G4 Powerbook (15") and sleep

[Duncan Farrow]

Hi,

I have just got a new G4 Powerbook (15"), 1GHz with 1GB of RAM. The price
was too good to resist!

Anyway, I'm finding that under sleep I'm losing about 1% per hour of sleep
of battery. This seems much more than I used to lose with my Wallstreet
under 8.6. Is this normal? If I put the machine for sleep over the weekend,
there's barely enough power to wake.

I know you're supposed to shutdown for extended periods of non-use but I
like to have the machine up and running quickly when I pull it out. Of
course, I could leave it plugged in but then it would not be very secure.

Cheers,

Duncan.

[Thomas Otto]

> I have just got a new G4 Powerbook (15"), 1GHz with 1GB of RAM. The price

> was too good to resist!
>
> Anyway, I'm finding that under sleep I'm losing about 1% per hour of sleep
> of battery. This seems much more than I used to lose with my Wallstreet
> under 8.6. Is this normal?

Unfortunately yes. My 12" iBook2 with 386MB RAM is a loosing power at
about the same rate.
Actually IIRC the 1GB of RAM are probably only worsening the problem
since the complete Gigabyte has to be kept under a certain current so
that it doesn't loose its content.
(I don't know if something is developed yet that can distinguish "used"
from unallocated RAM and powers of the latter when suspending).

> If I put the machine for sleep over the weekend,
> there's barely enough power to wake.

Isn't the power-up time just long enough to get something to eat and
dring before starting to work? ;-)

> I know you're supposed to shutdown for extended periods of non-use but I
> like to have the machine up and running quickly when I pull it out. Of
> course, I could leave it plugged in but then it would not be very secure.

I'm quite exited about the software suspend-to-disk in the coming
Linux 2.6 kernel which will save the RAM to disk and then completely
power off. The bootup time will then only be the time it takes to copy
the memory stored on the harddrive back to RAM (ok, plus kernel
initialisation).
I wonder why OS X doesn't have something similar in Panther with laptops
being a key market for Apple.

-Thomas

[Thomas Reed]

In article <BB4C00B4.18D1%duncanfarrow@iprimus.com.au>, Duncan Farrow
<duncanfarrow@iprimus.com.au> wrote:

> Of
> course, I could leave it plugged in but then it would not be very secure.

I'm afraid I don't understand this part. How is leaving it plugged in
less secure?

I leave my PowerBook plugged in all the time when I'm not using it (and
most of the time when I am). That way, when I need the battery power,
I'm always at full capacity.

--
-Thomas

e-mail me at thomasareed at philadelphia.net minus phil

[Duncan Farrow]

Thomas Reed wrote:

>
> In article <BB4C00B4.18D1%duncanfarrow@iprimus.com.au>, Duncan Farrow
> <duncanfarrow@iprimus.com.au> wrote:
>

> > Of
> > course, I could leave it plugged in but then it would not be very secure.

>
> I'm afraid I don't understand this part. How is leaving it plugged in
> less secure?
>
> I leave my PowerBook plugged in all the time when I'm not using it (and
> most of the time when I am). That way, when I need the battery power,
> I'm always at full capacity.
>
> --
> -Thomas
>
> e-mail me at thomasareed at philadelphia.net minus phil

If I leave the machine on my desk, plugged into the power, it is less
secure from theft than if it is locked away somewhere out of sight.

I lost a machine a few years ago in such circumstances (a 5300 so it
didn't matter too much!)

Cheers,
Duncan.

[Neill Massello]

Duncan Farrow <duncanfarrow@iprimus.com.au> wrote:

> If I leave the machine on my desk, plugged into the power, it is less
> secure from theft than if it is locked away somewhere out of sight.

Have you considered the security cable products, such as Kensington's
<http://www.microsaver.com/>?

[Duncan Farrow]

On 30/7/03 9:52 AM, in article
1fyv72g.mhslo916s50ggN%neillmassello@earthlink.net , "Neill Massello"
<neillmassello@earthlink.net> wrote:

> Duncan Farrow <duncanfarrow@iprimus.com.au> wrote:
>

>> If I leave the machine on my desk, plugged into the power, it is less
>> secure from theft than if it is locked away somewhere out of sight.

>
> Have you considered the security cable products, such as Kensington's
> <http://www.microsaver.com/>?

I have one of these - but I don't trust it to hold out against a determined
thief that has gone to the trouble of breaking into my office in the dead of
night. I've not tried it but I bet a hand-held cable cutter or bolt cutters
could get through a cable lock in less than a minute. The thief can then
remove the part still attached to the powerbook at their leisure.

