Hard drives again...
Andy Hewitt
preparation for future plans, and so I can incorporate my Aperture
library into it.
I'm looking at going for a 1TB drive, which is far far more than I have
in data at the moment, so should last a fair while.
Whats the latest in terms of reliability with hard drives? I tend to
avoid WD and Maxtor drives (mainly from past experience and reputation),
and Samsung (on past experience with other products). I've always had
good service from Seagate and Hitachi drives, having two 500GB Hitachis,
and two of three Seagates fof smaller sizes.
As this is for a TM drive, and speed isn't a great priority, what about
these new 5900rpm Seagate drives I'm seeing. They're at a very good
price (cost is a big factor for me), and look like they could be quieter
and maybe even more reliable (if you believe the tech blurb).
Cheers.
--
Andy Hewitt
<http://web.me.com/andrewhewitt1/>
Duncan Kennedy
<thewil...@me.com> writes
>It's getting time to upgrade my Time Machine drive (again!), mostly in
>preparation for future plans, and so I can incorporate my Aperture
>library into it.
>
>I'm looking at going for a 1TB drive, which is far far more than I have
>in data at the moment, so should last a fair while.
>
>Whats the latest in terms of reliability with hard drives? I tend to
>avoid WD and Maxtor drives (mainly from past experience and reputation),
>and Samsung (on past experience with other products). I've always had
>good service from Seagate and Hitachi drives, having two 500GB Hitachis,
>and two of three Seagates fof smaller sizes.
>
"Click of Death" noises from a specific model of big Seagate (I *think,
1TB). Last year much the same happened to a lot of people with a
specific model of 500MB.
I know no more but as I'm looking at new drives too I'm beginning to
wonder if there are any left where somebody has not reported something!
--
Duncan K
Downtown Dalgety Bay
Mark Ingle
>
> I'm looking at going for a 1TB drive, which is far far more than I have
> in data at the moment, so should last a fair while.
>
have been so unreliable that I am swapping them out. The last batch of
external Seagates we had were appalling - over 50% failure within a
short period of time with the dreaded click of death... I would never
buy another Seagate/Maxtor drive again; shame as a couple of years ago
they were very good. The only good thing I can say about Seagate is
that their warranty service is very good; we've had to use it a lot and
they do swap out the drives pretty quickly.
I currently replacing them with Hitachi ones (probably have about 15 so
far) and so far they have been totally troublefree.
We also have a few external Western Digital drives, and though we have
had one or two failures, they've generally been good and the MyBook ones
we have look quite good and have FW.
Having said that, I haven't had any issues with Samsung drives, though
we haven't used enough of them to have a sufficiently large sample.
Andy Hewitt
Righto, thanks. I've done a little searching on the XLR8 site, and that
seems to agree with your finding too, rather a lot of later Seagate
failures. I was wondering about the later Samsungs too, they seem to get
good reviews, and little about failures. I think my Seagates are at
least two years old, so were still good then. I've been happy enough
with the Hitachis of late, although they can be a bit noisy.
Elliott Roper
<thewil...@me.com> wrote:
> It's getting time to upgrade my Time Machine drive (again!), mostly in
> preparation for future plans, and so I can incorporate my Aperture
> library into it.
>
> I'm looking at going for a 1TB drive, which is far far more than I have
> in data at the moment, so should last a fair while.
>
> Whats the latest in terms of reliability with hard drives? I tend to
> avoid WD and Maxtor drives (mainly from past experience and reputation),
> and Samsung (on past experience with other products). I've always had
> good service from Seagate and Hitachi drives, having two 500GB Hitachis,
> and two of three Seagates fof smaller sizes.
I'm using 1TB WD greens. All six are still a bit new, so I have no data
on longevity. They are quiet, run cool and transfer data a bit quicker
than I was led to expect.
I chickened out of Time Machining my Aperture Library on grounds of
size and consistency. I would have turned off time machine while in
Aperture anyway to avoid TM grabbing an inconsistent copy at an
unfortunate time. I figure Aperture locking one out while updating
vaults was done for a reason. My photos are far too valuable to me to
entrust them to Time Machine's high handed pruning malarkey. Instead I
rotate through 3 vaults, each on WD greens, one of which is stored off
site. My Library is on a 2 drive stripe of WD greens. The vaults are
held on single drives - there is not much advantage in speed and no
point in doubling the probability of a drive death taking out the
vault.
I have relegated Time Machine to dealing with accidental deletions
within my user folder and a narrow 1 day window for failure of my user
folder on the system drive. Everything else is handled through
SuperDuper scripts and manual vault updates on a daily basis.
--
To de-mung my e-mail address:- fsnospam$elliott$$
PGP Fingerprint: 1A96 3CF7 637F 896B C810 E199 7E5C A9E4 8E59 E248
T i m
wrote:
>> Having said that, I haven't had any issues with Samsung drives, though
>> we haven't used enough of them to have a sufficiently large sample.
>
>Righto, thanks. I've done a little searching on the XLR8 site, and that
>seems to agree with your finding too, rather a lot of later Seagate
>failures. I was wondering about the later Samsungs too, they seem to get
>good reviews, and little about failures.
For some reason I get a feeling of pleasure when I get a mate to get
me in a drive (based on cost / capacity) and it turns out to be a
Samsung. So far they seem to be quiet, reasonably fast and <touching
reconstituted wood fibres> reliable.
Cheers, T i m
p.s. The Hitachi HTS542516K95A00 I put in this Mini is also ok. Fairly
quick and pretty quiet.
Andy Hewitt
I might still do something similar anyway, at least keep my vault.
However, I thought they'd sorted the Aperture/TM issues now - doesn't TM
ignore the Aperture library if Aperture is open?
> I have relegated Time Machine to dealing with accidental deletions
> within my user folder and a narrow 1 day window for failure of my user
> folder on the system drive. Everything else is handled through
> SuperDuper scripts and manual vault updates on a daily basis.
I also use SuperDuper to do a backup clone of my main hard drive.
Andy Hewitt
Excellent, thanks, that's pretty much what I'm finding as I do more
searching - although I see there were some issues with the Samsung F1s.
Elliott Roper
<thewil...@me.com> wrote:
> Elliott Roper <nos...@yrl.co.uk> wrote:
>
<snip>
> > I chickened out of Time Machining my Aperture Library on grounds of
> > size and consistency. I would have turned off time machine while in
> > Aperture anyway to avoid TM grabbing an inconsistent copy at an
> > unfortunate time.
<snip>
>
> I might still do something similar anyway, at least keep my vault.
> However, I thought they'd sorted the Aperture/TM issues now - doesn't TM
> ignore the Aperture library if Aperture is open?
I didn't know that. It makes it mildly useless, which is an improvement
from downright dangerous. I get a choice of quitting Aperture and
deading around waiting for Time Machine or keeping Aperture running and
deading around waiting for a vault update to complete. I like to keep
Aperture running for long periods, so TM would serve little purpose for
me. I rotate through vaults whenever I import new stuff and before
clearing the images off the the camera's CF card. Well, at least when
it includes shots I really can't lose. The rest of the time I'm a bit
reckless.
A quick google and a local test still favours vault over Time Machine.
A poster at
http://www.apertureprofessional.com/showthread.php?t=12015
asserts that Time Machine wrote 730MB of his 25GB library and that was
the size of the database inside his library - all for changing the star
rating on a single image. I repeated his experiment using vaults. 1
Update vault. 2 change star rating on a single image. 3 Update vault.
According to Activity monitor, less than 10 MB was written to all disks
during the second vault update, which took about 30 sec. (My Aperture
Library is currently pushing 400GB and the stuff inside it that looks
database-ish is about 500MB) I therefore conclude that the vault update
does not attempt to re-write the whole database and that Time Machine
probably does.
Andy Hewitt
Ok, thanks for that. I haven't actually done any testing myself, as my
TM drive currently doesn't have enough space to add my 120GB Aperture
library. Perhaps just another drive for an additional Vault may be
better, and leave the rest as it is.
Jon B
I've been using Samsungs for the big stuff based on recommendations from
Odie. Not seen any issues. Then buy a roll your own case so you know
what spec drive is inside.
--
Jon B
Above email address IS valid.
<http://www.bramley-computers.co.uk/> Apple Laptop Repairs.
Andy Hewitt
Thanks, I think that's the way I'm going.
I have a modified PC tower case already set up to take a number of extra
drives :-). I could probably get another 2 or 3 in there for the cost of
a bridge adapter.
J. J. Lodder
> Andy Hewitt <thewil...@me.com> wrote:
>
> > It's getting time to upgrade my Time Machine drive (again!), mostly in
> > preparation for future plans, and so I can incorporate my Aperture
> > library into it.
> >
> > I'm looking at going for a 1TB drive, which is far far more than I have
> > in data at the moment, so should last a fair while.
> >
> > Whats the latest in terms of reliability with hard drives? I tend to
> > avoid WD and Maxtor drives (mainly from past experience and reputation),
> > and Samsung (on past experience with other products). I've always had
> > good service from Seagate and Hitachi drives, having two 500GB Hitachis,
> > and two of three Seagates fof smaller sizes.
> >
> > As this is for a TM drive, and speed isn't a great priority, what about
> > these new 5900rpm Seagate drives I'm seeing. They're at a very good
> > price (cost is a big factor for me), and look like they could be quieter
> > and maybe even more reliable (if you believe the tech blurb).
> >
> I've been using Samsungs for the big stuff based on recommendations from
> Odie.
Internet folklore doesn't become more reliable by someone echoing it.
And the big stuff is precisely Samsung's problem:
they are lagging far behind WD and also Hitachi
with the 2 TB drives.
> Not seen any issues. Then buy a roll your own case so you know
> what spec drive is inside.
With HD size increasing so rapidly
it is a pity to stop at 1.5 TB,
which will be obsolete sooner,
Jan
Rowland McDonnell
> It's getting time to upgrade my Time Machine drive (again!), mostly in
> preparation for future plans, and so I can incorporate my Aperture
> library into it.
>
> I'm looking at going for a 1TB drive, which is far far more than I have
> in data at the moment, so should last a fair while.
>
> Whats the latest in terms of reliability with hard drives?
No idea, but 1.5TB was more bits per penny than 1TB last time I looked -
which wasn't so long ago. I'd not buy a 1TB HDD for that reason.
I've still never had a Seagate pack up on me.
[snip]
> As this is for a TM drive, and speed isn't a great priority, what about
> these new 5900rpm Seagate drives I'm seeing. They're at a very good
> price (cost is a big factor for me), and look like they could be quieter
> and maybe even more reliable
Why more reliable? Don't bet on the lower speed - it's all about what
the mechanics are designed to do. 7200rpm and 5400rpm are in any case
more or less the same sort of speed - erm, max wear is on start up and
shut down, the rest of the time the main disc bearing gets hydrodynamic
lubrication, so wear there is practically non-existent.
>(if you believe the tech blurb).
You won't want to hear my suggestion of `get a twin HDD FW800 enclosure
and run a RAID mirror with a pair of 1.5TB Seagate Barracudas'.
Rowland.
--
Remove the animal for email address: rowland....@dog.physics.org
Sorry - the spam got to me
http://www.mag-uk.org http://www.bmf.co.uk
UK biker? Join MAG and the BMF and stop the Eurocrats banning biking
Rowland McDonnell
[snip]
> Internet folklore doesn't become more reliable by someone echoing it.
That which applies to others applies also to you.
> And the big stuff is precisely Samsung's problem:
> they are lagging far behind WD and also Hitachi
> with the 2 TB drives.
Meaning what, exactly? That Samsung has got its 1.5TB (current best
price) HDD sorted out but is having trouble making the larger ones? So
what? Doesn't mean anything about the reliability or performance of the
HDDs that it has got in full-scale series production.
> > Not seen any issues. Then buy a roll your own case so you know
> > what spec drive is inside.
>
> With HD size increasing so rapidly
> it is a pity to stop at 1.5 TB,
> which will be obsolete sooner,
I've got a 20MB HDD that's still useful.
1.5TB - obsolete sooner? But that's up to the user. No need to ditch
it if it's useful - and it will be very useful for a long time if the
Mac it's being used on isn't expected to hold a huge pile of videos.
That's all you have to do, really - well, I'd not be expecting to put a
significant video library on HDD until they get up to nearer 10TB.
Rowland.
(pining for a 1 Gbit/s internet connection)
J. J. Lodder
> J. J. Lodder <nos...@de-ster.demon.nl> wrote:
>
> [snip]
>
> > Internet folklore doesn't become more reliable by someone echoing it.
>
> That which applies to others applies also to you.
Right. A lot of opinions is still folklore,
not a statistic.
> > And the big stuff is precisely Samsung's problem:
> > they are lagging far behind WD and also Hitachi
> > with the 2 TB drives.
>
> Meaning what, exactly? That Samsung has got its 1.5TB (current best
> price) HDD sorted out but is having trouble making the larger ones? So
> what? Doesn't mean anything about the reliability or performance of the
> HDDs that it has got in full-scale series production.
>
> > > Not seen any issues. Then buy a roll your own case so you know
> > > what spec drive is inside.
> >
> > With HD size increasing so rapidly
> > it is a pity to stop at 1.5 TB,
> > which will be obsolete sooner,
>
> I've got a 20MB HDD that's still useful.
Useful, yet obsolete, for it can't be justified economically.
It consumes far too much power/bit.
> 1.5TB - obsolete sooner? But that's up to the user. No need to ditch
> it if it's useful - and it will be very useful for a long time if the
> Mac it's being used on isn't expected to hold a huge pile of videos.
> That's all you have to do, really - well, I'd not be expecting to put a
> significant video library on HDD until they get up to nearer 10TB.
Burning DVDs isn't realy worthwhle anymore.
Too much work, almost no savings.
suitable only for things you are unlikely ever to look at again,
yet don't want to throw away,
Jan
Andy Hewitt
> Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
>
> > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> >
> > [snip]
> >
> > > Internet folklore doesn't become more reliable by someone echoing it.
> >
> > That which applies to others applies also to you.
>
> Right. A lot of opinions is still folklore,
> not a statistic.
In this case, I am with Rowland. It's not folklore if you're dealing
with simple factual statements. Such as 'this drive failed after 6
months', or 'that drive is still working well after 12 years of constant
use'.
They're not opinions, they're facts. Get enough of it, and you have
information to make informed decisions about buying.
> > > And the big stuff is precisely Samsung's problem:
> > > they are lagging far behind WD and also Hitachi
> > > with the 2 TB drives.
> >
> > Meaning what, exactly? That Samsung has got its 1.5TB (current best
> > price) HDD sorted out but is having trouble making the larger ones? So
> > what? Doesn't mean anything about the reliability or performance of the
> > HDDs that it has got in full-scale series production.
> >
> > > > Not seen any issues. Then buy a roll your own case so you know
> > > > what spec drive is inside.
> > >
> > > With HD size increasing so rapidly
> > > it is a pity to stop at 1.5 TB,
> > > which will be obsolete sooner,
> >
> > I've got a 20MB HDD that's still useful.
>
> Useful, yet obsolete, for it can't be justified economically.
> It consumes far too much power/bit.
Maybe, and my older 4x4 is less economical than a new one, but it's not
economical to dispose of it while it's still functioning fine.
I really have a thing about this chucking out perfectly functional items
to replace them with newer, supposedly more economical ones. So many
factors are forgotten about.
> > 1.5TB - obsolete sooner? But that's up to the user. No need to ditch
> > it if it's useful - and it will be very useful for a long time if the
> > Mac it's being used on isn't expected to hold a huge pile of videos.
> > That's all you have to do, really - well, I'd not be expecting to put a
> > significant video library on HDD until they get up to nearer 10TB.
>
> Burning DVDs isn't realy worthwhle anymore.
> Too much work, almost no savings.
> suitable only for things you are unlikely ever to look at again,
> yet don't want to throw away,
I agree with that. I'd probably need a full pack of 50 to do a backup of
all my data here.
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
>
> > It's getting time to upgrade my Time Machine drive (again!), mostly in
> > preparation for future plans, and so I can incorporate my Aperture
> > library into it.
> >
> > I'm looking at going for a 1TB drive, which is far far more than I have
> > in data at the moment, so should last a fair while.
> >
> > Whats the latest in terms of reliability with hard drives?
>
> No idea, but 1.5TB was more bits per penny than 1TB last time I looked -
> which wasn't so long ago. I'd not buy a 1TB HDD for that reason.
>
> I've still never had a Seagate pack up on me.
Nor have I, but the latest ones at 1TB seem to be failing - going by
reports I find in various forums.
> [snip]
>
> > As this is for a TM drive, and speed isn't a great priority, what about
> > these new 5900rpm Seagate drives I'm seeing. They're at a very good
> > price (cost is a big factor for me), and look like they could be quieter
> > and maybe even more reliable
>
> Why more reliable?
Erm, that was actually the question I was asking! Are they in fact
*more* reliable, as the marketing speak suggests, or is that just
cobblers?
> Don't bet on the lower speed - it's all about what
> the mechanics are designed to do. 7200rpm and 5400rpm are in any case
> more or less the same sort of speed - erm, max wear is on start up and
> shut down, the rest of the time the main disc bearing gets hydrodynamic
> lubrication, so wear there is practically non-existent.
IIRC, didn't Google post the data about all their server drives a while
back, and found that there was neglible difference between continuous
running, and stop/start running.
Ah yes...
http://labs.google.com/papers/disk_failures.html
...only 2% more after 2 years in use.
> >(if you believe the tech blurb).
>
> You won't want to hear my suggestion of `get a twin HDD FW800 enclosure
> and run a RAID mirror with a pair of 1.5TB Seagate Barracudas'.
