Re: Re: XO battery/performance[ Devel Digest, Vol 76, Issue 21]
Could you please do another test, to verify that screen is not configuring the serial port any differently with or without ,n,8,1. You don't need to connect the two laptops together, just use the one you use as the serial terminal. The test uses the stty program to display the serial port configuration. Here you are. However I need to have the the 2 XOs connected because without it /dev/ttyUSB0 does not exist and if I connect just the adapter then stty fails with resource busy $ screen /dev/ttyUSB0 115200 * Today the output was readable! :-? [screen is terminating] [olpc@xo-74-39-1a ~]$ stty --all /dev/ttyUSB0 speed 115200 baud; rows 0; columns 0; line = 0; intr = ^C; quit = ^\; erase = ^?; kill = ^H; eof = ^D; eol = undef; eol2 = undef; swtch = undef; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 100; time = 2; -parenb -parodd cs8 -hupcl -cstopb cread clocal -crtscts -ignbrk brkint ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl ixon -ixoff -iuclc -ixany -imaxbel -iutf8 -opost -olcuc -ocrnl -onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0 -isig -icanon iexten -echo echoe echok -echonl -noflsh -xcase -tostop -echoprt echoctl echoke $ screen /dev/ttyUSB0 115200,n,8,1 [screen is terminating] [olpc@xo-74-39-1a ~]$ stty --all /dev/ttyUSB0 speed 115200 baud; rows 0; columns 0; line = 0; intr = ^C; quit = ^\; erase = ^?; kill = ^H; eof = ^D; eol = undef; eol2 = undef; swtch = undef; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 100; time = 2; -parenb -parodd cs8 -hupcl -cstopb cread clocal -crtscts -ignbrk brkint ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl ixon -ixoff -iuclc -ixany -imaxbel -iutf8 -opost -olcuc -ocrnl -onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0 -isig -icanon iexten -echo echoe echok -echonl -noflsh -xcase -tostop -echoprt echoctl echoke [olpc@xo-74-39-1a ~]$ screen /dev/ttyUSB0 9600,n,8,1 **HERE I get the questionmark in back diamond output** [screen is terminating] [olpc@xo-74-39-1a ~]$ stty --all /dev/ttyUSB0 speed 9600 baud; rows 0; columns 0; line = 0; intr = ^C; quit = ^\; erase = ^?; kill = ^H; eof = ^D; eol = undef; eol2 = undef; swtch = undef; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 100; time = 2; -parenb -parodd cs8 -hupcl -cstopb cread clocal -crtscts -ignbrk brkint ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl ixon -ixoff -iuclc -ixany -imaxbel -iutf8 -opost -olcuc -ocrnl -onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0 -isig -icanon iexten -echo echoe echok -echonl -noflsh -xcase -tostop -echoprt echoctl echoke ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Re: XO battery/performance[ Devel Digest, Vol 76, Issue 21]
On Tue, Jun 12, 2012 at 11:07:44PM -0700, Yioryos Asprobounitis wrote: James wrote: Could you please do another test, to verify that screen is not configuring the serial port any differently with or without ,n,8,1. You don't need to connect the two laptops together, just use the one you use as the serial terminal.? The test uses the stty program to display the serial port configuration. Here you are. Thanks. Your results prove the ,n,8,1 is ineffective, and it must have been line noise that caused your unreadable output last time. However I need to have the the 2 XOs connected because without it /dev/ttyUSB0 does not exist and if I connect just the adapter then stty fails with resource busy stty will fail with device or resource busy if screen is running at the time that stty is run. Or, modemmanager may have been trying to send AT commands to the serial adapter to see if it is a modem. This is one reason why I don't like using our OLPC Fedora builds as serial terminals! http://wiki.laptop.org/go/Serial_adapters#Linux has a note about modemmanager. So it probably wasn't connecting the two laptops that resolved the device or resource busy error for you. It was the passing of time, enough to let modemmanager finish probing. $ screen /dev/ttyUSB0 115200 * Today the output was readable! :-? Good. Therefore the problem was line noise, probably caused by the contacts bouncing as they were mated, or by an earth loop. [olpc@xo-74-39-1a ~]$ screen /dev/ttyUSB0 9600,n,8,1 **HERE I get the questionmark in back diamond output** Good. Totally expected, because the EC would be transmitting at 115200 baud, with the serial adapter configured for 9600 baud, and the resulting garbage data is translated by Terminal