[David Glover]

On 31/7/03 1:19 AM, in article [email]vifog3bogds528@corp.supernews.com[/email], "Sloopy"
<sloopy@hangon.xyz> wrote:

> In article <BB4D5D3B.18F1%duncanfarrow@iprimus.com.au>,
> Duncan Farrow <duncanfarrow@iprimus.com.au> wrote:
>

>> On 30/7/03 9:52 AM, in article
>> 1fyv72g.mhslo916s50ggN%neillmassello@earthlink.net , "Neill Massello"
>> <neillmassello@earthlink.net> wrote:
>>

>>> Duncan Farrow <duncanfarrow@iprimus.com.au> wrote:
>>>
>>>> If I leave the machine on my desk, plugged into the power, it is less
>>>> secure from theft than if it is locked away somewhere out of sight.
>>>
>>> Have you considered the security cable products, such as Kensington's
>>> <http://www.microsaver.com/>?

>>
>> I have one of these - but I don't trust it to hold out against a determined
>> thief that has gone to the trouble of breaking into my office in the dead of
>> night. I've not tried it but I bet a hand-held cable cutter or bolt cutters
>> could get through a cable lock in less than a minute. The thief can then
>> remove the part still attached to the powerbook at their leisure.

>
> I think the Kensington product is designed to keep a light-fingered
> co-employed from "borrowing" your laptop.
>

Although, as Neill says, it would be fairly easy to cut the cable with the
right tool, it wouldn't be easy (if it's possible at all) to get the
Kensington fitting out without making a mess of the case. Kensington's logic
is that a laptop with an obviously buggered case is not as saleable,
therefore not worth stealing.

As always, nothing can stop a thief who has enough time and tools. All you
can do is slow them down - hopefully enough that they'll stick to easier
things.

David

[Simon Slavin]

In article <slrnbii5ld.inh.t-use@ID-685.user.dfncis.de>,
Martin Trautmann <t-use@gmx.net> wrote:

>I don't have any clue about the precision of this power display - but
>7 % for 10 hours (?) seems to be pretty fair: so 70 % would be 100 hours
>or four days?
>
>Ok, I feel it should last for about 720 hours (one month)

There are two big drains on power while the machine is asleep:
RAM and wake-up circuitry. Modern laptop design has done wonders
with both but we're still not down to the 3% a day you want.

While I've seen other people talking about storing the contents
of RAM on disk and powering the RAM down while the laptop is
asleep, there are problems with this approach too:

a) If you have 512Meg of RAM you need to keep 512Meg of drive
free at all times.
b) Writing 512 Meg of RAM to disk and reading it back off again
takes a lot of power. If most of your sleeps are for less than
four hours you actually lose out using this technique.
c) Writing 512 Meg of RAM to disk and reading it back off again
takes a lot of time. When people close the lid of their laptop
they want to see the sleep-light immediately; when they open
the lid they want to see the display immediately.
d) Saving memory to disk is a lot more subject to failure than
just doing nothing. If the process failed during saving it's a
pain to recover from.

[Stephan Schulz]

In article <BB52042A966829D0F2@10.0.1.3>, Simon Slavin wrote:

>In article <slrnbii5ld.inh.t-use@ID-685.user.dfncis.de>,
>Martin Trautmann <t-use@gmx.net> wrote:
>

>>I don't have any clue about the precision of this power display - but
>>7 % for 10 hours (?) seems to be pretty fair: so 70 % would be 100 hours
>>or four days?
>>
>>Ok, I feel it should last for about 720 hours (one month)

>
>There are two big drains on power while the machine is asleep:
>RAM and wake-up circuitry. Modern laptop design has done wonders
>with both but we're still not down to the 3% a day you want.
>
>While I've seen other people talking about storing the contents
>of RAM on disk and powering the RAM down while the laptop is
>asleep, there are problems with this approach too:
>
>a) If you have 512Meg of RAM you need to keep 512Meg of drive
>free at all times.

If you have 512 MB of RAM, you probably have a 30GB Disk. PC
notebooks just keep a separate partition around to make things easier
and faster (no file system to speak of, just dump everything
equentially).

>b) Writing 512 Meg of RAM to disk and reading it back off again
>takes a lot of power. If most of your sleeps are for less than
>four hours you actually lose out using this technique.
>c) Writing 512 Meg of RAM to disk and reading it back off again
>takes a lot of time. When people close the lid of their laptop
>they want to see the sleep-light immediately; when they open
>the lid they want to see the display immediately.

Right. But it's not an either-or, it's a "I want both". On my
second-to-last Linux notebook, a Toshiba, I could choose wether to
suspend to RAM (instantly, draws power while sleeping), or wether to
suspend to disk (about 10-15 seconds, no power use at all)[1].

It would be nice to have this ability on my Powerbook as well.

>d) Saving memory to disk is a lot more subject to failure than
>just doing nothing. If the process failed during saving it's a
>pain to recover from.

Is it that much of a problem? My Toshiba was a 1996 or 1997 make, and
it worked flawlessly.

Bye,

Stephan

[1] On my last Linux notebook, I had the same choice, but
unfortunately, closing the lid would activate the wake-up function
of the suspend-to-ram function, so it was basically useless.
--
-------------------------- It can be done! ---------------------------------
Please email me as [email]schulz@informatik.tu-muenchen.de[/email] (Stephan Schulz)
----------------------------------------------------------------------------