Easily done here, I am using the old FW800 bridge from my twin drive box
in my PC tower conversion. I'd only need two matching drives - which I
can't really afford right now.
J. J. Lodder
> J. J. Lodder <nos...@de-ster.demon.nl> wrote:
>
> > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> >
> > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > >
> > > [snip]
> > >
> > > > Internet folklore doesn't become more reliable by someone echoing it.
> > >
> > > That which applies to others applies also to you.
> >
> > Right. A lot of opinions is still folklore,
> > not a statistic.
>
> In this case, I am with Rowland. It's not folklore if you're dealing
> with simple factual statements. Such as 'this drive failed after 6
> months', or 'that drive is still working well after 12 years of constant
> use'.
Technology evolves so rapidly that what may have been wrong
with drives of more than three years ago is no longer relevant
for what you buy today.
> They're not opinions, they're facts. Get enough of it, and you have
> information to make informed decisions about buying.
No. For opinions to become more than folklore
you need to know what happened to a representative sample.
> > > > And the big stuff is precisely Samsung's problem:
> > > > they are lagging far behind WD and also Hitachi
> > > > with the 2 TB drives.
> > >
> > > Meaning what, exactly? That Samsung has got its 1.5TB (current best
> > > price) HDD sorted out but is having trouble making the larger ones? So
> > > what? Doesn't mean anything about the reliability or performance of the
> > > HDDs that it has got in full-scale series production.
> > >
> > > > > Not seen any issues. Then buy a roll your own case so you know
> > > > > what spec drive is inside.
> > > >
> > > > With HD size increasing so rapidly
> > > > it is a pity to stop at 1.5 TB,
> > > > which will be obsolete sooner,
> > >
> > > I've got a 20MB HDD that's still useful.
> >
> > Useful, yet obsolete, for it can't be justified economically.
> > It consumes far too much power/bit.
>
> Maybe, and my older 4x4 is less economical than a new one, but it's not
> economical to dispose of it while it's still functioning fine.
You wouldn't feel like that about a rambling car with a V8 engine,
I guess. (consuming for example 20l/100km)
> I really have a thing about this chucking out perfectly functional items
> to replace them with newer, supposedly more economical ones. So many
> factors are forgotten about.
You spend about a kWh in a few days for 20 MB storage,
when you can have 2 TB for the same kWh.
> > > 1.5TB - obsolete sooner? But that's up to the user. No need to ditch
> > > it if it's useful - and it will be very useful for a long time if the
> > > Mac it's being used on isn't expected to hold a huge pile of videos.
> > > That's all you have to do, really - well, I'd not be expecting to put a
> > > significant video library on HDD until they get up to nearer 10TB.
> >
> > Burning DVDs isn't realy worthwhle anymore.
> > Too much work, almost no savings.
> > suitable only for things you are unlikely ever to look at again,
> > yet don't want to throw away,
>
> I agree with that. I'd probably need a full pack of 50 to do a backup of
> all my data here.
And blueray is much worse per gigabyte, and still way too small,
even if the price drops dramatically.
Jan
Jaimie Vandenbergh
wrote:
>> Don't bet on the lower speed - it's all about what
>> the mechanics are designed to do. 7200rpm and 5400rpm are in any case
>> more or less the same sort of speed - erm, max wear is on start up and
>> shut down, the rest of the time the main disc bearing gets hydrodynamic
>> lubrication, so wear there is practically non-existent.
>
>IIRC, didn't Google post the data about all their server drives a while
>back, and found that there was neglible difference between continuous
>running, and stop/start running.
>
>Ah yes...
>
>http://labs.google.com/papers/disk_failures.html
>
>...only 2% more after 2 years in use.
Domestic disks use fluid bearings these days, there isn't a metal ball
in there at all any more. Started about 2002ish.
Cheers - Jaimie
--
My swerver room, my patch panels. By the time they figure out why none of the
ports on their floor box work anymore I'll be done, dusted and down the pub
with a pint of something brewed with yeast that was smarter than they are.
-- Matt S Trout, asr
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
>
> > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> >
> > > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> > >
> > > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > > >
> > > > [snip]
> > > >
> > > > > Internet folklore doesn't become more reliable by someone echoing it.
> > > >
> > > > That which applies to others applies also to you.
> > >
> > > Right. A lot of opinions is still folklore,
> > > not a statistic.
> >
> > In this case, I am with Rowland. It's not folklore if you're dealing
> > with simple factual statements. Such as 'this drive failed after 6
> > months', or 'that drive is still working well after 12 years of constant
> > use'.
>
> Technology evolves so rapidly that what may have been wrong
> with drives of more than three years ago is no longer relevant
> for what you buy today.
For sure, but you can still gather enough to see a 'trent' in the data,
and you'd need a crystal ball to know how any current drive will behave
in three years time.
However, again looking at factual data, the general trend tends to show
that drives fail mostly in the first three months, and then increasingly
so beyond two years.
> > They're not opinions, they're facts. Get enough of it, and you have
> > information to make informed decisions about buying.
>
> No. For opinions to become more than folklore
> you need to know what happened to a representative sample.
Yes, I think Google's report made that clear, and I'd say they did have
quite a representative sample.
These still aren't 'opinions' though, unless somebody says, 'that drive
looks crap, I think it'll fail early'. We can only judge these things on
the information we have available, and to some extent that manufacturers
can only do the same as well. If drives are failing earlier than your
rivals products, it could take a couple of years of data gatherign to
find that out, not to mention to what extent, and what the cause is.
[..]
> > > Useful, yet obsolete, for it can't be justified economically.
> > > It consumes far too much power/bit.
> >
> > Maybe, and my older 4x4 is less economical than a new one, but it's not
> > economical to dispose of it while it's still functioning fine.
>
> You wouldn't feel like that about a rambling car with a V8 engine,
> I guess. (consuming for example 20l/100km)
I would expect the owner of such a car to not give a toss about the
environment though.
> > I really have a thing about this chucking out perfectly functional items
> > to replace them with newer, supposedly more economical ones. So many
> > factors are forgotten about.
>
> You spend about a kWh in a few days for 20 MB storage,
> when you can have 2 TB for the same kWh.
No idea, I haven't checked.
> > > > 1.5TB - obsolete sooner? But that's up to the user. No need to ditch
> > > > it if it's useful - and it will be very useful for a long time if the
> > > > Mac it's being used on isn't expected to hold a huge pile of videos.
> > > > That's all you have to do, really - well, I'd not be expecting to put a
> > > > significant video library on HDD until they get up to nearer 10TB.
> > >
> > > Burning DVDs isn't realy worthwhle anymore.
> > > Too much work, almost no savings.
> > > suitable only for things you are unlikely ever to look at again,
> > > yet don't want to throw away,
> >
> > I agree with that. I'd probably need a full pack of 50 to do a backup of
> > all my data here.
>
> And blueray is much worse per gigabyte, and still way too small,
> even if the price drops dramatically.
Yes, Blueray is still far too expensive to be a viable option for now.
Even dual layer disks are marginal.
Andy Hewitt
> On Tue, 8 Dec 2009 21:23:50 +0000, thewil...@me.com (Andy Hewitt)
> wrote:
>
> >> Don't bet on the lower speed - it's all about what
> >> the mechanics are designed to do. 7200rpm and 5400rpm are in any case
> >> more or less the same sort of speed - erm, max wear is on start up and
> >> shut down, the rest of the time the main disc bearing gets hydrodynamic
> >> lubrication, so wear there is practically non-existent.
> >
> >IIRC, didn't Google post the data about all their server drives a while
> >back, and found that there was neglible difference between continuous
> >running, and stop/start running.
> >
> >Ah yes...
> >
> >http://labs.google.com/papers/disk_failures.html
> >
> >...only 2% more after 2 years in use.
>
> Domestic disks use fluid bearings these days, there isn't a metal ball
> in there at all any more. Started about 2002ish.
I didn't know they used ball-bearings anyway, I always thought they were
fluid types.
Jaimie Vandenbergh
wrote:
>Jaimie Vandenbergh <jai...@sometimes.sessile.org> wrote:
>
>> On Tue, 8 Dec 2009 21:23:50 +0000, thewil...@me.com (Andy Hewitt)
>> wrote:
>>
>> >> Don't bet on the lower speed - it's all about what
>> >> the mechanics are designed to do. 7200rpm and 5400rpm are in any case
>> >> more or less the same sort of speed - erm, max wear is on start up and
>> >> shut down, the rest of the time the main disc bearing gets hydrodynamic
>> >> lubrication, so wear there is practically non-existent.
>> >
>> >IIRC, didn't Google post the data about all their server drives a while
>> >back, and found that there was neglible difference between continuous
>> >running, and stop/start running.
>> >
>> >Ah yes...
>> >
>> >http://labs.google.com/papers/disk_failures.html
>> >
>> >...only 2% more after 2 years in use.
>>
>> Domestic disks use fluid bearings these days, there isn't a metal ball
>> in there at all any more. Started about 2002ish.
>
>I didn't know they used ball-bearings anyway, I always thought they were
>fluid types.
I've got a 40meg Rodime upstairs somewhere with a visible main
bearing! 1989 vintage, I think.
Cheers - Jaimie
--
"Some people think that noise abatement should be a higher priority
for ATC. I say safety is noise abatement. You have no idea how much
noise it makes to have a 737 fall out of the sky after an accident."
-- anonymous air traffic controller
J. J. Lodder
> On Tue, 8 Dec 2009 21:23:50 +0000, thewil...@me.com (Andy Hewitt)
> wrote:
>
> >> Don't bet on the lower speed - it's all about what
> >> the mechanics are designed to do. 7200rpm and 5400rpm are in any case
> >> more or less the same sort of speed - erm, max wear is on start up and
> >> shut down, the rest of the time the main disc bearing gets hydrodynamic
> >> lubrication, so wear there is practically non-existent.
> >
> >IIRC, didn't Google post the data about all their server drives a while
> >back, and found that there was neglible difference between continuous
> >running, and stop/start running.
> >
> >Ah yes...
> >
> >http://labs.google.com/papers/disk_failures.html
> >
> >...only 2% more after 2 years in use.
>
> Domestic disks use fluid bearings these days, there isn't a metal ball
> in there at all any more. Started about 2002ish.
Still have a ball bearing one. (WD 120 GB)
The whine is so bad (has been for for 5 years now)
that I use it only for automatic back-ups,
with automatic eject.
So power consumption is no reason to scrap it,
and the Cube it is attached to
won't take a bigger one anyway.
We'll see how long it whines on,
Jan
zoara
> > I really have a thing about this chucking out perfectly functional
> > items
> > to replace them with newer, supposedly more economical ones. So many
> > factors are forgotten about.
>
> You spend about a kWh in a few days for 20 MB storage,
> when you can have 2 TB for the same kWh.
Manufacturing and disposal costs are also a factor.
-zoara-
--
email: nettid1 at fastmail dot fm
zoara
> J. J. Lodder <nos...@de-ster.demon.nl> wrote:
>
> > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid>
> > wrote:
> >
> > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > >
> > > [snip]
> > >
> > > > Internet folklore doesn't become more reliable by someone
> > > > echoing it.
> > >
> > > That which applies to others applies also to you.
> >
> > Right. A lot of opinions is still folklore,
> > not a statistic.
>
> In this case, I am with Rowland. It's not folklore if you're dealing
> with simple factual statements. Such as 'this drive failed after 6
> months', or 'that drive is still working well after 12 years of
> constant
> use'.
> They're not opinions, they're facts. Get enough of it, and you have
> information to make informed decisions about buying.
And there's your problem. How do you know that the people repeating
these figures are not only noticing those effects that they expect to
see? It's called confirmation bias, and is one of the reasons people
believe things like homeopathy; they notice the times the patient
"quickly" gets better but not when it takes ages, or when they get
worse.
It's the same with hard disks. If you're predisposed to the feeling that
(say) Seagates are unreliable, every time you get a Seagate fail you
will notice it as proof that Seagates are crap, and every time something
else fails you will dismiss it as an anomoly. That applies whether
you're a home user dealing with a disk or two a year, or a business
dealing with hundreds a year.
The only way to determine this sort of thing for sure is with a proper
analysis, where confirmation bias is controlled. Otherwise it's all just
folklore.
That said, I don't think it's worth wasting any time doing such a thing.
There are as many different opinions as to which manufacturers are
reliable as there are manufacturers, so it's reasonable to assume that
they're all pretty much the same and - unless you're dealing with
massive volumes - the difference between reliability is so small that
you won't notice it.
> > > I've got a 20MB HDD that's still useful.
> >
> > Useful, yet obsolete, for it can't be justified economically.
> > It consumes far too much power/bit.
>
> Maybe, and my older 4x4 is less economical than a new one, but it's
> not
> economical to dispose of it while it's still functioning fine.
>
> I really have a thing about this chucking out perfectly functional
> items
> to replace them with newer, supposedly more economical ones. So many
> factors are forgotten about.
I agree with this completely. Remember when everyone started replacing
all their lightbulbs with CFLs? They didn't realise that there was a
difference between replacing a bulb when it blows and replacing bulbs
which were currently functional; I had a few friends boast how
environmentally friendly they were because they'd just "replaced every
bulb in the house with energy saving bulbs". Inevitably they were the
ones who went and replaced all those CFLs with incandescants when they
decided the light was crappy (so doubly environmentally unsound, then).
Having said that, I don't think I could find a 20MB hard disk useful in
any circumstance. If that's all the space I needed, it would be easier
to make space on one of my SD cards or USB sticks than to dig out a 20MB
disk from wherever I'd left it last time I used it...
Pd
> The plural of anecdote is not data.
I think you'll find that is entirely dependent on relativistic
subjectivity. If (using the general "I" for the sake of argument) I see
one instance, whatever I conclude is the truth. If you see a million
contradictory instances, you're just wrong. And insulting, with it.
--
Pd
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
> > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
[..]
> > In this case, I am with Rowland. It's not folklore if you're dealing
> > with simple factual statements. Such as 'this drive failed after 6
> > months', or 'that drive is still working well after 12 years of
> > constant
> > use'.
>
> The plural of anecdote is not data.
Where did you get anecdote from?
> > They're not opinions, they're facts. Get enough of it, and you have
> > information to make informed decisions about buying.
>
> And there's your problem. How do you know that the people repeating
> these figures are not only noticing those effects that they expect to
> see? It's called confirmation bias, and is one of the reasons people
> believe things like homeopathy; they notice the times the patient
> "quickly" gets better but not when it takes ages, or when they get
> worse.
It's certainly true that we'll generally mainly hear about only failed
units. Over a period of time the discrepancies should even out though,
and with enough data, you can get a good idea what's what. It's enough
to make decisions about purchases IMHO. If you had scanned Google, say,
and found dozens of reports of a particular drive failing prematurely,
would *you* still buy it? For sure there may be bias against, or for,
particular brands, but even that should even out over time.
All you can look for are 'trends', and use this for an informed decision
on buying things. It's not to say that buying one brand *will* result in
a failure, or another *won't*. It's just about minimising the risks, and
if a bit of research can help that, why not?
> It's the same with hard disks. If you're predisposed to the feeling that
> (say) Seagates are unreliable, every time you get a Seagate fail you
> will notice it as proof that Seagates are crap, and every time something
> else fails you will dismiss it as an anomoly. That applies whether
> you're a home user dealing with a disk or two a year, or a business
> dealing with hundreds a year.
Not at all, I know that as a rule Seagates have been reliable over a
long period, and using my own experience of drives, I have found this to
be so. Likewise I have used Hitachi, which have been reliable too.
Reading what others say matches my own experience. I also note that
there's some reliability issue with the Samsung F1 Spinpoint, but not so
with the F2 and F3. AFAIK, there's only reports of the latest 1TB
Seagates being unreliable, and I'm sure they won't be for too long
before they fix it.
> The only way to determine this sort of thing for sure is with a proper
> analysis, where confirmation bias is controlled. Otherwise it's all just
> folklore.
I don't know how this can be biased, unless you only take the 'this is
crap' statements. I certainly wouldn't make a decision based on a low
number of comments, but take note of just the fact of failure. If a
drive has failed, and is failing in larger numbers than others, where's
the bias? Of course it doesn't become accurate if not all failures are
reported, but as a rule many consumers are more than happy to have a
moan about a failed unit.
> That said, I don't think it's worth wasting any time doing such a thing.
> There are as many different opinions as to which manufacturers are
> reliable as there are manufacturers, so it's reasonable to assume that
> they're all pretty much the same and - unless you're dealing with
> massive volumes - the difference between reliability is so small that
> you won't notice it.
I don't use opinions, I use information based on actual failures. If a
drive has failed, then it's failed, there's no 'opinion' about that.
> > > > I've got a 20MB HDD that's still useful.
> > >
> > > Useful, yet obsolete, for it can't be justified economically.
> > > It consumes far too much power/bit.
> >
> > Maybe, and my older 4x4 is less economical than a new one, but it's
> > not
> > economical to dispose of it while it's still functioning fine.
> >
> > I really have a thing about this chucking out perfectly functional
> > items
> > to replace them with newer, supposedly more economical ones. So many
> > factors are forgotten about.
>
> I agree with this completely. Remember when everyone started replacing
> all their lightbulbs with CFLs? They didn't realise that there was a
> difference between replacing a bulb when it blows and replacing bulbs
> which were currently functional; I had a few friends boast how
> environmentally friendly they were because they'd just "replaced every
> bulb in the house with energy saving bulbs". Inevitably they were the
> ones who went and replaced all those CFLs with incandescants when they
> decided the light was crappy (so doubly environmentally unsound, then).
Aye, that's one that gets me too. Indeed there are many cases where a
CFL is not ideal anyway (such as short duration usage).
