iRex Iliad was not designed for Suspend/Hibernate

[segatang]

My wish is the iRex can provide us the better iLiad which has long battery duration, fast file loading and page-turning, and quickly resume from any kind of reserving power mode.

[CommanderROR]

Not just your wish Segatang...^^

Actually, all that should be possible with the current Iliad and some proper Software...

[Mambo]

What I find annoying that they admit that the power management could be improved. They could do it. Remember: the audio is not turned off! One element of saving is to turn off the audio. 33 Mhz CPU and lower power on the DRAM would be enough.

But they do the implementation so slow. They don't focus on this issue! They don't realize that power management issue should be solved first and everything else comes after that! This is what looks like completely ignoring the user community, it does not matter what else they do well.

[Henry Loenwind]

Originally Posted by Mambo

They don't focus on this issue! They don't realize that power management issue should be solved first and everything else comes after that! This is what looks like completely ignoring the user community, it does not matter what else they do well.

I think they are having all people who can work on power management doing so. Just remember, an application programmmer, who can do PDF zooming, usually is just about no help when it comes to kernel hacking (aka driver programming).

[jæd]

Originally Posted by Mambo

But they do the implementation so slow. They don't focus on this issue! They don't realize that power management issue should be solved first and everything else comes after that! This is what looks like completely ignoring the user community, it does not matter what else they do well.

I think Irex have given their reason for that in their post... Which would you have...? An Illiad that might last for 21 hours but has a good chance of crashing before that....? Or a hibernate mode that only returns to life every other time...?

As mentioned before, there is a lot of back-seat developing on these forums... I'm pretty sure Irex are developing as fast as they can, and from feedback on the beta program they seem to be doing fine.

[Mambo]

Originally Posted by jæd

I think Irex have given their reason for that in their post... Which would you have...? An Illiad that might last for 21 hours but has a good chance of crashing before that....? Or a hibernate mode that only returns to life every other time...?

As mentioned before, there is a lot of back-seat developing on these forums... I'm pretty sure Irex are developing as fast as they can, and from feedback on the beta program they seem to be doing fine.

Am I the only one complaining here about power management?

They have enough testers because they closed down the sign up. They said they need time for testing, not for developing. Otherwise if they don't have enough developers, why don't they hire more?

What they have provided is not a reason. It is just something to make people silent. My iliad crashes sometimes already, so that is not an excuse.

Originally Posted by jæd

Or a hibernate mode that only returns to life every other time...?

This must be a joke isn't it? I don't see the difference against the current situation. When I turn on the device 1 out of 2 times I have to switch off, because it does not see my CF card. Which would I have?
I can see the choice is between instability without power management and instability with power management. The last one please.

There is no such thing as bug free software.

[Mambo]

Originally Posted by Henry Loenwind

I think they are having all people who can work on power management doing so. Just remember, an application programmmer, who can do PDF zooming, usually is just about no help when it comes to kernel hacking (aka driver programming).

Sorry, I did not know that on Linux too there are programmers who do not know anything about the kernel and hardware like on Windows

Maybe it is so, but still he can do alpha-testing, he can read the code what other people have written and learn things and eventually be helpful. Nobody is born to be application programmer. "Power management" is not something that you learn in the university, but you can learn if you WANT by experimenting I guess (with the guidance of a seasoned programmer). At least this is what I would do and so far I always managed to learn new things.

[CommanderROR]

mambo, you are definately not the only one complaining here...:-)

The trouble as I see it is that iRex is doing things back-to-front.
Usually, you build the Core-OS, implement the main elements like menu-structure, power-states and drivers for memory, display, controls,...
Then you start creating applications. You can always make small improvements and fix bugs in the Core-Os on the way, but starting on the basic functions after you have implemented advanced features like PDF zooming isn't really a good idea...

And now, to make it even worse, it seems the Hardware, the only thing we hadn't complained about before, is somehow not capable of allowing proper power-saving.

In case somebody has not really seen the problem yet...
Even if we get those 21 hours of battery-life, that is still less than 1 day of leaving your Iliad switched on without doing anything. It means you'll still have to switch it off after every useage to preserve battery life. You'll still have to endure the long wait until the device powers back up. I can't imagine a useage where this would not be a problem.

