OLPC:News: Difference between revisions
No edit summary |
No edit summary |
||
| Line 5: | Line 5: | ||
=LAPTOP NEWS= |
=LAPTOP NEWS= |
||
1. Brazil: The presidency has created a new working group inside the |
|||
1. Libya and OLPC signed an MOU in which they agreed to work together towards the deployment of one OLPC laptop for every school-age child in Libya and contributing laptops to poor African nations. OLPC will provide the support needed by Libya to plan and carry out such a deployment. The signing took place during a visit by Nicholas, Walter Bender, and Khaled Hassounah, where they met with members of the newly formed OLPC4Libya steering committee. The committee was presented with the outline of a comprehensive plan to distribute the laptops, create connectivity and server infrastructure, and prepare teachers and students. Libya will send a team to the OLPC office in Cambridge to focus on software (Linux), infrastructure, and education content. |
|||
Ministry of Education to focus solely on laptops and learning. The group is |
|||
comprehensive, covering all the necessary departments: basic education; |
|||
teacher and content development; technology; distance support; and |
|||
integration and coordination. This group will coordinate all activities |
|||
needed for the deployment of laptops. |
|||
2. Michail Bletsas and Barry Vercoe gave a 20-minute progress report in the |
|||
2. Nicholas spoke to the International Federation of University Presidents in Seoul on Thursday. Most Asian and many other universities were present. The appeal of OLPC was obvious and the talk deeply appreciated. |
|||
Media Lab's Digital Life sponsor meeting followed by a flawless performance |
|||
of Barry's laptop ensemble. OLPC monopolized the audience's questions and |
|||
overly positive feedback. |
|||
3. Display: Acceptance-criteria testing for the LCD panels began this week. |
|||
3. Michail attended the WiMax World conference in Boston. Nortel made big news in the show announcing their strong commitment to WiMax. |
|||
Various display artifacts were combined to create worst-case panel |
|||
configurations. These panels were graded by Mary Lou. Most artifacts were |
|||
barely (if at all) noticeable. More detailed acceptance criteria will be |
|||
established (called fitness-of-use testing) in mid-November with materials |
|||
from the B1 build. |
|||
4. This week Marcelo Tosatti and Dan Williams continued work on the Marvell |
|||
4. Display: The new dual-mode panels—with double the reflectance—were |
|||
Libertas driver, trying to get it into shape to use in B1. This requires |
|||
hand-carried back by Mary Lou were installed successfully into the motherboard-daughter card test-sets at OLPC. With work by Jim Gettys and Walter Bender, preliminary optimal font size tests were started. A 9-point font (25 pixels) seems ideal for readability as a guideline for development. |
|||
getting enough things working to be able to use it with NetworkManager, one |
|||
of the standard Linux utilities. The driver, as it came from Marvell, is |
|||
adequate for embedded platforms, but isn't ready for desktop use. A lot of |
|||
work has gone into getting things into shape. Dan and John Palmieri also |
|||
spent time building a NetworkManager front end for B1. |
|||
5. Dan also worked with the Thai team to get a group of demos together to |
|||
5. Taipei: Mark Foster has been working with the teams at Quanta, Marvell, and the OLPC software team to bring up a new version of the CAFE chip on the first OLPC “pre-B Test” motherboards. The boards are up and running, so the teams are now concentrating on general debugging. The motherboards are all running Linux, and all three of the CAFE subsystems (camera interface, SD Card controller, and the NAND Flash controller), are available for further driver development. |
|||
take back with them (they returned home after a three-week stay at OLPC). |
|||
6. Marco Gritti and Dan have done a huge amount of work over the last |
|||
6. Additionally, samples of the latest Alps touchpad are also on their way to the software team. These samples include a newer touchpad controller ASIC that is intended for final production, as well as containing the unique OLPC keyboard code functionality. This keyboard code, essentially a superset of that used by conventional PCs, provides direct support for the Sugar user interface, adds convenient shortcut keys, and enables OLPC's unique "analog slider" function keys. |
|||
couple of weeks on Sugar, getting ready for the B1. The shell is working |