> Having said that, I don't think I could find a 20MB hard disk useful in
> any circumstance. If that's all the space I needed, it would be easier
> to make space on one of my SD cards or USB sticks than to dig out a 20MB
> disk from wherever I'd left it last time I used it...
Depends, it's easy to make comparisons based on our own needs, but there
must be a few users out there still on older equipment - I still have a
Mac SE in my loft, which works perfectly, along with its 6MB drive. It
could be possible to locate an old Apple printer, and get some word
processing donw with it.
Jaimie Vandenbergh
wrote:
>Not at all, I know that as a rule Seagates have been reliable over a
>long period, and using my own experience of drives, I have found this to
>be so. Likewise I have used Hitachi, which have been reliable too.
>Reading what others say matches my own experience. I also note that
>there's some reliability issue with the Samsung F1 Spinpoint, but not so
>with the F2 and F3. AFAIK, there's only reports of the latest 1TB
>Seagates being unreliable, and I'm sure they won't be for too long
>before they fix it.
But you're hitting exactly the point that's being made. "As a rule"
maps directly to "in my experience, plus the feel I get from the
crowd's comments". It's all anecdotes, because there's no control for
the source data - you don't have success rates, uptimes, birth/death
dates, drive types, use scenarios, anything. Just notes that "I had
one of these and it died".
This clearly isn't useful for data, because (for example) I have
different views. For NAS stuffing purposes I was watching 1Tb+ drives
a lot this year - the mess Seagate made with their 1, 1.5 and 2Tb
drive firmwares was a total crapfest, which I guess you've missed. I
won't be touching their 3.5" drives until a later series, but their
2.5" drives have had no issues.
Hitachi I've not heard anything bad or good about recently, but last
time I was working with laptops (IBMs, 2000-2006) on a mass basis we
had about 50% failure rate of the 2.5" drives over the three year
laptop replacement cycle.
All anecdotes.
I still use them to inform my own purchasing (weighted against
availability of preferred makes), but I don't consider those opinions
to be statistically valid.
Cheers - Jaimie
--
"On two occasions I have been asked [by members of Parliament],
'Pray, Mr. Babbage, if you put into the machine wrong figures, will
the right answers come out?' I am not able rightly to apprehend the
kind of confusion of ideas that could provoke such a question."
J. J. Lodder
> Andy Hewitt <thewil...@me.com> wrote:
> > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> >
> > > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid>
> > > wrote:
> > >
> > > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > > >
> > > > [snip]
> > > >
> > > > > Internet folklore doesn't become more reliable by someone
> > > > > echoing it.
> > > >
> > > > That which applies to others applies also to you.
> > >
> > > Right. A lot of opinions is still folklore,
> > > not a statistic.
> >
> > In this case, I am with Rowland. It's not folklore if you're dealing
> > with simple factual statements. Such as 'this drive failed after 6
> > months', or 'that drive is still working well after 12 years of
> > constant
> > use'.
>
> The plural of anecdote is not data.
Gosh. I wish I had said that,
Jan
--
"You will Oscar, you will."
Rowland McDonnell
> J. J. Lodder <nos...@de-ster.demon.nl> wrote:
>
> > Andy Hewitt <thewil...@me.com> wrote:
[snip]
> > > > Useful, yet obsolete, for it can't be justified economically.
> > > > It consumes far too much power/bit.
> > >
> > > Maybe, and my older 4x4 is less economical than a new one, but it's not
> > > economical to dispose of it while it's still functioning fine.
> >
> > You wouldn't feel like that about a rambling car with a V8 engine,
> > I guess. (consuming for example 20l/100km)
>
> I would expect the owner of such a car to not give a toss about the
> environment though.
[snip]
A fuel-inefficient V8 is no environmental hazard, simply expensive. NO
SINGLE ANYTHING done by a human is a threat to the maintenance of a
global environment suitable for humans (which is what we need to worry
about).
Coal seams burning out of control due to human incompetence is injecting
more greenhouse gas into the air than all the old Jags in the world
would manage if they all ran 24 hours a day flat-out non-stop[1]. Then
there's the cow farts. Big `gas-guzzling' cars aren't simply not a
problem until you have to fill up with fuel, or overtake 'em on a narrow
street (i.e., anywhere near where I live[2]), park in a European city,
do a U-turn, and so on.
The great thing about the fuel tax increases here is that it's put most
of the really big vehicles off the road so I can make better progress on
my motorcycle.
The only reason big fancy cars are targetted as `environmental threat
number 1' is that yer average greenie hates rich bastards in flash cars
and wants to put 'em down - nothing more than that. I dislike rich
flashy pushy bastards in big fancy cars too, but that's because I hate
rich flashy pushy bastards. But I'm not a hypocrite and I'll call 'em
bastards to their face because they're bastards and I won't dishonestly
pretend that their choice of motor vehicle is causing a signifigant
increase in atmospheric CO_2, 'cos it ain't.
Rowland.
[1] But maybe not all the old Fords. I'd have to do some
calculations...
[2] Aside from the motorways, I suppose. Plural, yes, we are well
supplied with 'em.
Rowland McDonnell
> Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
>
> > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> >
> > [snip]
> >
> > > Internet folklore doesn't become more reliable by someone echoing it.
> >
> > That which applies to others applies also to you.
>
> Right. A lot of opinions is still folklore,
> not a statistic.
And a lot of measurements are statistics. What is your point?
[snip]
> > > > Not seen any issues. Then buy a roll your own case so you know
> > > > what spec drive is inside.
> > >
> > > With HD size increasing so rapidly
> > > it is a pity to stop at 1.5 TB,
> > > which will be obsolete sooner,
> >
> > I've got a 20MB HDD that's still useful.
>
> Useful, yet obsolete, for it can't be justified economically.
> It consumes far too much power/bit.
That is irrelevant to any justification - what matters is `what the
system it's part of can do per <measure>' for ALL relevant values of
<measure> including the aesthetic ones.
> > 1.5TB - obsolete sooner? But that's up to the user. No need to ditch
> > it if it's useful - and it will be very useful for a long time if the
> > Mac it's being used on isn't expected to hold a huge pile of videos.
> > That's all you have to do, really - well, I'd not be expecting to put a
> > significant video library on HDD until they get up to nearer 10TB.
>
> Burning DVDs isn't realy worthwhle anymore.
> Too much work, almost no savings.
You forget `backup on removeable media'. It occurs to me that I can't
afford to put all my video DVDs on HDD.
> suitable only for things you are unlikely ever to look at again,
> yet don't want to throw away,
And also for data transfer jobs. The paranoid pay attention to the
point that a DVD in the post is less likely to be subjected to snooping
than anything sent over the wires.
Rowland.
Rowland McDonnell
[snip]
> Domestic disks use fluid bearings these days, there isn't a metal ball
> in there at all any more. Started about 2002ish.
I was under the impression that they'd all been using fluid dynamic
bearings more or less since the dawn of time (such bearings/lubrication
predating HDDs, IIRC). What else other than a plain bearing using fluid
dynamic lubrication could provide the smoothness or reliability needed?
I can't think of anything.
But I was also under the impression that they all use a ball for the
bearing to settle down on to when the platter's aren't spinning - lifts
off when running, I gather (although I might have mis-understood that
bit).
If there's no ball, what do they rest on?
Rowland McDonnell
> Andy Hewitt <thewil...@me.com> wrote:
>
> > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> >
> > > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> > >
> > > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > > >
> > > > [snip]
> > > >
> > > > > Internet folklore doesn't become more reliable by someone echoing it.
> > > >
> > > > That which applies to others applies also to you.
> > >
> > > Right. A lot of opinions is still folklore,
> > > not a statistic.
> >
> > In this case, I am with Rowland. It's not folklore if you're dealing
> > with simple factual statements. Such as 'this drive failed after 6
> > months', or 'that drive is still working well after 12 years of constant
> > use'.
>
> Technology evolves so rapidly that what may have been wrong
> with drives of more than three years ago is no longer relevant
> for what you buy today.
>
> > They're not opinions, they're facts. Get enough of it, and you have
> > information to make informed decisions about buying.
>
> No. For opinions to become more than folklore
> you need to know what happened to a representative sample.
You are dishonestly misrepresenting measurements as mere opinion.
You are also missing the point that in engineering, it's hard to be sure
that one has a representative sample and the only thing to do is gather
more and more data making the best guess from the accumulated data that
one can.
Doing as you seem to suggest: totally ignoring historical data until
it's at 99% certainty (or whatever) as proven by a full-on statistical
analysis - well, that's crazy and not how engineering gets done.
> > > > > And the big stuff is precisely Samsung's problem:
> > > > > they are lagging far behind WD and also Hitachi
> > > > > with the 2 TB drives.
> > > >
> > > > Meaning what, exactly? That Samsung has got its 1.5TB (current best
> > > > price) HDD sorted out but is having trouble making the larger ones? So
> > > > what? Doesn't mean anything about the reliability or performance of the
> > > > HDDs that it has got in full-scale series production.
> > > >
> > > > > > Not seen any issues. Then buy a roll your own case so you know
> > > > > > what spec drive is inside.
> > > > >
> > > > > With HD size increasing so rapidly
> > > > > it is a pity to stop at 1.5 TB,
> > > > > which will be obsolete sooner,
> > > >
> > > > I've got a 20MB HDD that's still useful.
> > >
> > > Useful, yet obsolete, for it can't be justified economically.
> > > It consumes far too much power/bit.
> >
> > Maybe, and my older 4x4 is less economical than a new one, but it's not
> > economical to dispose of it while it's still functioning fine.
>
> You wouldn't feel like that about a rambling car with a V8 engine,
> I guess. (consuming for example 20l/100km)
What's that in sane units? Consumption per distance, please.
(20l/100km works out to have the units of area. That has a meaning,
believe it or not: the volume of a pipe of the specified cross-sectional
area swept by the passage of the vehicle in question will supply exactly
the right amount of fuel as you go along. Now you try to visualize the
meaning of miles per gallon in the same way - tricky, eh? So: which
units are the sane ones?)
Anyway, you're quite right, better off with an efficient 4 cylinder car
like one of these:
<http://www.ultimatecarpage.com/images/events/lemans2002/1930_bentley_bl
ower.jpg>
(I saw one of them - blower stuck on the front an' all - rattling and
clattering along the road one day when I was walking to college, years
ago. Astonishing! I saw what Enrico Bugatti meant when he called 'em
the fastest lorries on the road.)
> > I really have a thing about this chucking out perfectly functional items
> > to replace them with newer, supposedly more economical ones. So many
> > factors are forgotten about.
>
> You spend about a kWh in a few days for 20 MB storage,
> when you can have 2 TB for the same kWh.
Modern HDDs consume more juice than my old 20MB job. The storage space
doesn't matter - what matters is what you're doing with it and how much
energy it consumes. Not energy per bit, but total energy consumption
full stop.
A 20MB HDD is a good idea if it's still useful - and it's more
economical than buying a new HDD. Lower power and of course if you
already own one, no purchase cost to consider. And if it's only used
occasionally, its energy consumption will be lower still.
[snip]
Rowland.
Rowland McDonnell
> Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
>
> > Andy Hewitt <thewil...@me.com> wrote:
> >
> > > It's getting time to upgrade my Time Machine drive (again!), mostly in
> > > preparation for future plans, and so I can incorporate my Aperture
> > > library into it.
> > >
> > > I'm looking at going for a 1TB drive, which is far far more than I have
> > > in data at the moment, so should last a fair while.
> > >
> > > Whats the latest in terms of reliability with hard drives?
> >
> > No idea, but 1.5TB was more bits per penny than 1TB last time I looked -
> > which wasn't so long ago. I'd not buy a 1TB HDD for that reason.
> >
> > I've still never had a Seagate pack up on me.
>
> Nor have I, but the latest ones at 1TB seem to be failing - going by
> reports I find in various forums.
Righto. Pooey (?). Got a couple of 1.5TB Barracudas - didn't seem to
be any reliability problems when we looked (speed problems, firmware
issue, got sorted).
Gonna get another 1.5TB Barracuda. Got the current pair in RAID mirror.
Wanna spare disc...
> > [snip]
> >
> > > As this is for a TM drive, and speed isn't a great priority, what about
> > > these new 5900rpm Seagate drives I'm seeing. They're at a very good
> > > price (cost is a big factor for me), and look like they could be quieter
> > > and maybe even more reliable
> >
> > Why more reliable?
>
> Erm, that was actually the question I was asking! Are they in fact
> *more* reliable, as the marketing speak suggests, or is that just
> cobblers?
Ah. Oh. Dunno. :-)
> > Don't bet on the lower speed - it's all about what
> > the mechanics are designed to do. 7200rpm and 5400rpm are in any case
> > more or less the same sort of speed - erm, max wear is on start up and
> > shut down, the rest of the time the main disc bearing gets hydrodynamic
> > lubrication, so wear there is practically non-existent.
>
> IIRC, didn't Google post the data about all their server drives a while
> back, and found that there was neglible difference between continuous
> running, and stop/start running.
Hmm. Coo. Blimey. Well, well, well. That did not used to be the
case.
> Ah yes...
>
> http://labs.google.com/papers/disk_failures.html
>
> ...only 2% more after 2 years in use.
2% sounds like it's probably statistically significant given the sample
size, and so not negligible. Economically trivial, perhaps - and
definitely less than I was expecting.
Hmm.
Coo.
The more I hear about HDD reliability, the stranger it seems.
> > >(if you believe the tech blurb).
> >
> > You won't want to hear my suggestion of `get a twin HDD FW800 enclosure
> > and run a RAID mirror with a pair of 1.5TB Seagate Barracudas'.
>
> Easily done here, I am using the old FW800 bridge from my twin drive box
> in my PC tower conversion. I'd only need two matching drives - which I
> can't really afford right now.
Quite...
Rowland McDonnell
> thewil...@me.com (Andy Hewitt) rote:
[snip]
> >> Domestic disks use fluid bearings these days, there isn't a metal ball
> >> in there at all any more. Started about 2002ish.
> >
> >I didn't know they used ball-bearings anyway, I always thought they were
> >fluid types.
>
> I've got a 40meg Rodime upstairs somewhere with a visible main
> bearing! 1989 vintage, I think.
Yeah, but what sort of bearing?
Rowland McDonnell
> Andy Hewitt <thewil...@me.com> wrote:
[snip]
> > In this case, I am with Rowland. It's not folklore if you're dealing
> > with simple factual statements. Such as 'this drive failed after 6
> > months', or 'that drive is still working well after 12 years of
> > constant
> > use'.
>
> The plural of anecdote is not data.
Trite putdowns are not scientific reasoning - nor are they often right.
[snip]
Rowland.
J. J. Lodder
> J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > Andy Hewitt <thewil...@me.com> wrote:
>
> > > I really have a thing about this chucking out perfectly functional
> > > items
> > > to replace them with newer, supposedly more economical ones. So many
> > > factors are forgotten about.
> >
> > You spend about a kWh in a few days for 20 MB storage,
> > when you can have 2 TB for the same kWh.
>
> Manufacturing and disposal costs are also a factor.
HDs should (and can be) be recycled.
And lifetime energy cost of a HD
(if used full time)
is greater than selling price,
Jan
Rowland McDonnell
[snip]
> All anecdotes.
And all data, too. All data is reports: all data is anecdotes.
Scientifically valid data is anecdotes that lots of people come out
with, all agreeing with each other.
> I still use them to inform my own purchasing (weighted against
> availability of preferred makes), but I don't consider those opinions
> to be statistically valid.
Not safe to make detailed specific predictions, but safe to make rough
general predictions.
Do you even know the difference between `statistically valid' and `good
enough for rough predictions'? Anyone who's worked with reality knows
how to make predictions from inadequate data and if you've worked at it
you can make them pretty reliably (it's one of the things that makes us
superior to computers[1]). Me and Andy both have such experience, so we
take the same view of this sort of thing. We can do it - can you?
Rowland.
[1] Caveat: most of the time, most people are very bad at making
decisions based on past experience. It takes a special kind of attitude
and a lot of practice to get competent at it. Being slightly autistic
helps, as it happens - for things that don't involve personal
relationships, I should add.
Rowland McDonnell
> zoara <me...@privacy.net> wrote:
>
> > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > > Andy Hewitt <thewil...@me.com> wrote:
> >
> > > > I really have a thing about this chucking out perfectly functional
> > > > items
> > > > to replace them with newer, supposedly more economical ones. So many
> > > > factors are forgotten about.
> > >
> > > You spend about a kWh in a few days for 20 MB storage,
> > > when you can have 2 TB for the same kWh.
> >
> > Manufacturing and disposal costs are also a factor.
>
> HDs should (and can be) be recycled.
Which is a cost, as is making the things in the first place. Energy
gets used in both cases - energy which cannot (according to modern
conventional consensus scientific ideas) ever be recovered. The
universe is running down, and everything thrown away increases entropy.
We all bear guilt for accelerating the death of the entire universe.
I don't intend to worry about it.
> And lifetime energy cost of a HD
> (if used full time)
> is greater than selling price,
If it's not used full time, however - as is normal for most domestic
appliances...
Rowland.
Chris Ridd
> zoara <me...@privacy.net> wrote:
>
>> The plural of anecdote is not data.
>
> Gosh. I wish I had said that,
It's a great phrase. I saw someone use it on zfs-discuss, so sadly it
isn't a Zoara original :-(
--
Chris
Rowland McDonnell
> J. J. Lodder said:
>
> > zoara <me...@privacy.net> wrote:
> >
> >> The plural of anecdote is not data.
> >
> > Gosh. I wish I had said that,
>
> It's a great phrase. I saw someone use it on zfs-discuss, so sadly it
> isn't a Zoara original :-(
It's also counter-factual. Data is nothing but anecdotes - a collection
of reports is what data *IS*.
