System Software Teleconference Minutes, 2007-03-13

[Mitch Bradley]

2007-03-13

Chris, Richard, Andres, Jim, Jordan, Mitch, David

Status

* stable build 303 woo-hoo!

  + Need a series of stable builds.  Try to do it sanely, don't break 
things.

 >> We need to establish some battery life benchmarks.

AI: cjb - measure battery life on both NiMH and LiFe packs.

* are we done with release notes for 303?

  + Will be done by the end of this meeting.

* Test-1 goals
  + 2 hours on NiMH
  + 3 hours on LiFe
  + mesh network
  + UI improvements for choosing APs
  + battery charging problems solved
  + Ebook reader
  + server - a few services, rsync and backup scripts
  + calculator (sugar app)
  + pixel paint (sugar app)

* USB performance problem

  Choices:
    a) Continue with uncached memory investigation
    b) Disable workaround
    c) Take the performance hit

  Decision: disable workaround, reevaluate at B3

AI: dilinger, wmb: make sure released kernel works with fastpath FW

AI: richard: send wmb the power rail ECO

Marvell may have a PCI card with CaFe ASIC.

AI: jg - touch base with Wad about making a DCON breakout board.  Don't 
forget mechanical issues around panel support and cabling.
        
* Suspend/resume

AI: wmb: publish a suspend/resume FW binary.

* Battery problems - what's left?

  - Trickle charge stops after 6 V, doesn't switch over to full charging 
mode
  - LiFe batteries do some funny things with the lights

* Tinderbox/power measurement

  - Waiting on Richard
  -> Chris can talk to the device via the network

* Priorities
  1) Test-1 requirements
  2) Suspend/resume

AI: wmb - file a trac to figure out how to do LX board id.  assign to wad

Current action items:

AI: Andres - look into what it would take to uncache USB descriptors.
- In progress

AI: Richard - update trac tickets after releasing WLAN_RESET firmware
- We have a workaround, but we aren't convinced that it is really a
  host problem as Ronak's comment on the ticket suggests.

AI: dcbw and marcopg to investigate mesh/avahi not working.
      done, but we're seeing crashing reports
- No information

AI: Richard to work with Ronak  & Michailis to get him the wireless
wakeup info he needs.
- In progress - waiting for Quanta's input on EC signals.  Wad is making 
a GPIO map.

Previous action items:

AI: USB analyzer - do we have it in Cambridge
+ Yes!