Example:
Casual Reading: I go to the lavatory, I pick up my Iliad. By the time it's booted and the book I was reading is on the screen, I can often switch it off again. Cool eh?

"Professional" Reading: You need to look something up for a paper you're just preparing. With paper-books you would go to the shelf, pick the book up and search for the page you need. With the Iliad, you have to pick it up, wait for it to boot and then you can actually start using it's advantages (once searching inside books is supported). In many cases you'd have found what you are looking for in the paper book before the Iliad was even booted. Cool eh?

Note-Taking: You are at home, the phone rings and you answer it, talk to the guy/lady on the other end. She then gives you a number or describes a route or something which you would like to write down. With a piece of paper you can start taking notes as soon as you find a pen. With a properly suspended e-ink device you can start taking notes almost right away. With the Iliad? I hope you have a good memory and still know everything she said by the tiem the Iliad is finally ready to go. You could of course write it down on paper and then write it on the Iliad once it is ready to go, but that would be...really cool eh?

On the Go: You are a busy man/woman with a full schedule and you keep it all on your Iliad. You are on a trip and meet some business partners. You have to work out a time and place for a new meeting so you check your schedule. Or rahter, you go all red and embarassed because your collegues various handhelds (PocketPC, smartphones, Palms, Tablet-PC...) are all there and waiting for input while your Iliad is showing it's pretty black-and-white booting screen. You could of course say: "It's eink, I can use it for ever and ever and don't have to recharge it.." but that would be a lie would it not?

Convinced that this a serious problem? I guess so!
But please, if you can find applications where this won't be a problem and where the Iliad would be a worthy replacement for pen and paper (or books) then feel free to point them out. I'm always willing to learn.

On another note, I talked to the guys from Jinke again and they have interesting plans for 2007.

Read this excerpt carefully...

We devote to the ebook devices; we plan to release 3 series and 6 types new products in 2007. And all these products adopt Linux OS and e Ink display. Some types have the 10 inches e Ink display. We focus on the communication function in our product next year, which include GPRS, WIFI, CDMA and 3G. We don’t design on/off switch on the products, all these can support suspend. When users charge the device one time, it can continue to use one month. We ever plan to adopt the solar charger.

Sounds good eh? Sounds like just about everybody apart from iRex thinks that you should use the strenghts of eink technology as best you can?

[jæd]

Originally Posted by Mambo

Otherwise if they don't have enough developers, why don't they hire more? ... Maybe it is so, but still he can do alpha-testing, he can read the code what other people have written and learn things and eventually be helpful.

In theory, yes, but in practise both new developers and training/learning new skills takes time... Its not an instant-on situation either...

Originally Posted by Mambo

This must be a joke isn't it? I don't see the difference against the current situation. When I turn on the device 1 out of 2 times I have to switch off, because it does not see my CF card. Which would I have?

Sounds like a dodgy cf card... I use mine with a cf card and have never seen this problem...

Originally Posted by Mambo

I can see the choice is between instability without power management and instability with power management. The last one please.

But it wouldn't be... It would be a choice between software (without power management software) with x number of bugs, or this, as well as power management with y number of bugs in it.

If you didn't attempt to minimise the number of bugs you would have x *and* y number of bugs. And imagine the wailing and gnashing of teeth if 2.8 doubled the number of bugs...?

[CommanderROR]

well...that's why we have beta testers isn't it? Through the buggy software at us and we'll help fix it...that's what beta testers do usually...

And I don't think it's a dodgy CF card...I never had that problem before (and I did not change my card), but 2.7 introduced it and it didn't go away with 2.7.1 so it's probably not found yet.

And jaed...we are not talking about WHEN iRex will be implementing power-management, we are talking about the fact that they admitted that they CAN'T implement it at all...at least the kind most of us were expecting...

[Mambo]

Jaed, I think both of us have a point.