Citizen Jimserac
> > > wrote:
>
> > > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
>
> > > > [snip]
>
> > > > > Internet folklore doesn't become more reliable by someone
> > > > > echoing it.
>
> > > > That which applies to others applies also to you.
>
> > > Right. A lot of opinions is still folklore,
> > > not a statistic.
>
> > In this case, I am with Rowland. It's not folklore if you're dealing
> > with simple factual statements. Such as 'this drive failed after 6
> > months', or 'that drive is still working well after 12 years of
> > constant
> > use'.
>
> The plural of anecdote is not data.
>
> > They're not opinions, they're facts. Get enough of it, and you have
> > information to make informed decisions about buying.
>
> And there's your problem. How do you know that the people repeating
> these figures are not only noticing those effects that they expect to
> see? It's called confirmation bias, and is one of the reasons people
> believe things like homeopathy; they notice the times the patient
> "quickly" gets better but not when it takes ages, or when they get
> worse.
Actually the main reason Homeopathy is used is because it works, has
been working for a couple of hundred years and will continue to work.
And that means really works, not placebo, not trickery nor none of the
rest of the nonsense that is written about it.
Homeopathy has been adapted and used by some very hard thinking and
sceptical MD's and other health professionals and experts. You think
these folk don't know a cure when they see one or can't tell the
difference if an illness gradually resolved itself of its own accord
or was improved by Homeopathy? Think they can't tell the difference
between a placebo effect and a real cure? Think again.
Think that high dilutions with ALL the molecules of the curative
substance diluted away can't have a biological effect? Think again
(Ennis, Inflammation Research, vol 53, p181 - yes, it's been repeated
and confirmed by others, yes the biological effect stimulated acted as
though the missing molecules were still there, no the BBC's
documentary did not repeat her experiment and did not "refute it").
>
> It's the same with hard disks. If you're predisposed to the feeling that
> (say) Seagates are unreliable, every time you get a Seagate fail you
> will notice it as proof that Seagates are crap, and every time something
> else fails you will dismiss it as an anomoly. That applies whether
> you're a home user dealing with a disk or two a year, or a business
> dealing with hundreds a year.
>
> The only way to determine this sort of thing for sure is with a proper
> analysis, where confirmation bias is controlled. Otherwise it's all just
> folklore.
Well true but the folklore does have some signficance beyond a certain
point.
I'm sure the engineers and their managers at Iomega were VERY
comfortable with all their statistical studies showing their zip drive
ready for production,
but a huge number of users of their ZIP drive product eventually
discovered what the "click of death"
was - some after a good long period of using the product and others
after only a short time, perhaps less than year. The accumulated
opinion of people with these experiences IS significant. I know I'm
never going to buy another Iomega product again and yet I also know
that my experiences with just one item from their product line is no
proof that it or any other item design was flawed. But, if you look
at the history of other Iomega products, a pattern begins to
emerge....
>
> That said, I don't think it's worth wasting any time doing such a thing.
> There are as many different opinions as to which manufacturers are
> reliable as there are manufacturers, so it's reasonable to assume that
> they're all pretty much the same and - unless you're dealing with
> massive volumes - the difference between reliability is so small that
> you won't notice it.
No, here I must disagree. They are NOT the same and gradually a
consensus of opinion will emerge,
sometimes even though not a majority and not at all necessarily a
"proof" that one or the other manufacturer
is superiour. Often such opinions are based more on intangibles
such as customer service, documentation, ease of replacement of a
flawed product and, most importantly, acknowledgement of a flawed
product.
By way of example, I had a bad experience with an HP "All in One"
Desktop printer - it had a TERRRIBLE DESIGN FLAW in which it would
REFUSE TO SCAN A DOCUMENT in the built in scanner if one was out of
ink for the PRINTER!! Think about the stupidity of such a design.
Of course this was mentioned nowhere in the documentation. Of course
it probably saved them some manufacturing costs to do such a think but
the stupidity of it is simply astonishing. On the basis of this
experience I promptly added HP to my AVOID list. Now imagine how many
thousand of people eventually encountered the same stupidity and
reacted the same way.
Citizen Jimserac
zoara
> zoara <me...@privacy.net> wrote:
>
> > Andy Hewitt <thewil...@me.com> wrote:
> > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> [..]
Jamie's answered a lot of this better than I could, so just some quick
points...
> > > In this case, I am with Rowland. It's not folklore if you're
> > > dealing
> > > with simple factual statements. Such as 'this drive failed after 6
> > > months', or 'that drive is still working well after 12 years of
> > > constant
> > > use'.
> >
> > The plural of anecdote is not data.
>
> Where did you get anecdote from?
>
> > > They're not opinions, they're facts. Get enough of it, and you
> > > have
> > > information to make informed decisions about buying.
> >
> > And there's your problem. How do you know that the people repeating
> > these figures are not only noticing those effects that they expect
> > to
> > see? It's called confirmation bias, and is one of the reasons people
> > believe things like homeopathy; they notice the times the patient
> > "quickly" gets better but not when it takes ages, or when they get
> > worse.
>
> It's certainly true that we'll generally mainly hear about only failed
> units.
Not just that - we'll hear about the "expected" failures more than the
unexpected ones. Those that confirm the reporter's feelings that brand X
is the unreliable one. If two brand X drives fail and one brand Y drive
fails, it's most likely the brand X failures will be reported and the
brand Y won't.
> Over a period of time the discrepancies should even out though,
> and with enough data, you can get a good idea what's what. It's enough
> to make decisions about purchases IMHO. If you had scanned Google,
> say,
> and found dozens of reports of a particular drive failing prematurely,
> would *you* still buy it?
No, but I am not immune to confirmation bias or the misjudgement of
risk.
And Googling any brand of drive will give loads of reported failures...
I just tend to buy on price and reported loudness these days.
> For sure there may be bias against, or for,
> particular brands, but even that should even out over time.
That's not what "confirmation bias" means. It means that people tend to
put more weight on events that confirm their already-held beliefs. If
someone believes Americans are stupid, they'll notice more instances of
when Americans say something stupid than when Americans say something
intelligent. But it works the other way too - if someone believes
Americans are intelligent, they'll notice the intelligent comments more
than the stupid ones.
[....]
> > It's the same with hard disks. If you're predisposed to the feeling
> > that
> > (say) Seagates are unreliable, every time you get a Seagate fail you
> > will notice it as proof that Seagates are crap, and every time
> > something
> > else fails you will dismiss it as an anomoly. That applies whether
> > you're a home user dealing with a disk or two a year, or a business
> > dealing with hundreds a year.
>
> Not at all, I know that as a rule Seagates have been reliable over a
> long period,
I was using "Seagate" as an example, not a specific. Replace with a
brand that you consider unreliable if you want :) My comment was
explaining confirmation bias, not saying which brand is unreliable.
> > That said, I don't think it's worth wasting any time doing such a
> > thing.
> > There are as many different opinions as to which manufacturers are
> > reliable as there are manufacturers, so it's reasonable to assume
> > that
> > they're all pretty much the same and - unless you're dealing with
> > massive volumes - the difference between reliability is so small
> > that
> > you won't notice it.
>
> I don't use opinions, I use information based on actual failures. If a
> drive has failed, then it's failed, there's no 'opinion' about that.
Well, obviously. But "Brand X is unreliable" is an opinion, even if it's
followed up with "because I've had loads of their drives fail".
Look at me and my repeated iBook / MacBook failures. I'm now of the
opinion that the "consumer-level" laptops from Apple are absolute junk
and best avoided; but that opinion is purely anecdotal and coloured by
confirmation bias. If Apple's consumer laptops were failing on the
levels I was experiencing, they'd be out of business by now. But they
aren't; I just had bad luck. By the end of it I was unsurprised when a
machine went, as my confirmation bias said they were junk; but I bet
failure rates of consumer laptops are so close to the Pro laptops that
it can't be measured outside of Apple's recall / repair figures.
[...]
> > Having said that, I don't think I could find a 20MB hard disk useful
> > in
> > any circumstance. If that's all the space I needed, it would be
> > easier
> > to make space on one of my SD cards or USB sticks than to dig out a
> > 20MB
> > disk from wherever I'd left it last time I used it...
>
> Depends, it's easy to make comparisons based on our own needs, but
> there
> must be a few users out there still on older equipment - I still have
> a
> Mac SE in my loft, which works perfectly, along with its 6MB drive. It
> could be possible to locate an old Apple printer, and get some word
> processing donw with it.
Of course, I know other people could make use of it. But I don't think I
could.
zoara
> Chris Ridd <chri...@mac.com> wrote:
>
> > J. J. Lodder said:
> >
> > > zoara <me...@privacy.net> wrote:
> > >
> > >> The plural of anecdote is not data.
> > >
> > > Gosh. I wish I had said that,
> >
> > It's a great phrase. I saw someone use it on zfs-discuss, so sadly
> > it
> > isn't a Zoara original :-(
>
> It's also counter-factual. Data is nothing but anecdotes - a
> collection
> of reports is what data *IS*.
Pithiness is inversely proportional to accuracy.
zoara
Oh, it's definitely not. Can't remember where I first heard it (possibly
Ben Goldacre?), but it's a (reasonably) popular phrase. And lovely, too.
zoara
Zing!
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
>
> > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
[..]
> > > I've still never had a Seagate pack up on me.
> >
> > Nor have I, but the latest ones at 1TB seem to be failing - going by
> > reports I find in various forums.
>
> Righto. Pooey (?). Got a couple of 1.5TB Barracudas - didn't seem to
> be any reliability problems when we looked (speed problems, firmware
> issue, got sorted).
Indeed no, I believe it's only the 1TB ones that are affected.
> Gonna get another 1.5TB Barracuda. Got the current pair in RAID mirror.
> Wanna spare disc...
Well, having started this thread, and had a think, I'm going to wait for
now, thanks. I'm actually waiting for things to happen, which means I
can get sorted a bit better financially, and maybe get a new iMac - that
may change how I need my setup - although I'm still possibly thinking of
minimising costs, but I'd rather like a RAID mirror setup for my
backups.
[..]
> > > Why more reliable?
> >
> > Erm, that was actually the question I was asking! Are they in fact
> > *more* reliable, as the marketing speak suggests, or is that just
> > cobblers?
>
> Ah. Oh. Dunno. :-)
Of course they also say they're quiter and more energy efficient, but I
suspect there are other ways to achieve that.
> > > Don't bet on the lower speed - it's all about what
> > > the mechanics are designed to do. 7200rpm and 5400rpm are in any case
> > > more or less the same sort of speed - erm, max wear is on start up and
> > > shut down, the rest of the time the main disc bearing gets hydrodynamic
> > > lubrication, so wear there is practically non-existent.
> >
> > IIRC, didn't Google post the data about all their server drives a while
> > back, and found that there was neglible difference between continuous
> > running, and stop/start running.
>
> Hmm. Coo. Blimey. Well, well, well. That did not used to be the
> case.
>
> > Ah yes...
> >
> > http://labs.google.com/papers/disk_failures.html
> >
> > ...only 2% more after 2 years in use.
>
> 2% sounds like it's probably statistically significant given the sample
> size, and so not negligible. Economically trivial, perhaps - and
> definitely less than I was expecting.
Oh for sure, with the amount of drives that somebody like Google will be
consuming, a 2% difference is indeed significant, but at home, that
equates to a very low reduction is working life on a single unit.
> Hmm.
>
> Coo.
>
> The more I hear about HDD reliability, the stranger it seems.
Indeed. Certainly nothing can be taken for granted.
Andy Hewitt
> Jaimie Vandenbergh <jai...@sometimes.sessile.org> wrote:
>
> [snip]
>
> > All anecdotes.
>
> And all data, too. All data is reports: all data is anecdotes.
> Scientifically valid data is anecdotes that lots of people come out
> with, all agreeing with each other.
>
> > I still use them to inform my own purchasing (weighted against
> > availability of preferred makes), but I don't consider those opinions
> > to be statistically valid.
>
> Not safe to make detailed specific predictions, but safe to make rough
> general predictions.
>
> Do you even know the difference between `statistically valid' and `good
> enough for rough predictions'? Anyone who's worked with reality knows
> how to make predictions from inadequate data and if you've worked at it
> you can make them pretty reliably (it's one of the things that makes us
> superior to computers[1]). Me and Andy both have such experience, so we
> take the same view of this sort of thing. We can do it - can you?
Wow.
Yes, that makes sense really, I spent most of my life having to analyse
incomplete information in order to make decisions, it was my job for
many years. I could also use a build up of experience, both mine and
others, to help diagnose faults based on a variety of common defects
that affected certain products. It mostly worked very well, occasionally
it'd end up with a unique cause. It's far more efficient that ignoring
all the information and starting from scratch every time.
Andy Hewitt
> On Tue, 8 Dec 2009 23:19:25 +0000, thewil...@me.com (Andy Hewitt)
> wrote:
[..]
> >> Domestic disks use fluid bearings these days, there isn't a metal ball
> >> in there at all any more. Started about 2002ish.
> >
> >I didn't know they used ball-bearings anyway, I always thought they were
> >fluid types.
>
> I've got a 40meg Rodime upstairs somewhere with a visible main
> bearing! 1989 vintage, I think.
Blimey.
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
>
> > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
[..]
> > > You wouldn't feel like that about a rambling car with a V8 engine,
> > > I guess. (consuming for example 20l/100km)
> >
> > I would expect the owner of such a car to not give a toss about the
> > environment though.
>
> [snip]
>
> A fuel-inefficient V8 is no environmental hazard, simply expensive. NO
> SINGLE ANYTHING done by a human is a threat to the maintenance of a
> global environment suitable for humans (which is what we need to worry
> about).
>
> Coal seams burning out of control due to human incompetence is injecting
> more greenhouse gas into the air than all the old Jags in the world
> would manage if they all ran 24 hours a day flat-out non-stop[1]. Then
> there's the cow farts. Big `gas-guzzling' cars aren't simply not a
> problem until you have to fill up with fuel, or overtake 'em on a narrow
> street (i.e., anywhere near where I live[2]), park in a European city,
> do a U-turn, and so on.
>
> The great thing about the fuel tax increases here is that it's put most
> of the really big vehicles off the road so I can make better progress on
> my motorcycle.
Actually, it doesn't necessarily. I have a car in a higher tax bracket,
but it's had the effect of reducing both its value and saleability to
the point where I can't get rid of it, or afford to change to a smaller
car. It's actually more economical for me to stick with it for now.
What really annoys me is that I pay more tax than my BiL who has an
older gas guzzler that's far less friendly than mine is.
> The only reason big fancy cars are targetted as `environmental threat
> number 1' is that yer average greenie hates rich bastards in flash cars
> and wants to put 'em down - nothing more than that. I dislike rich
> flashy pushy bastards in big fancy cars too, but that's because I hate
> rich flashy pushy bastards. But I'm not a hypocrite and I'll call 'em
> bastards to their face because they're bastards and I won't dishonestly
> pretend that their choice of motor vehicle is causing a signifigant
> increase in atmospheric CO_2, 'cos it ain't.
Agreed.
Andy Hewitt
> Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
>
> > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> >
> > [snip]
> >
> > > Internet folklore doesn't become more reliable by someone echoing it.
> >
> > That which applies to others applies also to you.
>
> Right. A lot of opinions is still folklore,
> not a statistic.
Actually, the data I'm gathering is mostly heresay. There is no
evidence, you just have to trust what people are saying - although I see
no reason to suspect that the majority are making up stories that their
drives have failed!
Jaimie Vandenbergh
wrote:
>Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
>
>> Jaimie Vandenbergh <jai...@sometimes.sessile.org> wrote:
>>
>> [snip]
>>
>> > All anecdotes.
>>
>> And all data, too. All data is reports: all data is anecdotes.
>> Scientifically valid data is anecdotes that lots of people come out
>> with, all agreeing with each other.
You're playing Humpty Dumpty again, redefining words to make an
argument.
>> > I still use them to inform my own purchasing (weighted against
>> > availability of preferred makes), but I don't consider those opinions
>> > to be statistically valid.
>>
>> Not safe to make detailed specific predictions, but safe to make rough
>> general predictions.
Not even that. Without reliable data and analysis thereof, all you're
doing is making yourself feel positive about your decision.
Sometimes your decision may be right, sometimes it may be wrong. This
may or may not be a result of the process you made the decision by.
With domestic hard drives, the penalty is a couple of hours recovering
a backup onto another drive and a half hour handling an RMA. Almost
insignificant effort, so I'm happy to make myself feel happy by buying
things that anecdotally seem like good buys. Really, I have no idea
what the reliability will be on the device.
>> Do you even know the difference between `statistically valid' and `good
>> enough for rough predictions'? Anyone who's worked with reality knows
>> how to make predictions from inadequate data and if you've worked at it
>> you can make them pretty reliably (it's one of the things that makes us
>> superior to computers[1]). Me and Andy both have such experience, so we
>> take the same view of this sort of thing. We can do it - can you?
Yes, I can.
By the way, your phrasing there is in such a manner that it appears
both supercilious and insulting. I frankly have no idea whether the
intent is there, but the appearance is.
>Wow.
>
>Yes, that makes sense really, I spent most of my life having to analyse
>incomplete information in order to make decisions, it was my job for
>many years. I could also use a build up of experience, both mine and
>others, to help diagnose faults based on a variety of common defects
>that affected certain products. It mostly worked very well, occasionally
>it'd end up with a unique cause. It's far more efficient that ignoring
>all the information and starting from scratch every time.
Yes, me too - and these days I have large lumps of company money
riding on my accuracy. But that's a very different thing to buying the
occasional hard drive, so don't be derailed by sudden changes in
scale.