|||
pretty well and the browser activity has a lot of the nice style elements |
|||
that Eben Eliason and the Pentagram team have worked on. They also have |
|||
some group-browsing functions done. The Sugar frame now supports the system |
|||
tray spec, so other programs can embed icons and functionality into the |
|||
frame. We'll use this for our network and power status at some point. |
|||
7. Chris Blizzard reports that the team is generating builds at 4PM every |
|||
7. The OLPC document format spec rev2 has support for block macros, granular tables of contents, and layout hinting. The format is named “Crossmark,” and is believed to be highly suitable for (collaborative) book writing and editing and use as an e-book format. |
|||
day. The team at Red Hat is adding their daily work, as appropriate, to |
|||
those builds. This means that we can test drivers, package updates and |
|||
changes every morning when we come in. It has also set internal deadlines |
|||
so that everyone integrates around the same time. |
|||
8. We've also started to include a newer D-Bus to our builds to fix a |
|||
8. One month ago, without warning, Sun released Open Firmware under a BSD license. After much deliberation, we reached the decision that rather than continuing to use Linux itself as boot loader, we will transition to Open Firmware (OFW), while retaining LinuxBIOS. The date for a transition has not yet been set. Open Firmware has been in use on both Sun and PPC hardware (e.g. Apple) for approaching a decade, and has shipped in quantity millions. Mitch Bradley, currently working for OLPC is its original author. |
|||
number of bugs and a newer copy of HAL which saves a lot of memory and is |
|||
much faster. |
|||
9. Mitch cleaned up the SPI (system programming interface) recovery code, |
|||
9. Our firmware development process and strategy has crystallized, and is a major relief to us all. Chris Ball has worked on our internal BIOS build system and procedures, setting up a Xen virtual machine containing a stable system for building the OLPC BIOS. The build system has successfully built LinuxBIOS with both Linux and Open Firmware as bootloaders. A release process has been proposed and is being reviewed. |
|||
merged it with the in-system re-flashing code in OFW, and tested it. In the |
|||
process of testing CAFE NAND support, Mitch solidified support in the OFW |
|||
code base and developed some diagnostics for the NAND chips themselves. |
|||
10. Chris Ball has set up a software stress-test environment and has run it |
|||
10. Chris has also established a tinderbox to enable continuous build testing and performance testing of the OLPC system, from installing firmware, booting, installation onto flash and ensuring the user interface has started. We intend to add performance tests in the future (See http://mailman.laptop.org/pipermail/devel/2006-October/002565.html). |
|||
on each of the configurations of our boards, helping to expose an |
|||
unreliable RAM part, which has now been disqualified from our second-source |
|||
parts list. The tinderbox has its first performance test, tracking general |
|||
Python performance for each of our OS builds. The tinderbox will also be |
|||
used for stress testing of the B1 board in the coming week. |
|||
11. Multi-media: Scott Nelson and Greg Wright from Real Networks performed |
|||
11. Jon Corbet has RGB565 mode (or native mode) working in the camera driver now. Jon will submit a new patch to the OLPC tree today. The camera driver is getting close to completion. The remaining items are support for image sizes other than VGA (which should be easy) and wiring up more controls. |
|||
detailed testing of an A1 board. While their findings were “a result of |
|||
only a few days research and therefore should not be considered final or |
|||
complete,” they nonetheless give an idea of the lower bound of the laptop's |
|||
capabilities, which will only improve. Their conclusion: the device is “a |
|||
capable multi-media platform for everything short of high bit-rate video |
|||
content and large frame sizes.” |
|||
Revision as of 21:26, 21 October 2006
LAPTOP NEWS
1. Brazil: The presidency has created a new working group inside the Ministry of Education to focus solely on laptops and learning. The group is comprehensive, covering all the necessary departments: basic education; teacher and content development; technology; distance support; and integration and coordination. This group will coordinate all activities needed for the deployment of laptops.