Maybe my 4GB Kingston CF card is dodgy. But there are things that are not rational:
1. If I put it into a card reader and attach to a PC it ALWAYS works.
2. It does not matter how many times I take out and put back into the iLiad, it will be not recognized, once it is not recognized on startup. It will be only recognized when I restart the iLiad. This tells me that something is not right with the iLiad.
3. When I restart the iLiad, the card ALWAYS works. Never happened so far to me, that it would be necessary to restart the iLiad 2 times to make the card working again.

On the other hand what I am talking about is I still believe that they can improve the power management. I still don't want to send back the product and a pseudo-suspend mode would suit me. though, I deeply share CommanderRORs feelings.

My point is, that they just seem to have different attitude to the whole thing. It does look like that this issue is not so important for them. They have intentionally adopted an approach of small improvements. This is they strategy. This is an intentional and not an emerging strategy. At this pace it can take forever the development. All I want to see is concentrated efforts for a solution.

I said I am ready for beta testing. But I was not contacted by them. They seem like have enough beta testers by now. So what keeps them back from delivering a test version? Remember they did not say they can not do it, they don't have enough programmers, they said "thorough testing" what takes long. This is what sounds strange to me.

By the way there is no such feature that could not be turned off, so if the device get instable, it is possible to switch back to "safe mode". And what they say, that it is difficult to add a new option to the settings application. Ridiculous.

[b_k]

i want to add some things here.
As a linux user (Ubuntu 6.10 on a T42) I know that powermanagement is troublesome. There are some tricky things, you only need a driver that goes on rampage after STR/STHDD and you loose your Display after suspend. I can see the problems there, even as a non-programmer. But I would have thought a tiny/embedded system would have less problematic points. Especially when the hardware was designed to run a Linux OS. The thing about the buttons is probably that they are no ACPI or keyboard buttons and therefore can't wake the iLiad from a real suspend to RAM. (thats what I think, and some of you already mentioned here or at iRex's forum.)

[k2r]

Originally Posted by b_k

The thing about the buttons is probably that they are no ACPI or keyboard buttons and therefore can't wake the iLiad from a real suspend to RAM.

Yepp, I don't know too much about hardware-design but I doubt that they are polling the buttons for no reason.
Actually I think that this aspect of the iLiad might be an excellent cantidate for http://TheDailyWtf.com/

But I'm still very confident that there is the potential for expanding the iLiad runtime on battery by the factor 5.

If they only would release sufficient sourcecode and instructions how to unbrick the devices...

[kusmi]

Being part of the beta-tester group, I'm also not so sure, if this beta-test group would be a big help on solving power-management issues: Usually developers want some steps to reproduce the fault. But with such low-level stuff like power-managment, we could just report: "Oh, my screen does not work", but could not give any more hints (e.g. steps to reproduce would be always: I put iLiad to suspend...), as we don't have access to internals (e.g. no debugger, no logfiles, etc).

So I agree, we could do some long-term stability testing (e.g. it could be that only every 10th suspend goes ballistic or if you suspend for a long period of time) but I guess iRex does not need us beta-testers for the first implementation round, as it would be relatively trivial to test, if suspend works or not: By just putting iLiad to suspend after they did some code changes :-)

I could also imaging, that this stuff is quite hard to develop, as you would need to test it on real hardware (no simulator), but could be wrong, as I don't develop such low-level stuff :-)

[Snappy!]

Originally Posted by Henry Loenwind

You're quite right, just one little detail: A PC has multiple standby modes, S3 is (I think) "suspend to RAM", while S1 leaves the CPU running and only switches off some other hardware.

So what iRex is aiming for is "Standby S1", with the CPU still running but with a reduced frequency.

If I had to guess why they cannot do S3/"suspend to RAM" or even "suspend to flash", I'd say the hardware buttons have no way to wake up the CPU. So the CPU has to run to check if a button is pressed.

erm ... I doubt so ... wakeup triggers are mostly handled by onboard auxilliary controller chips and not the main CPU itself. They 'just' need such a controller to wait for triggers, very much like how the various IO controller chips can wake up a whole PC through a phone or LAN line even with the whole system switched off.

Granted the IO controller is using some power, but should be miniscule compared to the main CPU running.