Cheers - Jaimie
--
Beer has food value, but food has no beer value.
Jaimie Vandenbergh
Anecdote doesn't mean what Rowland is saying it does.
You can all check your OSX dictionary, and the Oxford Concise has the
same definition ("entertaining" instead of "amusing"). My OED2 has
broken (C-DILLA license mismanager farked up again) so I can't paste
from that. And to stave off prescriptive vs descriptive wars, they all
match the way everyone except Rowland has been using the word here.
Cheers - Jaimie
--
...most SF writers are small blokes; they spent a lot of time grubbing
around on the floor for old SF mags, not stretching up to the top shelf
for pornography... As an aside, Douglas Adams is quite tall.
- Terry Pratchett
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
> > zoara <me...@privacy.net> wrote:
> >
> > > Andy Hewitt <thewil...@me.com> wrote:
> > > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > [..]
> > It's certainly true that we'll generally mainly hear about only failed
> > units.
>
> Not just that - we'll hear about the "expected" failures more than the
> unexpected ones. Those that confirm the reporter's feelings that brand X
> is the unreliable one. If two brand X drives fail and one brand Y drive
> fails, it's most likely the brand X failures will be reported and the
> brand Y won't.
I'm not seeing that in the reports I'm finding. Quite the opposite now,
I've been seeing much more specific comments about individual drive unit
models.
> > Over a period of time the discrepancies should even out though,
> > and with enough data, you can get a good idea what's what. It's enough
> > to make decisions about purchases IMHO. If you had scanned Google,
> > say,
> > and found dozens of reports of a particular drive failing prematurely,
> > would *you* still buy it?
>
> No, but I am not immune to confirmation bias or the misjudgement of
> risk.
>
> And Googling any brand of drive will give loads of reported failures...
> I just tend to buy on price and reported loudness these days.
>
> > For sure there may be bias against, or for,
> > particular brands, but even that should even out over time.
>
> That's not what "confirmation bias" means. It means that people tend to
> put more weight on events that confirm their already-held beliefs. If
> someone believes Americans are stupid, they'll notice more instances of
> when Americans say something stupid than when Americans say something
> intelligent. But it works the other way too - if someone believes
> Americans are intelligent, they'll notice the intelligent comments more
> than the stupid ones.
Yes, OK, I see your point, although I really don't believe it applies
quite so much to products, or at least a brand name. Certainly
discussions found in various forums seem to find that comments are made
about specific products, rather than an entire brand being crap
(although there are some brands that will be related to cheaply made
products).
> > > It's the same with hard disks. If you're predisposed to the feeling
> > > that
> > > (say) Seagates are unreliable, every time you get a Seagate fail you
> > > will notice it as proof that Seagates are crap, and every time
> > > something
> > > else fails you will dismiss it as an anomoly. That applies whether
> > > you're a home user dealing with a disk or two a year, or a business
> > > dealing with hundreds a year.
> >
> > Not at all, I know that as a rule Seagates have been reliable over a
> > long period,
>
> I was using "Seagate" as an example, not a specific. Replace with a
> brand that you consider unreliable if you want :) My comment was
> explaining confirmation bias, not saying which brand is unreliable.
Understood.
> > > That said, I don't think it's worth wasting any time doing such a
> > > thing.
> > > There are as many different opinions as to which manufacturers are
> > > reliable as there are manufacturers, so it's reasonable to assume
> > > that
> > > they're all pretty much the same and - unless you're dealing with
> > > massive volumes - the difference between reliability is so small
> > > that
> > > you won't notice it.
> >
> > I don't use opinions, I use information based on actual failures. If a
> > drive has failed, then it's failed, there's no 'opinion' about that.
>
> Well, obviously. But "Brand X is unreliable" is an opinion, even if it's
> followed up with "because I've had loads of their drives fail".
>
> Look at me and my repeated iBook / MacBook failures. I'm now of the
> opinion that the "consumer-level" laptops from Apple are absolute junk
> and best avoided; but that opinion is purely anecdotal and coloured by
> confirmation bias. If Apple's consumer laptops were failing on the
> levels I was experiencing, they'd be out of business by now. But they
> aren't; I just had bad luck. By the end of it I was unsurprised when a
> machine went, as my confirmation bias said they were junk; but I bet
> failure rates of consumer laptops are so close to the Pro laptops that
> it can't be measured outside of Apple's recall / repair figures.
Yes, but you're taking into account only the information of an
individual (i.e. you!). I'd consider information from a larger number of
sources myself! :-)
That becomes much more valid then.
I used to use terms like 'they're crap', or 'they're reliable' with
cars, which are of course my own (informed) opinions. However, that was
based on actually having to turn spanners on them, and fix the things
that went wrong, and know how they'd been put together. I also had to
handle a mass of warranty claim data (�250k/yr), and once I'd worked on
a number or marques, it's possible to make comparisons.
Of course it's not an opinion if I say that (at the time) 40% of Ford's
have something fail within the warranty period, and Hondas only have 4%
- which would be based on the actual repairs carried out.
I use a similar technique with everything I buy, and use information to
make the best buy I can. You have to learn to filter out the crap
comments, with what is relevant - such as ignoring silly comments that
something is just crap, to those that specify actual failures.
It's not perfect for sure, but it's got to be better than nothing.
Andy Hewitt
> On Wed, 9 Dec 2009 17:32:38 +0000, thewil...@me.com (Andy Hewitt)
> wrote:
[..]
> >Wow.
> >
> >Yes, that makes sense really, I spent most of my life having to analyse
> >incomplete information in order to make decisions, it was my job for
> >many years. I could also use a build up of experience, both mine and
> >others, to help diagnose faults based on a variety of common defects
> >that affected certain products. It mostly worked very well, occasionally
> >it'd end up with a unique cause. It's far more efficient that ignoring
> >all the information and starting from scratch every time.
>
> Yes, me too - and these days I have large lumps of company money
> riding on my accuracy. But that's a very different thing to buying the
> occasional hard drive, so don't be derailed by sudden changes in
> scale.
Of course not, but I'm always going to try and buy things based on the
best information I can find. It's all about minimising the chances of a
failure, even with a single unit.
Pd
> A fuel-inefficient V8 is no environmental hazard, simply expensive. NO
> SINGLE ANYTHING done by a human is a threat to the maintenance of a
> global environment suitable for humans (which is what we need to worry
> about).
No single drop admits responsibility for the flood.
I did read somewhere that just 16 cargo ships burning shite bunker oil
transporting stupid plastic crapola 'toys' from China for McDonalds to
put in their "HappyMeals" contribute as much to global warming and
atmospheric pollution as the entire world's fleet of cars. Must find a
reference for that.
--
Pd
Pd
> > You wouldn't feel like that about a rambling car with a V8 engine,
> > I guess. (consuming for example 20l/100km)
>
> What's that in sane units? Consumption per distance, please.
What, like litres of petrol consumed per kilometres of distance
travelled? Oh, wait...
--
Pd
Pd
> I spent most of my life having to analyse
> incomplete information in order to make decisions
I'd say all of your life. And all of everybody else's life too. Every
single decision we ever make is based on incomplete information, and the
apparent balance of probability.
That's what annoys me the most about these discussions, is that one
person (not you Andy) draws the line in a different place from somebody
else, then argues that their particular choice of location for that line
is the only sane, sensible, logical, allowable one and everybody else is
not only wrong, but a complete idiot for being wrong.
--
Pd
Pd
Chris Ridd
Are you sure you're reading the right newsgroup? That doesn't sound
very familiar.
--
Chris
Pd
> Actually the main reason Homeopathy is used is because it works, has
> been working for a couple of hundred years and will continue to work.
> And that means really works, not placebo, not trickery nor none of the
> rest of the nonsense that is written about it.
Same as religion. We know it works, we know prayer works, but it works
in a cunning way that does not lend itself to petty scientific
examination and is indistinguishable from random chance.
> Homeopathy has been adapted and used by some very hard thinking and
> sceptical MD's and other health professionals and experts. You think
> these folk don't know a cure when they see one or can't tell the
> difference if an illness gradually resolved itself of its own accord
> or was improved by Homeopathy?
That's exactly what I think.
> Think they can't tell the difference
> between a placebo effect and a real cure?
It's been shown with actual placebos and real cures, so why would these
hard thinking and clever experts suddenly not be fooled by homeopathy?
> Think again.
Okay.... well, I thought again and I still think hard thinking and
sceptical MDs and other such experts can't necessarily tell the
difference between a disease that resolved itself and would have
regardless of the homeopathic treatment, and one that would not have
resolved itself without the homeopathic treatment.
Chances are that if someone is having homeopathic treatment, there is
probably someone praying for that person's recovery. How do you prove
that the cure was effected by homeopathy rather than prayer? How can you
prove that someone isn't being prayed for?
> Think that high dilutions with ALL the molecules of the curative
> substance diluted away can't have a biological effect?
I hope not - what about all the poo particles that those water molecules
have been in contact with? Why don't they have a biological effect?
> Think again
> (Ennis, Inflammation Research, vol 53, p181 - yes, it's been repeated
> and confirmed by others, yes the biological effect stimulated acted as
> though the missing molecules were still there, no the BBC's
> documentary did not repeat her experiment and did not "refute it").
I'd really like to believe Ennis's results, I like those kinds of
mysteries and unexplained phenomena. In fact I *do* believe the results,
now I'm just waiting for some testable theory to explain it, and to
account for the obvious anomalies. The Horizon programme is a bit of a
fly in the Ennis ointment, though. It didn't repeat her experiment - it
used a much more rigorous examination with double blind and controls,
and found no effect at all.
I'm not saying homeopathy doesn't work, but I do think if it does work
it's not for the reasons homeopaths think it is.
--
Pd
Pd
Oops, sorry, you're right - wrong newsgroup.
I meant that to go into uk.rec.linedrawing.disputed.
--
Pd
T i m
wrote:
>I'm not saying homeopathy doesn't work, but I do think if it does work
>it's not for the reasons homeopaths think it is.
You know when people say 'they fought their cancer', I wonder if we do
have the powers within us to affect such things (and so could partly
explain many non scientific / provable 'solutions')?
Like if they had just given up, would it have made any difference? Is
there more than one level of auto reaction ... fight, do nothing, give
up?
T i m
Jim
>
> I'm not saying homeopathy doesn't work, but I do think if it does work
> it's not for the reasons homeopaths think it is.
I strongly suspect it works for the same reasons that having plenty of
bedrest and keeping yourself properly hydrated works.
Jim
--
http://www.ursaMinorBeta.co.uk http://twitter.com/GreyAreaUK
"A Microsoft spokesman said: "We spent five years and $350million
developing a system that would make it really easy for us not to
give you your money back." The Daily Mash
Woody
> On Thu, 10 Dec 2009 10:23:44 +0000, peter...@gmail.invalid (Pd)
> wrote:
>
Ooh link time!
> >I'm not saying homeopathy doesn't work, but I do think if it does work
> >it's not for the reasons homeopaths think it is.
<http://www.thedailymash.co.uk/news/health/ten-in-ten-homeopathic-prescr
iptions-contain-mistakes-200912032281/>
> You know when people say 'they fought their cancer', I wonder if we do
> have the powers within us to affect such things (and so could partly
> explain many non scientific / provable 'solutions')?
<http://news.bbc.co.uk/1/hi/health/8398728.stm>
Except that is both scientific and provable
--
Woody
Pd
> T i m <ne...@spaced.me.uk> wrote:
>
> > On Thu, 10 Dec 2009 10:23:44 +0000, peter...@gmail.invalid (Pd)
> > wrote:
> >
>
> Ooh link time!
>
> > >I'm not saying homeopathy doesn't work, but I do think if it does work
> > >it's not for the reasons homeopaths think it is.
>
> <http://www.thedailymash.co.uk/news/health/ten-in-ten-homeopathic-prescr
> iptions-contain-mistakes-200912032281/>
"If any homeopaths disagree, I suggest they mix up a tonic of one part
my balls to three parts my big, fat, hairy arse."
Made Oi larf.
> > You know when people say 'they fought their cancer', I wonder if we do
> > have the powers within us to affect such things (and so could partly
> > explain many non scientific / provable 'solutions')?
>
> <http://news.bbc.co.uk/1/hi/health/8398728.stm>
> Except that is both scientific and provable
Of course we have the power within us - conventional medicine doesn't
cure anything at all, except perhaps the tiny percentage of purely
mechanical interventions. All of it is simply aiding the body's own
ability to heal itself.
--
Pd
T i m
wrote:
>T i m <ne...@spaced.me.uk> wrote:
>
>> On Thu, 10 Dec 2009 10:23:44 +0000, peter...@gmail.invalid (Pd)
>> wrote:
>>
>
>Ooh link time!
>
>> >I'm not saying homeopathy doesn't work, but I do think if it does work
>> >it's not for the reasons homeopaths think it is.
>
><http://www.thedailymash.co.uk/news/health/ten-in-ten-homeopathic-prescr
>iptions-contain-mistakes-200912032281/>
Ridiculous <takes ivy hat off and puts foil one back on> [1]
>
>> You know when people say 'they fought their cancer', I wonder if we do
>> have the powers within us to affect such things (and so could partly
>> explain many non scientific / provable 'solutions')?
>
><http://news.bbc.co.uk/1/hi/health/8398728.stm>
>
>Except that is both scientific and provable
Thought it might be ..
Cheers, T i m
[1] On a rare visit to the doc the other day I took my prescription
into the chemist. I'd had been prescribed some antihistamine for 'an
allergy to something' [2] and I asked the chemist should this
prescription work, would I have to be on it for ever. They said yes or
I could take some homeopathic stuff ..
[2] It started when I got this Mac?
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
>
> > I spent most of my life having to analyse
> > incomplete information in order to make decisions
>
> I'd say all of your life. And all of everybody else's life too. Every
> single decision we ever make is based on incomplete information, and the
> apparent balance of probability.
Quite so, although I was trying to make it clear that I specifically did
this for a job, so am probably more attuned to it than average.
> That's what annoys me the most about these discussions, is that one
> person (not you Andy) draws the line in a different place from somebody
> else, then argues that their particular choice of location for that line
> is the only sane, sensible, logical, allowable one and everybody else is
> not only wrong, but a complete idiot for being wrong.
Yes, although that happens in many newgroups - try arguing in the
subject of digital cameras for instance. That makes ours look like an
argument with the Chuckle Brothers.
Citizen Jimserac
> Citizen Jimserac <jimse...@gmail.com> wrote:
> > Actually the main reason Homeopathy is used is because it works, has
> > been working for a couple of hundred years and will continue to work.
> > And that means really works, not placebo, not trickery nor none of the
> > rest of the nonsense that is written about it.
>
> Same as religion. We know it works, we know prayer works, but it works
> in a cunning way that does not lend itself to petty scientific
> examination and is indistinguishable from random chance.
You might as well compare chemotherapy with religion too. We know it
kills cancer cells.
Unfortunately, it does a whole lot more.
>
> > Homeopathy has been adapted and used by some very hard thinking and
> > sceptical MD's and other health professionals and experts. You think
> > these folk don't know a cure when they see one or can't tell the
> > difference if an illness gradually resolved itself of its own accord
> > or was improved by Homeopathy?
>
> That's exactly what I think.
> > Think they can't tell the difference
> > between a placebo effect and a real cure?
>
> It's been shown with actual placebos and real cures, so why would these
> hard thinking and clever experts suddenly not be fooled by homeopathy?
>
> > Think again.
>
> Okay.... well, I thought again and I still think hard thinking and
> sceptical MDs and other such experts can't necessarily tell the
> difference between a disease that resolved itself and would have
> regardless of the homeopathic treatment, and one that would not have
> resolved itself without the homeopathic treatment.
Well your opinion is fine so long as you are willing to believe that
these medical people, the vast majority of which were and are among
some of the finest medical minds of their day, were somehow, without
knowing it, fooling themselves. Some of the Homeopathic journals
were published with a continuity of over a century - and again and
again from different Doctors, in different countries, at different
times, are reported exactly similar curative effects based on the
symptomologies with the same types of remedies. That means you
believe in the longest sustained example of mass self delusion in the
history of medicine, indeed in all of science.
OK, how about the following example, from "The Patient, Not the Cure"
by Dr. Majorie Blackie:
She tells of parents who had called her to the bedside of a child who
had been given up for dead by conventional doctors. The child had
Purpura and was bleeding through his pores, eyes, nose. Dr. Blackie,
using crazy Homeopathy theories prescribed Lachesis in a Homeopathic
doseage, a snake venom which soon stopped the bleeding. - the child
survived. Dr. Blackie later overheard another MD explaining the case
in the hospital to a bunch of students as a rare example of an illness
which had "spontaneously resolved" itself. (My thoughts, the flu or
some other self resolving illnesses, yes, maybe but hemorrhagic
Purpura? Unlikely). No mention of the Homeopathic administration was
made. In Homeopathy theory, as best as I can understand it (my field
is Oriental medicine and herbology, I'm not a Homepath), Homeopathic
doses (no, or practically no molecules of the Lachesis left) the
rattlesnake venom has the effect of triggering a response in the
subject which causes a cessation of bleeding. It sure as hell was not
placebo, this kid was not long away from death.
>
> Chances are that if someone is having homeopathic treatment, there is
> probably someone praying for that person's recovery. How do you prove
> that the cure was effected by homeopathy rather than prayer? How can you
> prove that someone isn't being prayed for?
>
> > Think that high dilutions with ALL the molecules of the curative
> > substance diluted away can't have a biological effect?
>
> I hope not - what about all the poo particles that those water molecules
> have been in contact with? Why don't they have a biological effect?
