it for some drive letter ordering practicalities.
>
> > You need to first create 3 primary partitions before you can create
> > the one and only extended?
>
> The difference between a primary partition and an extended partition is that
> the extended partition is a oontainer that can contain logical parititons.
>
> The classic setup is a primary and an extended with logical paritions. I
> still use it because I like to be win9x compatible because I am yet another
> luddist.

All right Peter, I think I've got it. Thanks,
Jeff

[straightnut]

On Dec 12, 7:44 am, Nick Brown > wrote:
>
> The possibilities are:
>
> - Between 1 and 4 primary partitions, and no extended partition.
> - Between 0 and 3 primary partitions, followed by an extended
> partition (which itself needs to contain one or more logical
> partitions).
>
> So you can create an extended partition over the whole of a physical
> drive, if you're intending to use it solely for data, as only primary
> partitions are bootable. Windows will insist upon installing itself
> onto a primary partition.
>
> -Nick

Thanks. I've got it now.
Jeff

[David Morgan \(MAMS\)]

"straightnut" > wrote in message ...
> On Dec 12, 7:44 am, Nick Brown > wrote:
> >
> > The possibilities are:
> >
> > - Between 1 and 4 primary partitions, and no extended partition.
> > - Between 0 and 3 primary partitions, followed by an extended
> > partition (which itself needs to contain one or more logical
> > partitions).
> >
> > So you can create an extended partition over the whole of a physical
> > drive, if you're intending to use it solely for data, as only primary
> > partitions are bootable. Windows will insist upon installing itself
> > onto a primary partition.
> >
> > -Nick
>
> Thanks. I've got it now.
> Jeff


PartitionMagic 8 will not allow you to worry about this. However, if you
are on XP and you want the hard drive letters to be sequential and
to preceed the CD/DVD drive letters... you must disconnect those
drives before partitioning the hard disk.

[Nick Brown]

On 12 Dec, 20:40, "David Morgan \(MAMS\)" /Odm>
wrote:

> PartitionMagic 8 will not allow you to worry about this. However, if you
> are on XP and you want the hard drive letters to be sequential and
> to preceed the CD/DVD drive letters... you must disconnect those
> drives before partitioning the hard disk.

Right. Or use the XP disk management tool to change the CD drive
letter to something further down the alphabet before creating the
other partitions.

-Nick

[Arny Krueger]

"David Morgan (MAMS)" /Odm> wrote in
message news:DGX7j.11483$bW.3906@trnddc07
> "straightnut" > wrote in message
> ...
>> On Dec 12, 7:44 am, Nick Brown
>> > wrote:
>>>
>>> The possibilities are:
>>>
>>> - Between 1 and 4 primary partitions, and no extended
>>> partition.
>>> - Between 0 and 3 primary partitions, followed by an
>>> extended partition (which itself needs to contain one
>>> or more logical partitions).
>>>
>>> So you can create an extended partition over the whole
>>> of a physical drive, if you're intending to use it
>>> solely for data, as only primary partitions are
>>> bootable. Windows will insist upon installing itself
>>> onto a primary partition.
>>>
>>> -Nick
>>
>> Thanks. I've got it now.
>> Jeff
>
>
> PartitionMagic 8 will not allow you to worry about this.
> However, if you
> are on XP and you want the hard drive letters to be
> sequential and
> to preceed the CD/DVD drive letters... you must
> disconnect those
> drives before partitioning the hard disk.

Or, change all the drive letters but the boot drive after the fact with XP's
disk manager.

[David Morgan \(MAMS\)]

"Arny Krueger" > wrote in message . ..
> "David Morgan (MAMS)" /Odm> wrote in
> message news:DGX7j.11483$bW.3906@trnddc07
> > "straightnut" > wrote in message
> > ...
> >> On Dec 12, 7:44 am, Nick Brown
> >> > wrote:
> >>>
> >>> The possibilities are:
> >>>
> >>> - Between 1 and 4 primary partitions, and no extended
> >>> partition.
> >>> - Between 0 and 3 primary partitions, followed by an
> >>> extended partition (which itself needs to contain one
> >>> or more logical partitions).
> >>>
> >>> So you can create an extended partition over the whole
> >>> of a physical drive, if you're intending to use it
> >>> solely for data, as only primary partitions are
> >>> bootable. Windows will insist upon installing itself
> >>> onto a primary partition.
> >>>
> >>> -Nick
> >>
> >> Thanks. I've got it now.
> >> Jeff
> >
> >
> > PartitionMagic 8 will not allow you to worry about this.
> > However, if you
> > are on XP and you want the hard drive letters to be
> > sequential and
> > to preceed the CD/DVD drive letters... you must
> > disconnect those
> > drives before partitioning the hard disk.

> Or, change all the drive letters but the boot drive after the fact with XP's
> disk manager.

My experience is that "auto-detect" is the most accurate means to
accomplish this. Nothing gets altered in the registry and the need
to compensate for previous drive references is bypassed.

Unplug-em.... partition, plug 'em back in.... process complete.