[OLPC New Zealand] [Testing] Testing Summary: 3 December 2011 - Auckland, New Zealand

[James Cameron]

Thanks for testing.

On Sun, Dec 04, 2011 at 11:44:52AM +1300, Tabitha Roder wrote:
> Testing Summary: 3 December 2011 - Auckland, New Zealand
> Who: Carlos, Charlie, Fabiana, John, Oliver, Tabitha, Tom
> Carlos Nazareno (Naz) from OLPC Philippines came today. We discussed game
> development events and encouraging more volunteers (among other things).
> 
> We received two new prototypes! (we think they are C1s but there is no
> information on the C1 page linked from the bottom of http://wiki.laptop.org/go/
> XO_1.75_B1). The report motherboard 1C1 and one is SKU201 while the
> other is SKU202. We have written up SKUs on
> http://wiki.laptop.org/go/OLPC_New_Zealand/ XO_inventory. These new
> laptops have 1GB ram and 8GB storage, one has a high school keyboard
> and the other has a "covered membrane" keyboard, where the hard white
> plastic of the case flows between all the membrane keys. We don't like
> the covered membrane, but perhaps it works better with smaller
> fingers?

This keyboard will change again, reducing the flex of the cover.

It is different, and therefore difficult for me, but I have large
fingers.  I agree that smaller fingers might help.  I've just asked my
wife, Petria, for a quick check with her smaller fingers, and she
typed, without making any corrections:

"This keyboard is ifficut, to use becaiuse of the angle of ttacfor th
keys. A bit ike when theh space bar ws hrd to ress with teh thumb."

It would teach vertical angle of attack.

> Kid, the covered membrane unit, has a very faulty touchpad, it
> exhibits the same problems we reported in
> https://dev.laptop.org/ticket/11171 but also, sometimes the pointer
> flies all over the screen and clicks on random stuff before getting
> stuck in a corner. Reboot seems to be the only fix when this happens
> -- changing to the text console and back does not help. I have only
> seen this in the operating system, if it helps we can try to reproduce
> it in the open firmware test thing.  [...]

Yes, please test again with "test /mouse" at the ok prompt.  I'm fairly
certain that what you are seeing is one of the suspend and resume
problems relating to the touchpad; #11379.  Testing at the ok prompt
will prove that for us.  Then, in the meanwhile, you might temporarily
disable automatic power management, reboot, and see if the problem goes
away.  Until the problem is fixed, temporarily turn off automatic power
management, but please don't make this a habit because we actually need
that tested as well.

> open firmware prints a warning when a zd4 is installed that it is much
> smaller than the available storage. I was successful increasing the
> size of the partition and the doing an online ext3 resize -- perhaps
> this should be done during first startup? Or could the open firmware
> fs-update do this?

Yes, a warning is issued, but it is harmless and you can proceed
regardless.  There is no need to increase the size of the partition or
do an online ext3 resize, but good to hear it worked for you.  If you
can write that up for other testers on our Wiki?

We have argued about doing this on first startup, but:

- the process may take an unacceptable time,

- we know that if the process is interrupted the filesystem may be
  corrupt and installation will have to be redone,

- deployments will generally order laptops with a known storage size,
  and so they will not likely need this feature,

fs-update can't do it because it doesn't have the capability.  Since the
Linux kernel can do it, the shortest path to having it done is to use
Linux.

> 
> XO-1.75 testing 11.3.1 build 16 Sugar 0.94.1 firmware Q4C06
> 
> Like the last build, the mouse pointer gets corrupted on a regular
> bases such that it looks like a QR code.

Heh.  Nice description.  #11491.

> Power management is back in my settings

Yay!

> Changed timezone to Auckland. Our new probably C1 laptops were
> delivered with the time set to Jan 2000 but seem to be holding the
> correct time after setting it with date. Our B1s are also behaving
> better than they did with last week's build.

Good, that means the date and time problems were most likely the
software problems.

> General feedback - slow.
> 
> All our XO-1.75s are having trouble getting onto the wifi, both at the
> windsor and at home. This is new, the B1s had no problems with the
> build we tested last week. At the windsor, the first one started had
> os16 but had not done the firmware update, this laptop got on the wifi
> straight away. The rest got os16 and the firmware update and couldn't
> get on. The first one got the firmware update and a reboot and then
> couldn't get on. Later we re-set the time on them and rebooted them
> and some of them got on the wifi. At home, so far, none have got on
> the wifi. I'm open any advice on how to debug this, the symptoms are a
> pulsing wifi thing and then eventually (why does wifi take so long to
> fail??) the password dialog appears. Sometimes the password dialog
> doesn't appear at the front and you can't see it but it steals the
> keyboard focus. You can find it with alt-tab.

Almost certainly loss of wireless related to suspend and resume.  #11510

> We found on at least one of our B1s the keyboard stopped working, you
> couldn't even switch to the text console. This might be related to
> going into and out of power saving or maybe the wifi password focus
> stealer?

It is the automatic power management, yes.  I've never seen that with
the automatic power management disabled.

> On all our XO-1.75s we loose the first keypress after coming out of
> power saving.

This won't be fixed in B1 or C1, but is expected to be fixed in C2.

> Speak - starts but no speaking and no lips moving. There is a high
> pitched whine (Tom estimates 10KHz) on clicky keys C1 and then it
> locked up and stopped updating the display. The other C1 also had no
> speaking and no lips moving and then locked in the same way. Both are
> successfully killed using the force quit option if you stop from the
> frame.

The estimated 10kHz tone is likely to be aliasing.  See #11334 for
technical background.  The lockup is unfamiliar, but if automatic power
management was enabled we know it happens.

-- 
James Cameron
http://quozl.linux.org.au/