>
> > Think again
> > (Ennis, Inflammation Research, vol 53, p181 - yes, it's been repeated
> > and confirmed by others, yes the biological effect stimulated acted as
> > though the missing molecules were still there, no the BBC's
> > documentary did not repeat her experiment and did not "refute it").
>
> I'd really like to believe Ennis's results, I like those kinds of
> mysteries and unexplained phenomena. In fact I *do* believe the results,
> now I'm just waiting for some testable theory to explain it, and to
> account for the obvious anomalies.
That's what we are all waiting for. We must let genuine scientists
and researchers see if there is a major breakthrough to be had here.
What you've got, what I've got, what the armchair pseudo-sceptics who
write articles and make money publishing "debunking" books is nothing
more than opinion.
>The Horizon programme is a bit of a
> fly in the Ennis ointment, though. It didn't repeat her experiment - it
> used a much more rigorous examination with double blind and controls,
> and found no effect at all.
The Ennis experiment is what turned me from being a complete sceptic
into someone who realized that Homeopathy might just be real.
After laughing at Homeopathy or being appalled at the absolute
nonsense of it, as most people do when first learning of its methods
of high dilution remedy preparation, I began to wonder what all the
fuss was about regarding constant attacks in the news media against
it. If it were such utter nonsense, why would all these "sceptical
debunking" articles be necessary?? Of course it goes without saying
that if it does work and the scientific go of its remedy mechanisms
can be had, a major part of the pharmaceutical industry, along with
billions in profits goes away (boo hoo).
Then I read of the Ennis experiment and of the BBC "refutation".
Suspicious about a TV demo "experiment", I read more and learned that
Ennis herself was suspicious of her result. She spent some weeks
trying to contact the producers and learn who their researcher was.
They eventually gave her the name of the researcher, a man named Guy
Turnbull, and she contacted him, only to learn that in setting up his
BBC version of her experiment, he had quietly added Ammonium
chloride. But this chemical kills Basophil cells, the very cells
which are the subject of the test in the experiment. In addition, he
made this change without telling anyone including not telling the
scientific referee from the British Royal Society. So, although the
common mythology is that her experiment failed in the documentary, the
truth is that their experiment was rendered worthless by the
unannounced change in the protocol. Was this done intentionally or
by accident? I don't know.
Here are some links describing the BBC's "experiment" and their
unannounced "variation".
Here is Researcher Ennis' own comments about the inadequacies and
outright misrepresentaton of the BBC experiment:
www.homeopathyoz.org/downloads/EnnisRebuttal-Dec03.pdf
Next are various relevant comments. It is only a few short paragraphs
but should set you straight on this issue.
http://www.homeopathic.com/articles/view,56
Good details on the whole history of Ennis experiment can be found
here:
http://www.homeopathy.ac.nz/full-text-articles/ultra-dilute-solutions-lies-and-videotape-now-truths/
> I'm not saying homeopathy doesn't work, but I do think if it does work
> it's not for the reasons homeopaths think it is.
On this one I may agree with you!
Thanks
Citizen Jimserac (James Pannozzi)
Jaimie Vandenbergh
<jims...@gmail.com> wrote:
>Well your opinion is fine so long as you are willing to believe that
>these medical people, the vast majority of which were and are among
>some of the finest medical minds of their day, were somehow, without
>knowing it, fooling themselves.
I'm always willing to believe that experts can be fooling themselves.
So are they, which is why double-blind experimental protocols are so
important. For anyone interested in verifiable reproducible results,
at least.
Cheers - Jaimie
--
"It's only work when somebody makes you do it." - Calvin
zoara
> zoara <me...@privacy.net> wrote:
>
> > Andy Hewitt <thewil...@me.com> wrote:
> > > zoara <me...@privacy.net> wrote:
> > >
> > > > Andy Hewitt <thewil...@me.com> wrote:
> > > > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > > [..]
> > > It's certainly true that we'll generally mainly hear about only
> > > failed
> > > units.
> >
> > Not just that - we'll hear about the "expected" failures more than
> > the
> > unexpected ones. Those that confirm the reporter's feelings that
> > brand X
> > is the unreliable one. If two brand X drives fail and one brand Y
> > drive
> > fails, it's most likely the brand X failures will be reported and
> > the
> > brand Y won't.
>
> I'm not seeing that in the reports I'm finding. Quite the opposite
> now,
> I've been seeing much more specific comments about individual drive
> unit
> models.
The problem is, you simply cannot know about the reports that are not
being made (a brand Y drive failing in the above example).
> > > For sure there may be bias against, or for,
> > > particular brands, but even that should even out over time.
> >
> > That's not what "confirmation bias" means. It means that people tend
> > to
> > put more weight on events that confirm their already-held beliefs.
> > If
> > someone believes Americans are stupid, they'll notice more instances
> > of
> > when Americans say something stupid than when Americans say
> > something
> > intelligent. But it works the other way too - if someone believes
> > Americans are intelligent, they'll notice the intelligent comments
> > more
> > than the stupid ones.
>
> Yes, OK, I see your point, although I really don't believe it applies
> quite so much to products, or at least a brand name. Certainly
> discussions found in various forums seem to find that comments are
> made
> about specific products, rather than an entire brand being crap
> (although there are some brands that will be related to cheaply made
> products).
It applies to specific products too. There's a good explanation on
Wikipedia. It's probably amplified on the Internet as if you have a
BigStor CBF-22 fail, you can google it, find that others have had
failures, and decide it's a dud. You replace it with a FastDisk Thunder
44, which you find to be flawless.
Then every time you see someone on your usenet groups report a failure
with the BigStor CBF-22, you may see it as proof that that drive is
faulty (disregarding the more frequent reports of failures with the
FastDisk Thunders).
It's worse when Internet memes get involved. If someone gives a name to
a problem (like DeskStars being renamed DeathStars) the idea becomes
more memorable and so more effective; confirmation bias will increase
dramatically.
It's human nature. We strive to find patterns in order to understand the
world around us. You can't change it, but you can accept it.
> > Look at me and my repeated iBook / MacBook failures. I'm now of the
> > opinion that the "consumer-level" laptops from Apple are absolute
> > junk
> > and best avoided; but that opinion is purely anecdotal and coloured
> > by
> > confirmation bias. If Apple's consumer laptops were failing on the
> > levels I was experiencing, they'd be out of business by now. But
> > they
> > aren't; I just had bad luck. By the end of it I was unsurprised when
> > a
> > machine went, as my confirmation bias said they were junk; but I bet
> > failure rates of consumer laptops are so close to the Pro laptops
> > that
> > it can't be measured outside of Apple's recall / repair figures.
>
> Yes, but you're taking into account only the information of an
> individual (i.e. you!). I'd consider information from a larger number
> of
> sources myself! :-)
But that's precisely the point - I *am* taking into account loads of
sources. I look on the Internet and see loads and loads of reports of
problems with consumer-level Apple laptops. I see problems with them
reported here all the time. Pro machines have a lot fewer reports of
issues.
But it's simply confirmation bias at work. I've decided, through my
experience, that consumer Apple laptops are poor quality. So when I see
reports of this it confirms this belief to me, and I trust those reports
more than reports that say otherwise.
You - or moreso someone who has had problems with the Pro machines and
not the consumer ones - will look at the same Internet and see a
different - or even opposite - take being told; because even with the
same information, different biases produce different conclusions.
> That becomes much more valid then.
It just *feels* more valid. As a concrete example, would you say that
people here are reporting a disproportionate number of problems with
consumer laptops than any other Mac? I'd say they are, and vastly so; I
trust people on this group enough that I could take this as proof that
the consumer laptops are crap.
But if you - or other people - are reading this same group and not
seeing the same levels of problems with consumer laptops, there must be
something going on. Same data, different conclusions. How? It's just
confirmation bias at work.
> I used to use terms like 'they're crap', or 'they're reliable' with
> cars, which are of course my own (informed) opinions. However, that
> was
> based on actually having to turn spanners on them, and fix the things
> that went wrong, and know how they'd been put together. I also had to
> handle a mass of warranty claim data (£250k/yr), and once I'd worked
> on
> a number or marques, it's possible to make comparisons.
Yep. But most people won't experience enough drives to make a
statistically valid argument. They aren't the sort of thing you turn
spanners on, so you can get no sense of the quality of engineering. They
pretty much either work or they don't, and if you have three go pop in a
year, what's to say you're not just unlucky? Maybe the next hundred you
buy will be flawless?
And even with spanner-turning, confirmation bias will be at work. Strip
a thread on a "crap" car and you'll likely remember it. Strip it on a
"good" car and you're more likely to dismiss it.
> Of course it's not an opinion if I say that (at the time) 40% of
> Ford's
> have something fail within the warranty period, and Hondas only have
> 4%
> - which would be based on the actual repairs carried out.
Actually, it is. It wouldn't be an opinion if you said "40% of Fords we
worked on". But as you don't know the total failure rate, your
experience could be (I emphasise *could be*) very unusual.
> I use a similar technique with everything I buy, and use information
> to
> make the best buy I can. You have to learn to filter out the crap
> comments, with what is relevant - such as ignoring silly comments that
> something is just crap, to those that specify actual failures.
>
> It's not perfect for sure, but it's got to be better than nothing.
I'm not arguing that it has no benefit, just that the benefit is a lot
smaller than common sense might imply. I look at reviews too, and I base
my decisions on them. But I know it's more about making me feel I'm
making the right choice than actually having any measurable effect on
the experience I have after buying the drive. As such I take just a few
minutes to confirm that the drive has at least some good reviews, and
leave it at that.
Oh, and this only applies to measures of reliability; if I was buying
something where the features and experience differed (iPhone versus
Droid, etc) I'd place a hell of a lot more importance on reviews. But
reliability is very difficult to measure.
Rowland McDonnell
> Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
[snip]
> > Do you even know the difference between `statistically valid' and `good
> > enough for rough predictions'? Anyone who's worked with reality knows
> > how to make predictions from inadequate data and if you've worked at it
> > you can make them pretty reliably (it's one of the things that makes us
> > superior to computers[1]). Me and Andy both have such experience, so we
> > take the same view of this sort of thing. We can do it - can you?
>
> Wow.
>
> Yes, that makes sense really, I spent most of my life having to analyse
> incomplete information in order to make decisions, it was my job for
> many years. I could also use a build up of experience, both mine and
> others, to help diagnose faults based on a variety of common defects
> that affected certain products. It mostly worked very well, occasionally
> it'd end up with a unique cause. It's far more efficient that ignoring
> all the information and starting from scratch every time.
... even when it doesn't work properly *all* the time, it's still worth
gambling on what *feels* like the right approach, yes? 'cos it
/usually/ works out pretty well, am I right?
Rowland.
--
Remove the animal for email address: rowland....@dog.physics.org
Sorry - the spam got to me
http://www.mag-uk.org http://www.bmf.co.uk
UK biker? Join MAG and the BMF and stop the Eurocrats banning biking
Rowland McDonnell
> J. J. Lodder <nos...@de-ster.demon.nl> wrote:
>
> > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> >
> > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > >
> > > [snip]
> > >
> > > > Internet folklore doesn't become more reliable by someone echoing it.
> > >
> > > That which applies to others applies also to you.
> >
> > Right. A lot of opinions is still folklore,
> > not a statistic.
>
> Actually, the data I'm gathering is mostly heresay.
Inadmissable in court - but not something to ignore in the real world.
> There is no
> evidence,
But there is - the testimony of a person. It's evidence. It might not
be particularly reliable evidence, but it's evidence.
> you just have to trust what people are saying
I find it's best to `provisionally trust' such reports - work on the
basis that they're mostly right, but always looking out for signs to the
contrary with a desire to ditch the existing model if good reason for
doing so is found.
>- although I see
> no reason to suspect that the majority are making up stories that their
> drives have failed!
Oh, I can suspect that - and often do. I have so far almost always
rejected the idea.
But sometimes, you'll come across a forum full of people who just lie to
strangers to put 'em wrong because they like doing that - the same sort
of person who directs you in completely the wrong direction when you ask
'em the way. Hmm: I've only ever had that done to me in Manchester.
Nowhere else in the world...
Rowland.
(the sort of well-bred polite middle class southerner who gets hatred
from /plenty/ of Mancs the instant he opens his mouth Scousers are
nothing like as bigotedly moronic about things - inbred idiots perhaps,
but not as maliciously nasty to me at least)
P.S. As per ushual, I do not characterise all inhabitants of Manchester
and Scouschester as above - I describe a minority as above. For those
who are reading purely because they wish to follow up with personal
abuse directed at me, please do not bother.
Rowland McDonnell
> Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
[snip]
> > The great thing about the fuel tax increases here is that it's put most
> > of the really big vehicles off the road so I can make better progress on
> > my motorcycle.
>
> Actually, it doesn't necessarily.
It did in fact do so round my way.
>I have a car in a higher tax bracket,
> but it's had the effect of reducing both its value and saleability to
> the point where I can't get rid of it, or afford to change to a smaller
> car. It's actually more economical for me to stick with it for now.
<heh> Check your annual fuel and tax bills - you might find that you're
wrong.
The way things are going, I'm going to have to learn to handle a car
soon. I pity whoever has the job of teaching me, given that I've been
riding bikes like a psychonutter for decades[1].
(I suspect few pupils begin by whining about the total absence of
acceleration and how that makes it so bloody dangerous trying to make
progress)
> What really annoys me is that I pay more tax than my BiL who has an
> older gas guzzler that's far less friendly than mine is.
[snip]
What annoys me is that you can buy a Ford Fiesta with a combined cycle
economy of about 76mpg (so the advert says, anyway) - but can one buy a
motorcycle with, say, - oh, even that much, let alone any more? My wife
riding her bike can get about 50mpg[2] - why so poor?
Rowland.
[1] Accidents involving other people caused by my riding: zero
Damage caused to other people by my riding: zero (bar pollution).
Noticeable injuries to me due to my riding: zero
I'm safe enough - just alarming to watch. Which is part of the plan:
defensive riding, don'tchaknow? I make sure the buggers do see me and
do decide to keep well out of my way until I'm long gone. So they
classify me as a psychonutter-suicide-on-two-wheels? Good! Well,
unless the person doing the classifying has a switch on his dashboard to
turn on some sirens and flashing blue lights. I like to give a
different impression to that type of driver, can't think why.
[2] She's been known to complain. She has to fill up after about 140
miles if I've not been riding her bike; rather sooner otherwise.
Rowland McDonnell
> Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
>
> > Andy Hewitt <thewil...@me.com> wrote:
> >
> > > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> [..]
> > > > I've still never had a Seagate pack up on me.
> > >
> > > Nor have I, but the latest ones at 1TB seem to be failing - going by
> > > reports I find in various forums.
> >
> > Righto. Pooey (?). Got a couple of 1.5TB Barracudas - didn't seem to
> > be any reliability problems when we looked (speed problems, firmware
> > issue, got sorted).
>
> Indeed no, I believe it's only the 1TB ones that are affected.
Righto. Well, the 1.5TB HDDs are a better price per bit when I last
checked, not long back.
> > Gonna get another 1.5TB Barracuda. Got the current pair in RAID mirror.
> > Wanna spare disc...
>
> Well, having started this thread, and had a think, I'm going to wait for
> now, thanks. I'm actually waiting for things to happen, which means I
> can get sorted a bit better financially, and maybe get a new iMac
Sensible. Computery stuff tends to get cheaper - or at least better
value, for a given value of `value' (hey, I've a working Mac Plus
whereas my 4G5 didn't even last four years).
> - that
> may change how I need my setup - although I'm still possibly thinking of
> minimising costs, but I'd rather like a RAID mirror setup for my
> backups.
We've got a pair of Taurus II enclosures - FW800 and hardware RAID. Had
a couple of kernel panics, too (3.06GHz Core2Duo iMac, 10.6.2 - and I do
mean `exactly two kernel panics'). No idea what caused 'em - could have
been the external kit hooked up at the time, which means suspecting that
there's something iffy about the things. <shrug> I like the boxes,
though - and given that I've been using external discs thus: two FW800
Taurus II RAID 2-disc enclosures daisy-chained to the iMac, then a FW400
(older) 2-disc enclosure hooked up to that. It's the sort of situation
where I suspect that my tendency to switch all the discs on in one go
might have been a cause for complaint on the part of something in the
chain - so I've stopped doing that, and not had a kernel panic since.
Who knows?
You wanna look?
<http://www.macpower.com.tw/files/manual/taurusraid_en.pdf>
> [..]
> > > > Why more reliable?
> > >
> > > Erm, that was actually the question I was asking! Are they in fact
> > > *more* reliable, as the marketing speak suggests, or is that just
> > > cobblers?
> >
> > Ah. Oh. Dunno. :-)
>
> Of course they also say they're quiter and more energy efficient, but I
> suspect there are other ways to achieve that.
The advertising says `More reliable' and `quieter' and `more energy
efficient' - but they don't give you the figures to compare so one never
knows what improvements if any exist, nor compared to what.
Newer HDDs do tend to be quieter than the old 'uns. And the 1.5TB HDDs
I've got seem to run cooler than some I've met - but hotter than others.
> > > > Don't bet on the lower speed - it's all about what
> > > > the mechanics are designed to do. 7200rpm and 5400rpm are in any case
> > > > more or less the same sort of speed - erm, max wear is on start up and
> > > > shut down, the rest of the time the main disc bearing gets hydrodynamic
> > > > lubrication, so wear there is practically non-existent.
> > >
> > > IIRC, didn't Google post the data about all their server drives a while
> > > back, and found that there was neglible difference between continuous
> > > running, and stop/start running.
> >
> > Hmm. Coo. Blimey. Well, well, well. That did not used to be the
> > case.