2. Michail Bletsas and Barry Vercoe gave a 20-minute progress report in the Media Lab's Digital Life sponsor meeting followed by a flawless performance of Barry's laptop ensemble. OLPC monopolized the audience's questions and overly positive feedback.
3. Display: Acceptance-criteria testing for the LCD panels began this week. Various display artifacts were combined to create worst-case panel configurations. These panels were graded by Mary Lou. Most artifacts were barely (if at all) noticeable. More detailed acceptance criteria will be established (called fitness-of-use testing) in mid-November with materials from the B1 build.
4. This week Marcelo Tosatti and Dan Williams continued work on the Marvell Libertas driver, trying to get it into shape to use in B1. This requires getting enough things working to be able to use it with NetworkManager, one of the standard Linux utilities. The driver, as it came from Marvell, is adequate for embedded platforms, but isn't ready for desktop use. A lot of work has gone into getting things into shape. Dan and John Palmieri also spent time building a NetworkManager front end for B1.
5. Dan also worked with the Thai team to get a group of demos together to take back with them (they returned home after a three-week stay at OLPC).
6. Marco Gritti and Dan have done a huge amount of work over the last couple of weeks on Sugar, getting ready for the B1. The shell is working pretty well and the browser activity has a lot of the nice style elements that Eben Eliason and the Pentagram team have worked on. They also have some group-browsing functions done. The Sugar frame now supports the system tray spec, so other programs can embed icons and functionality into the frame. We'll use this for our network and power status at some point.
7. Chris Blizzard reports that the team is generating builds at 4PM every day. The team at Red Hat is adding their daily work, as appropriate, to those builds. This means that we can test drivers, package updates and changes every morning when we come in. It has also set internal deadlines so that everyone integrates around the same time.
8. We've also started to include a newer D-Bus to our builds to fix a number of bugs and a newer copy of HAL which saves a lot of memory and is much faster.
9. Mitch cleaned up the SPI (system programming interface) recovery code, merged it with the in-system re-flashing code in OFW, and tested it. In the process of testing CAFE NAND support, Mitch solidified support in the OFW code base and developed some diagnostics for the NAND chips themselves.
10. Chris Ball has set up a software stress-test environment and has run it on each of the configurations of our boards, helping to expose an unreliable RAM part, which has now been disqualified from our second-source parts list. The tinderbox has its first performance test, tracking general Python performance for each of our OS builds. The tinderbox will also be used for stress testing of the B1 board in the coming week.
11. Multi-media: Scott Nelson and Greg Wright from Real Networks performed detailed testing of an A1 board. While their findings were “a result of only a few days research and therefore should not be considered final or complete,” they nonetheless give an idea of the lower bound of the laptop's capabilities, which will only improve. Their conclusion: the device is “a capable multi-media platform for everything short of high bit-rate video content and large frame sizes.”
Laptop News is archived at Laptop News.
You can subscribe to the OLPC community-news mailing list by visiting the laptop.org mailman site.
Press requests: please send email to press at laptop dot org.
MILESTONES
| Oct. 2006 | B-test boards become available |
| Aug. 2006 | Working prototype of the dual-mode display |
| 06 Jun. 2006 | First video with working prototype [1] |
| May 2006 | A-test boards become available |
| 28 Jan. 2006 | World Economic Forum, Switzerland UNDP and OLPC Sign Partnership Agreement news release |
| 13 Dec. 2005 | Quanta Computer Inc. to Manufacture Laptop (html)(pdf) |
| 16 Nov. 2005 | WSIS, Tunisia Prototype Unveiled by UN Secretary-General Kofi Annan Photos: (Image 1)
(Image 2) (Image 3) |
| Jan. 2005 | Laptop Intiative Officially Announced at World Economic Forum, Davos, Switzerland |
PRESS
Video
(Misc. videos of the laptop can be found here.)