> >
> > > Ah yes...
> > >
> > > http://labs.google.com/papers/disk_failures.html
> > >
> > > ...only 2% more after 2 years in use.
> >
> > 2% sounds like it's probably statistically significant given the sample
> > size, and so not negligible. Economically trivial, perhaps - and
> > definitely less than I was expecting.
>
> Oh for sure, with the amount of drives that somebody like Google will be
> consuming, a 2% difference is indeed significant, but at home, that
> equates to a very low reduction is working life on a single unit.
It's stats - it equates to precisly *NOTHING* when considering a single
unit, since stats do not apply to individuals.
[snip]
Rowland.
Rowland McDonnell
> Pd <peter...@gmail.invalid> wrote:
> > zoara <me...@privacy.net> wrote:
> >
> > > The plural of anecdote is not data.
> >
> > I think you'll find that is entirely dependent on relativistic
> > subjectivity. If (using the general "I" for the sake of argument) I
> > see
> > one instance, whatever I conclude is the truth. If you see a million
> > contradictory instances, you're just wrong. And insulting, with it.
>
> Zing!
Of course you shits like making shitty off-topic posts all joining in
with putting the boot into Rowland.
It's a social dynamic sort of thing - all you shits making closer ties
and setting up the social norm of `Rowland's a total fucking shit and
all complaints he makes are bullshit but all complaints about him are
valid' and the entire social grouping joins in.
You people who discuss me off this newsgroup - you know who you are.
You're the shits I'm talking about.
What a bunch of bastards you lot are.
For those who are reading purely because they wish to follow up with
personal abuse directed at me, please do not bother.
Rowland.
Rowland McDonnell
[snip]
> Actually the main reason Homeopathy is used is because it works, has
> been working for a couple of hundred years and will continue to work.
[snip]
Except that there is still no evidence that homeopathy works any better
than placebo - which is also an effective treatment for many medical
conditions, curing many people effectively, and has always done so and
always will do.
If you've any evidence that homoepoathy is better than placebo, do
publish links to it. If not, stop coming out with utter crap.
For those who are reading purely because they wish to follow up with
personal abuse directed at me, please do not bother.
Rowland.
(who has plenty of time for herbalists, but not homoepaths)
Rowland McDonnell
> thewil...@me.com (Andy Hewitt) wrote:
>
> >Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> >
> >> Jaimie Vandenbergh <jai...@sometimes.sessile.org> wrote:
> >>
> >> [snip]
> >>
> >> > All anecdotes.
> >>
> >> And all data, too. All data is reports: all data is anecdotes.
> >> Scientifically valid data is anecdotes that lots of people come out
> >> with, all agreeing with each other.
>
> You're playing Humpty Dumpty again, redefining words to make an
> argument.
[snip]
Bullshit.
I'm just explaining things in my usual straightforward fashion, and what
do I get? A reply that begins with malicious personal abuse.
Fuck off and die you shit.
For those who are reading purely because they wish to follow up with
personal abuse directed at me, please do not bother.
Rowland.
--
Rowland McDonnell
> Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> > Chris Ridd <chri...@mac.com> wrote:
> >
> > > J. J. Lodder said:
> > >
> > > > zoara <me...@privacy.net> wrote:
> > > >
> > > >> The plural of anecdote is not data.
> > > >
> > > > Gosh. I wish I had said that,
> > >
> > > It's a great phrase. I saw someone use it on zfs-discuss, so sadly
> > > it
> > > isn't a Zoara original :-(
> >
> > It's also counter-factual. Data is nothing but anecdotes - a
> > collection
> > of reports is what data *IS*.
>
> Pithiness is inversely proportional to accuracy.
Which principle certain applies to pithy phrases such as `pithiness is
inversely proportional to accuracy'[1]. It's a self-contradicting
statement and therefore very very funny.
E = mc^2 is *VERY* pithy - and no-one's ever seriously disputed it.
For those who are reading purely because they wish to follow up with
personal abuse directed at me, please do not bother.
Rowland.
[1] E = mc^2 is *VERY* pithy - and no-one's ever seriously disputed it.
Ditto e^{i \pi} = -1 (if you don't mind me speaking TeX for a bit).
Jon B
> Andy Hewitt <thewil...@me.com> wrote:
>
> > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> [snip]
>
> > > The great thing about the fuel tax increases here is that it's put most
> > > of the really big vehicles off the road so I can make better progress on
> > > my motorcycle.
> >
> > Actually, it doesn't necessarily.
>
> It did in fact do so round my way.
>
> >I have a car in a higher tax bracket,
> > but it's had the effect of reducing both its value and saleability to
> > the point where I can't get rid of it, or afford to change to a smaller
> > car. It's actually more economical for me to stick with it for now.
>
> <heh> Check your annual fuel and tax bills - you might find that you're
> wrong.
>
For me my annual fuel & tax bills for the ye olde 2.6 V6 cost me �1600.
The depreciation of the car is glacial [1]. If I changed to a newer
vehicle this would probably change to say �700-1000 , but, I'd lose more
than the difference in depreciation in buying the new vehicle the moment
I signed the car over.
Was shocked by someones I knew response to changing of car tax brackets,
potentially pushing their car tax up �300 a year, "I'm buying a new
car". Hang on, you're going to spend �8000 [2] to avoid spending �100 a
year. Bit rash, and imo a bit daft when his biggest cost on the Mondeo
he was driving was the �8000 he'd lost over the last 6 years in
depreciation.
[1] At worst it'd owe my �2000 purchase cost if it was to catch fire
tomorrow, but it's a desirable model in the range, I can't see it losing
very much over next year or two.
[2] After trade in
--
Jon B
Above email address IS valid.
<http://www.bramley-computers.co.uk/> Apple Laptop Repairs.
Woody
> You people who discuss me off this newsgroup - you know who you are.
> You're the shits I'm talking about.
Ooh thats good, I thought you were talking about me for a minute.
--
Woody
Woody
> the same sort
> of person who directs you in completely the wrong direction when you ask
> 'em the way. Hmm: I've only ever had that done to me in Manchester.
> Nowhere else in the world...
I have had it in Huish episcopi. Well and Portsmouth and Brighton.
To be fair, I don't think I have ever talked to anyone in Manchester.
--
Woody
zoara
> Yes, although that happens in many newgroups - try arguing in the
> subject of digital cameras for instance. That makes ours look like an
> argument with the Chuckle Brothers.
To you.
zoara
Hence the statement proves itself. Pithy, but inaccurate.
(there was an extra twist in the meaning of the word pithiness, too)
zoara
>
> I'd say all of your life. And all of everybody else's life too. Every
> single decision we ever make is based on incomplete information, and
> the
> apparent balance of probability.
Yup.
> That's what annoys me the most about these discussions, is that one
> person (not you Andy) draws the line in a different place from
> somebody
> else, then argues that their particular choice of location for that
> line
> is the only sane, sensible, logical, allowable one and everybody else
> is
> not only wrong, but a complete idiot for being wrong.
Yah, I'd just like to point out that my take on this conversation is not
that anyone is wrong, but that there may be more inaccuracy than is
realised.
It's more a "isn't it interesting that the judgements we make can be
apparently so very different from the truth" than a "you're so stupid
for believing this".
Just in case Andy or anyone else felt I was criticising....
Rowland McDonnell
> Andy Hewitt <thewil...@me.com> wrote:
> > zoara <me...@privacy.net> wrote:
> >
> > > Andy Hewitt <thewil...@me.com> wrote:
> > > > zoara <me...@privacy.net> wrote:
> > > >
> > > > > Andy Hewitt <thewil...@me.com> wrote:
> > > > > > J. J. Lodder <nos...@de-ster.demon.nl> wrote:
> > > > [..]
> > > > It's certainly true that we'll generally mainly hear about only
> > > > failed
> > > > units.
> > >
> > > Not just that - we'll hear about the "expected" failures more than
> > > the unexpected ones. Those that confirm the reporter's feelings that
> > > brand X is the unreliable one. If two brand X drives fail and one
> > > brand Y drive fails, it's most likely the brand X failures will be
> > > reported and the brand Y won't.
> >
> > I'm not seeing that in the reports I'm finding. Quite the opposite
> > now, I've been seeing much more specific comments about individual drive
> > unit models.
>
> The problem is, you simply cannot know about the reports that are not
> being made (a brand Y drive failing in the above example).
[snip]
Yes, but there is good reason to think since reports are made
more-or-less randomly, that one can trust that one has a representive
sample - so the fact that not all failures are reported ends up not
mattering in the slightest. In other words, you're wrong about it being
a problem.
This is what's called `human reasoning and problem solving' as is a
powerful tool to apply to the real world. You might not be able to work
things out that way, but don't sneer at those of us who can.
Rowland.
Rowland McDonnell
> peterd.n...@gmail.invalid (Pd) wrote:
> > Citizen Jimserac <jimse...@gmail.com> wrote:
> > > Actually the main reason Homeopathy is used is because it works, has
> > > been working for a couple of hundred years and will continue to work.
> > > And that means really works, not placebo, not trickery nor none of the
> > > rest of the nonsense that is written about it.
> >
> > Same as religion. We know it works, we know prayer works, but it works
> > in a cunning way that does not lend itself to petty scientific
> > examination and is indistinguishable from random chance.
>
> You might as well compare chemotherapy with religion too.
That's a crazy statement. Chemotherapy is the name for any and all drug
treatments. They all have effects which have been proven in clinical
trials to outperform placebo.
Religion is the name for a socially acceptable insanity which in many
cases causes the `believer' (as society calls sufferers of religious
delusions) to think that they've got an invisible friend who's always
with them and so on. Barking mad. A lot of these crazies use
propaganda from thousands of years ago to justify horrific acts of
brutality and cruelty.[1]
> We know it
> kills cancer cells.
> Unfortunately, it does a whole lot more.
[snip]
Any effective drug will be effective, and therefore have effects - in
that, drugs are indeed unlike homeopathic preparations, because they do
something. Powerful drugs have powerful effects. Drug firms and
doctors and nurses are pretty callous about side-effects - basically,
they don't care if you end up as the amazing bearded woman just so long
as the drug has the primary effect they were after[2].
<shrug> On the other hand, if it hadn't been for effective drugs, me
and my two brothers would probably not have survived to adulthood (we've
all had lung trouble, one way or another). Thank heavens for penicillin
and vaccinations!
I've never read of any study which shows that homoeopathy has any effect
at all.
If you know of *ANY* such study, please provide a reference.
Rowland.
[1] Just look at the policies of the nation state of Israel towards
the people who lived there before being evicted forcibly to give Jews a
homeland. I've heard the policies justified by Israeli settlers: `God
gave us this land'. No, it was the bloody Balfour Declaration - do
these crazy Israeli land-thieves *EVER* credit the UK with what they've
got by way of a country to live in? Not the crazy ones who are in the
new settlements: they claim is was all down to God.
And so do the Palestinians trying to blow 'em all up in retaliation.
I've gone off on one, haven't I? Finally, Rowland's solution to the
Middle East problem: get the Russians to build another Tzar Bomba, and
drop it on Jerusalem. Give the people a fortnight to get out, anyone
and anything left gets turned into radioactive glass.
No, the Dome of the Rock, the Western Wall, the `Church' of the Holy
Sepulchre (etc) are not `holy' sites, they are places of ill-omen and
lies and an excuse for the provoking of centuries of brutal conflict.
They have fights inside the `Church' I mentioned - during services, too.
`Christians' conducting a service actually exchanging blows with other
`Christians' involved in conducting the same service.
btw, anyone suggesting that it's anti-Jewish to point out that the
government of Israel is conducting a campaign of cultural and social
destruction against Palestinian Arabs in the region of Israel is a fully
paid up member of the Jewish Nazi Party as far as I'm concerned. And
yes I do know what I just typed. Address your complains to P.O. Box 666
in hell - I'm not interested.
[2] That is not a made-up example - I know a woman who was given a
beard by an NHS prescription. If she didn't shave, she'd have a fuller
beard than I can manage. The NHS refuses to remove the beard because
it's `only a cosmetic problem'. It was an NHS prescription which gave
her the beard - no warning, naturally...
Rowland McDonnell
> peter...@gmail.invalid (Pd) wrote:
>
> >I'm not saying homeopathy doesn't work, but I do think if it does work
> >it's not for the reasons homeopaths think it is.
>
> You know when people say 'they fought their cancer', I wonder if we do
> have the powers within us to affect such things (and so could partly
> explain many non scientific / provable 'solutions')?
Yes: epidemiological studies have proven that if you `fight' your cancer
as so many `brave' cancer sufferers do, you reduce your chances of
survival, and it's reckoned that the reason is increased stress and
things like that. Your best bet for a cancer cure is to relax to the
inevitability of death and chill out, so the studies show. Or so I've
read.
> Like if they had just given up, would it have made any difference?
Yep - it would have improved their chances.
> Is
> there more than one level of auto reaction ... fight, do nothing, give
> up?
Eh? Flight/fight is one pairing that comes from an animal's response to
immediate threat.
Rowland.
Rowland McDonnell
> Citizen Jimserac <jims...@gmail.com> wrote:
>
> >Well your opinion is fine so long as you are willing to believe that
> >these medical people, the vast majority of which were and are among
> >some of the finest medical minds of their day, were somehow, without
> >knowing it, fooling themselves.
>
> I'm always willing to believe that experts can be fooling themselves.
> So are they, which is why double-blind experimental protocols are so
> important. For anyone interested in verifiable reproducible results,
> at least.
As any fule kno, double-blind trials are somewhat discredited these
days.
Triple blind trials are the way to go in medicine - and all results
published, including those that do not show what the drug manufacturer
would like to show.
Almost *NO* drugs have been properly tested at all. Almost all drugs
have had no real testing. Those which have had testing have almost all
only had double-blind (flawed) tests, and negative studies have been
suppressed by the drug firms to give a false impression of drug
effectiveness.
Drug firms have been operating a deliberate con on us for decades.
There is good evidence, apparently, that of drugs introduced during my
lifetime (40+ years), very few offer a real improvement on what had gone
before. Some new drugs are actually more dangerous than the older ones
that they're meant to be `safer more effective more modern' versions of.
So we need triple blind trials (triple blind? The person analysing the
data doesn't know which group is which) and all results to be published.
It's beginnning to happen.
Rowland McDonnell
You too. You're one of the nastiest, rudest posters here. A malicious
lying bastard who likes making others miserable just for the hell of it.
You're a festering sore on the backside of humanity like so many others
here, and I despise you.
Rowland McDonnell
> Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
>
> > Andy Hewitt <thewil...@me.com> wrote:
> >
> > > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> > [snip]
> >
> > > > The great thing about the fuel tax increases here is that it's put most
> > > > of the really big vehicles off the road so I can make better progress on
> > > > my motorcycle.
> > >
> > > Actually, it doesn't necessarily.
> >
> > It did in fact do so round my way.
> >
> > >I have a car in a higher tax bracket,
> > > but it's had the effect of reducing both its value and saleability to
> > > the point where I can't get rid of it, or afford to change to a smaller
> > > car. It's actually more economical for me to stick with it for now.
> >
> > <heh> Check your annual fuel and tax bills - you might find that you're
> > wrong.
> >
>
> For me my annual fuel & tax bills for the ye olde 2.6 V6 cost me �ｿｽ1600.
Hmm. Not so bad.
> The depreciation of the car is glacial [1]. If I changed to a newer
> vehicle this would probably change to say �ｿｽ700-1000 ,
Not so much of a saving, then.
>but, I'd lose more
> than the difference in depreciation in buying the new vehicle the moment
> I signed the car over.
What about savings in servicing, if you're looking at full costs?
> Was shocked by someones I knew response to changing of car tax brackets,
> potentially pushing their car tax up �ｿｽ300 a year, "I'm buying a new
> car". Hang on, you're going to spend �ｿｽ8000 [2] to avoid spending �ｿｽ100 a
> year. Bit rash, and imo a bit daft when his biggest cost on the Mondeo
> he was driving was the �ｿｽ8000 he'd lost over the last 6 years in
> depreciation.
I know someone who bought a new car to replace their five year old one
(bought from new, one careful elderly lady owner, always serviced by the
original dealer) - because servicing costs had got silly. She was
spending more than the price of a second hand car on keeping it on the
road for the last couple of years. Skoda, it was. The electronics were
unreliable. Stupidly expensive to fix - and it was all going wrong from
the engine management computer to the fly-by-wire throttle.
> [1] At worst it'd owe my �ｿｽ2000 purchase cost if it was to catch fire
> tomorrow, but it's a desirable model in the range, I can't see it losing
> very much over next year or two.
> [2] After trade in
Whatever. Although I'm going to have to learn to drive a car...
Rowland.
Jon B
> Jon B <black...@jonbradbury.com> wrote:
>
> > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> >
> > > Andy Hewitt <thewil...@me.com> wrote:
> > >
> > > > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> > > [snip]
> > >
> > > > > The great thing about the fuel tax increases here is that it's put
> > > > > most of the really big vehicles off the road so I can make better
> > > > > progress on my motorcycle.
> > > >
> > > > Actually, it doesn't necessarily.
> > >
> > > It did in fact do so round my way.
> > >
> > > >I have a car in a higher tax bracket,
> > > > but it's had the effect of reducing both its value and saleability to
> > > > the point where I can't get rid of it, or afford to change to a smaller
> > > > car. It's actually more economical for me to stick with it for now.
> > >
> > > <heh> Check your annual fuel and tax bills - you might find that you're
> > > wrong.
> > >
> >
> > For me my annual fuel & tax bills for the ye olde 2.6 V6 cost me �1600.
>
> Hmm. Not so bad.
>
> > The depreciation of the car is glacial [1]. If I changed to a newer
> > vehicle this would probably change to say �700-1000 ,
>
> Not so much of a saving, then.
>
It's never gone to run for free, so there's always going to be a tax
charge, and some petrol charge. My biggest saving is by cutting mileage
down from 20k to 6-7k. Actually it's because of that I can have a fun
v6. Same as insurance, there's always going to be a cost, and it's
always going to be �100+ [1], boring little thinks were going to be
�200, regular family car �250, I'm paying under �300. I was actually
paying more last year for my 2.0 4 door saloon.
> >but, I'd lose more
> > than the difference in depreciation in buying the new vehicle the moment
> > I signed the car over.
>
> What about savings in servicing, if you're looking at full costs?
>
Again there's always going to be a cost, yes it's costing me a little
more in serving than my last car, but that was an absolute little gem of
a motor, cost �600, and I think I spent �1000 in three years on keeping
it on the road, and that includes full cambelt service using genuine
parts & a clutch, which is the bulk of it.
This one I've spent about �600 this year, probably need to spend another
�100-200 before the year is out, but, included in that is a PS pump
repair, and a brand spanking new radiator, so not costs likely to be
repeated next year. Well I hope not the radiator was a brand new genuine
ADK (as fitted by Audi originally) radiator. Bonus it was all of �42
[2]. The antifreeze & AF Fluid cost me more from the dealers than the
rad did.
Again I'd still expect to be spending something on a car, even if said I
halved it and was �500 vs �1000, total savings at best �1500.
Of course, the other reason I actually have it, is because I bloody well
enjoy it, I live on the edge of the Derbyshire Peak district, come a
nice day, it's roof down & off into the peaks with the V6 purring,
you're alive, it's wonderful & worth every bloody penny.
> > Was shocked by someones I knew response to changing of car tax brackets,
> > potentially pushing their car tax up �300 a year, "I'm buying a new
> > car". Hang on, you're going to spend �8000 [2] to avoid spending �100 a
> > year. Bit rash, and imo a bit daft when his biggest cost on the Mondeo
> > he was driving was the �8000 he'd lost over the last 6 years in
> > depreciation.
>
> I know someone who bought a new car to replace their five year old one
> (bought from new, one careful elderly lady owner, always serviced by the
> original dealer) - because servicing costs had got silly. She was
> spending more than the price of a second hand car on keeping it on the
> road for the last couple of years. Skoda, it was. The electronics were
> unreliable. Stupidly expensive to fix - and it was all going wrong from
> the engine management computer to the fly-by-wire throttle.
>
Well that can be a bit more understandable, if something is costing a
lot to keep on the road then you can question the reasoning for keeping
it.
> > [1] At worst it'd owe my �2000 purchase cost if it was to catch fire
> > tomorrow, but it's a desirable model in the range, I can't see it losing
> > very much over next year or two.
> > [2] After trade in
>
> Whatever. Although I'm going to have to learn to drive a car...
>
Well you've just have to grit your teeth and listen to what the
instructor says for a few weeks till you've got the extra letters on the
licence.
[1] Though I do know of some rare stories of people via cashback sites
getting their insurance down to about �15.
[2] Dropped gold, found the exact radiator same day on eBay in a
clearance sale from a radiator company. Finished next day. No other
bidders. Bargain.
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
>
> > Yes, although that happens in many newgroups - try arguing in the
> > subject of digital cameras for instance. That makes ours look like an
> > argument with the Chuckle Brothers.
>
> To you.
<groan>
To me
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
>
> > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
[..]
> > Yes, that makes sense really, I spent most of my life having to analyse
> > incomplete information in order to make decisions, it was my job for
> > many years. I could also use a build up of experience, both mine and
> > others, to help diagnose faults based on a variety of common defects
> > that affected certain products. It mostly worked very well, occasionally
> > it'd end up with a unique cause. It's far more efficient that ignoring
> > all the information and starting from scratch every time.
>
> ... even when it doesn't work properly *all* the time, it's still worth
> gambling on what *feels* like the right approach, yes? 'cos it
> /usually/ works out pretty well, am I right?
Of course. It's all about minimising the risks as far as I'm concerned.
As long as you *know* that it's not *always* right, you can be prepared
for that when things do go wrong.
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
> > zoara <me...@privacy.net> wrote:
[..]
> > I'm not seeing that in the reports I'm finding. Quite the opposite
> > now,
> > I've been seeing much more specific comments about individual drive
> > unit
> > models.
>
> The problem is, you simply cannot know about the reports that are not
> being made (a brand Y drive failing in the above example).
I then use the assumption that no reports usually means there's
*nothing* to report. People tend to be fairly vocal when things fail,
and rarely vocal when things work as expected - why should they, it's
just doing what it was bought to do.
[..]
> It applies to specific products too. There's a good explanation on
> Wikipedia. It's probably amplified on the Internet as if you have a
> BigStor CBF-22 fail, you can google it, find that others have had
> failures, and decide it's a dud. You replace it with a FastDisk Thunder
> 44, which you find to be flawless.
>
> Then every time you see someone on your usenet groups report a failure
> with the BigStor CBF-22, you may see it as proof that that drive is
> faulty (disregarding the more frequent reports of failures with the
> FastDisk Thunders).
It may apply to some, but it certainly doesn't to me. I do understand
that product reliability is dynamic, and from one model run to another
it can change dramatically - it may be just one component that needed
changing.
> It's worse when Internet memes get involved. If someone gives a name to
> a problem (like DeskStars being renamed DeathStars) the idea becomes
> more memorable and so more effective; confirmation bias will increase
> dramatically.
It's possible, but that doesn't work with me.
> It's human nature. We strive to find patterns in order to understand the
> world around us. You can't change it, but you can accept it.
Indeed.
[..]
> > Yes, but you're taking into account only the information of an
> > individual (i.e. you!). I'd consider information from a larger number
> > of
> > sources myself! :-)
>
> But that's precisely the point - I *am* taking into account loads of
> sources. I look on the Internet and see loads and loads of reports of
> problems with consumer-level Apple laptops. I see problems with them
> reported here all the time. Pro machines have a lot fewer reports of
> issues.
>
> But it's simply confirmation bias at work. I've decided, through my
> experience, that consumer Apple laptops are poor quality. So when I see
> reports of this it confirms this belief to me, and I trust those reports
> more than reports that say otherwise.
>
> You - or moreso someone who has had problems with the Pro machines and
> not the consumer ones - will look at the same Internet and see a
> different - or even opposite - take being told; because even with the
> same information, different biases produce different conclusions.
I just don't agree with that - certainly not for me anyway. I know that
I can look at a range of information, and use it to good effect, without
bias. I'm probably unusual in that way I suppose, I very rarely have an
opinion biased by, say, 'the media', and advertising almost never works
on me. Sometimes I may come to the wrong conclusions, but that'll be
because I've got incorrect data, or have misunderstood something.
> > That becomes much more valid then.
>
> It just *feels* more valid. As a concrete example, would you say that
> people here are reporting a disproportionate number of problems with
> consumer laptops than any other Mac? I'd say they are, and vastly so; I
> trust people on this group enough that I could take this as proof that
> the consumer laptops are crap.
I can't say I have, although it's not something I've checked either. I
only look up such things when I need too. However, I don't take one
single source into account, I'll read many forums, and reviews, and
decide whether a product is worth a risk. Ignore the one's that just say
'it's crap', and read the ones that say 'it's failed because...'
For one thing, using only this NG isn't a good idea as it's too small an
example, and in itself could just have a bias towards more users with
consumer level models, so are more likely to be reporting failures on
thos models anyway.
> But if you - or other people - are reading this same group and not
> seeing the same levels of problems with consumer laptops, there must be
> something going on. Same data, different conclusions. How? It's just
> confirmation bias at work.
Using data from all sources, or at least as many as is practical, I'd
say that there's probably not a lot in it with failures of Apple
products. Just a feeling though, as I haven't checked, and until I did I
wouldn't make any buying decisions. Even then, I'd narrow it down to the
products I'm interested in.
> > I used to use terms like 'they're crap', or 'they're reliable' with
> > cars, which are of course my own (informed) opinions. However, that
> > was
> > based on actually having to turn spanners on them, and fix the things
> > that went wrong, and know how they'd been put together. I also had to
> > handle a mass of warranty claim data (�250k/yr), and once I'd worked
> > on
> > a number or marques, it's possible to make comparisons.
>
> Yep. But most people won't experience enough drives to make a
> statistically valid argument. They aren't the sort of thing you turn
> spanners on, so you can get no sense of the quality of engineering. They
> pretty much either work or they don't, and if you have three go pop in a
> year, what's to say you're not just unlucky? Maybe the next hundred you
> buy will be flawless?
Indeed, too few units to make a valid choice as an individual.
> And even with spanner-turning, confirmation bias will be at work. Strip
> a thread on a "crap" car and you'll likely remember it. Strip it on a
> "good" car and you're more likely to dismiss it.
Not at all. Sorry, but the bias theory just doesn't work for *me*.
Besides, there'd be too many other factors involved to make it that
simple!
> > Of course it's not an opinion if I say that (at the time) 40% of
> > Ford's
> > have something fail within the warranty period, and Hondas only have
> > 4%
> > - which would be based on the actual repairs carried out.
>
> Actually, it is. It wouldn't be an opinion if you said "40% of Fords we
> worked on". But as you don't know the total failure rate, your
> experience could be (I emphasise *could be*) very unusual.
How do you know I didn't know the total failure rate? As it happens, I
did (for the UK, and the dealership, anyway).
> > I use a similar technique with everything I buy, and use information
> > to
> > make the best buy I can. You have to learn to filter out the crap
> > comments, with what is relevant - such as ignoring silly comments that
> > something is just crap, to those that specify actual failures.
> >
> > It's not perfect for sure, but it's got to be better than nothing.
>
> I'm not arguing that it has no benefit, just that the benefit is a lot
> smaller than common sense might imply. I look at reviews too, and I base
> my decisions on them. But I know it's more about making me feel I'm
> making the right choice than actually having any measurable effect on
> the experience I have after buying the drive. As such I take just a few
> minutes to confirm that the drive has at least some good reviews, and
> leave it at that.
>
> Oh, and this only applies to measures of reliability; if I was buying
> something where the features and experience differed (iPhone versus
> Droid, etc) I'd place a hell of a lot more importance on reviews. But
> reliability is very difficult to measure.
Yes of course, it can't be 100% accurate (it's possible, but not
practical), but you can still obtain enough information to make an
informed decision. As I've said elsewhere, it's just about minimising
the risks. I know that any product can develop a fault at any time, but
there are some that have a higher or lower chance than others.
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
[..]
> > Indeed no, I believe it's only the 1TB ones that are affected.
>
> Righto. Well, the 1.5TB HDDs are a better price per bit when I last
> checked, not long back.
For sure. It's not a price per bit that's my problem, it's the absolute
cost that I can't afford.
> > > Gonna get another 1.5TB Barracuda. Got the current pair in RAID mirror.
> > > Wanna spare disc...
> >
> > Well, having started this thread, and had a think, I'm going to wait for
> > now, thanks. I'm actually waiting for things to happen, which means I
> > can get sorted a bit better financially, and maybe get a new iMac
>
> Sensible. Computery stuff tends to get cheaper - or at least better
> value, for a given value of `value' (hey, I've a working Mac Plus
> whereas my 4G5 didn't even last four years).
Well, I could always get the old SE out of the loft, that does still
work OK.
> > - that
> > may change how I need my setup - although I'm still possibly thinking of
> > minimising costs, but I'd rather like a RAID mirror setup for my
> > backups.
>
> We've got a pair of Taurus II enclosures - FW800 and hardware RAID. Had
> a couple of kernel panics, too (3.06GHz Core2Duo iMac, 10.6.2 - and I do
> mean `exactly two kernel panics'). No idea what caused 'em - could have
> been the external kit hooked up at the time, which means suspecting that
> there's something iffy about the things. <shrug> I like the boxes,
> though - and given that I've been using external discs thus: two FW800
> Taurus II RAID 2-disc enclosures daisy-chained to the iMac, then a FW400
> (older) 2-disc enclosure hooked up to that. It's the sort of situation
> where I suspect that my tendency to switch all the discs on in one go
> might have been a cause for complaint on the part of something in the
> chain - so I've stopped doing that, and not had a kernel panic since.
> Who knows?
>
> You wanna look?
>
> <http://www.macpower.com.tw/files/manual/taurusraid_en.pdf>
Ah yes, I have seen those. Nice box :-)
[..]
> > Of course they also say they're quiter and more energy efficient, but I
> > suspect there are other ways to achieve that.
>
> The advertising says `More reliable' and `quieter' and `more energy
> efficient' - but they don't give you the figures to compare so one never
> knows what improvements if any exist, nor compared to what.
Indeed.
> Newer HDDs do tend to be quieter than the old 'uns. And the 1.5TB HDDs
> I've got seem to run cooler than some I've met - but hotter than others.
About average then? :-)
[..]
> > > 2% sounds like it's probably statistically significant given the sample
> > > size, and so not negligible. Economically trivial, perhaps - and
> > > definitely less than I was expecting.
> >
> > Oh for sure, with the amount of drives that somebody like Google will be
> > consuming, a 2% difference is indeed significant, but at home, that
> > equates to a very low reduction is working life on a single unit.
>
> It's stats - it equates to precisly *NOTHING* when considering a single
> unit, since stats do not apply to individuals.
Indeed not Hari Seldon ;-)
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
>
> > Rowland McDonnell <real-addr...@flur.bltigibbet.invalid> wrote:
> [snip]
>
> > > The great thing about the fuel tax increases here is that it's put most
> > > of the really big vehicles off the road so I can make better progress on
> > > my motorcycle.
> >
> > Actually, it doesn't necessarily.
>
> It did in fact do so round my way.
>
> >I have a car in a higher tax bracket,
> > but it's had the effect of reducing both its value and saleability to
> > the point where I can't get rid of it, or afford to change to a smaller
> > car. It's actually more economical for me to stick with it for now.
>
> <heh> Check your annual fuel and tax bills - you might find that you're
> wrong.
No, I'm not.
My annual fuel bill is about �600 at the moment (it was higher), and the
tax (at the moment) is �215. I might save a couple of hundred if I
changed to a smaller car.
The guv'mints policy on taxing has meant my car has depreciated by over
50% in two years (some �4-5000), and it was a fairly sought after model.
Now I can't even get a dealer to quote a trade-in price.
> The way things are going, I'm going to have to learn to handle a car
> soon. I pity whoever has the job of teaching me, given that I've been
> riding bikes like a psychonutter for decades[1].
>
> (I suspect few pupils begin by whining about the total absence of
> acceleration and how that makes it so bloody dangerous trying to make
> progress)
Having ridden for a number of years, I can sympathise.
> > What really annoys me is that I pay more tax than my BiL who has an
> > older gas guzzler that's far less friendly than mine is.
>
> [snip]
>
> What annoys me is that you can buy a Ford Fiesta with a combined cycle
> economy of about 76mpg (so the advert says, anyway) - but can one buy a
> motorcycle with, say, - oh, even that much, let alone any more? My wife
> riding her bike can get about 50mpg[2] - why so poor?
Yeah, I always found that too (always about 50mpg as well).
I'll bet if you made a bike engine to the same power/CC ratio, and rode
it with the same acceleration and speed, as a car, the bike would do a
lot better.
Andy Hewitt
> Andy Hewitt <thewil...@me.com> wrote:
>
[..]
> > Actually, the data I'm gathering is mostly heresay.
>
> Inadmissable in court - but not something to ignore in the real world.
Quite so.
> > There is no
> > evidence,
>
> But there is - the testimony of a person. It's evidence. It might not
> be particularly reliable evidence, but it's evidence.
I meant 'physical' evidence, but yes, you're right.
> > you just have to trust what people are saying
>
> I find it's best to `provisionally trust' such reports - work on the
> basis that they're mostly right, but always looking out for signs to the
> contrary with a desire to ditch the existing model if good reason for
> doing so is found.
Indeed, that's why I look in more than one forum, and check more than
one review or report I find in my searches.
> >- although I see
> > no reason to suspect that the majority are making up stories that their
> > drives have failed!
>
> Oh, I can suspect that - and often do. I have so far almost always
> rejected the idea.
>
> But sometimes, you'll come across a forum full of people who just lie to
> strangers to put 'em wrong because they like doing that - the same sort
> of person who directs you in completely the wrong direction when you ask
> 'em the way. Hmm: I've only ever had that done to me in Manchester.
> Nowhere else in the world...
Ok, not something I've considered much, although I have seen it on
occasions, but usually it's obvious so I ignore it.
Chris Ridd
> zoara <me...@privacy.net> wrote:
>
>> Andy Hewitt <thewil...@me.com> wrote:
>>> zoara <me...@privacy.net> wrote:
> [..]
>>> I'm not seeing that in the reports I'm finding. Quite the opposite
>>> now,
>>> I've been seeing much more specific comments about individual drive
>>> unit
>>> models.
>>
>> The problem is, you simply cannot know about the reports that are not
>> being made (a brand Y drive failing in the above example).
>
> I then use the assumption that no reports usually means there's
> *nothing* to report. People tend to be fairly vocal when things fail,
> and rarely vocal when things work as expected - why should they, it's
> just doing what it was bought to do.
Yes, but that still doesn't really help you - every product has
failures and no doubt owners of the duds will be vocal. But that
doesn't tell you *how many* of them proportionally are failing.
OTOH products which had well-advertised defects which got fixed might
be better buys than other products, because you *know* they've had lots
of extra development and QA. I'm thinking of the Seagate TB disks, and
possibly the Intel SSDs.
--
Chris