into black diamonds. (I didn't think to warn you of this, because I had planned that the tests you did were without any connection between the two laptops. Thanks for taking the initiative.) You can reduce the chance of contact noise derailing the transmission by connecting the two laptops in a specific order: - shutdown the target laptop, - remove power cable and batteries from the target laptop, - connect serial cable between laptops, - restart screen, - insert battery into the target laptop, which will start the EC running, and data will begin to appear, - proceed with your test. This ensures that at the time the contacts are mated, there is no software reading the data, and no data being transmitted. -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Re: XO battery/performance[ Devel Digest, Vol 76, Issue 21]
You can reduce the chance of contact noise derailing the transmission by connecting the two laptops in a specific order: or maybe adding ,n,8,1 in the command ?... (I still think is a good idea to add it in the wiki. Even with the notion that although not necessary may reduce unexpected results stemming from line noise or race with the modem manager ;) ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Announcing Q2F12 for XO-1
http://wiki.laptop.org/go/OLPC_Firmware_q2f12 This release contains a new .os command for checking what operating system build is installed, automatic power down if the lid is closed at the ok prompt, and a rework of the test menu to assist deployment repair centres. Test reports are encouraged. Fixes - add visual response to escape key when in secure mode, ticket #11609, - fix to vocabulary search order while in telnetd, ticket #11926, - rework of test menu for USB ports, ticket #11845, - keep menu visible after automatic tests, ticket #11847, - restore audio test sweep volume to original level, ticket #11846, - power down if lid closed while at ok prompt, ticket #11095, - fix SD card detection, ticket #11844, - test menu, enable ALPS touchpad driver, ticket #11902, - add .os command for test bed management, - remove USB serial adapter keyboard support, ticket #11871 and ticket #11887, - tidy test /memory display by copying from XO-1.75 implementation. Known Problems - USB drives with embedded hubs, or devices attached to hubs, may prevent the laptop from booting. This is being tracked at #11931, and is not a recent regression. -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Re: XO battery/performance[ Devel Digest, Vol 76, Issue 21]
On Wed, Jun 13, 2012 at 12:16:13AM -0700, Yioryos Asprobounitis wrote: You can reduce the chance of contact noise derailing the transmission by connecting the two laptops in a specific order: or maybe adding ,n,8,1 in the command ?... No, that will make no difference. (I still think is a good idea to add it in the wiki. Even with the notion that although not necessary may reduce unexpected results stemming from line noise or race with the modem manager ;) Heh. ,n,8,1 should not be added to the Wiki for the usage of the screen program, because it does nothing, and I would not have the Wiki deceive anybody. The line noise depends on the power supplies being used, and the amount of isolation, and the way in which the contacts are mated ... so many variables, if we documented them all we'd have a tutorial on using serial ports ... in an age where they are rarely used by most people. -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: olpc.fth question
Jerry, James and Martin: Adam and I thank you all ... a lot We are now 100% operational using 1 USB stick to update all versions of XO. We will add some more exception handling and 1.75 specifics to the procedures once we return to Canada, but the combination of OOB 4.1 and the olpc.fth boot are making the frequent process of updating/enhancing things while here in Kenya just fly!!! Cheers, KG On Tue, Jun 12, 2012 at 8:00 PM, James Cameron qu...@laptop.org wrote: On Tue, Jun 12, 2012 at 11:08:29AM -0400, Kevin Gordon wrote: Disclaimer: Newbie Forth question :-) Always welcome. We are trying to create a consolidated unsecured update stick. I worked on a secured update drive last week, so the techniques are on my mind. [...] So for those coming from a non-Forth background, we have hit a road block. Is there perhaps a way to store a 'possible' command into a variable then execute that 'variable' as a command, thereby perhaps bypassing any of the apparent syntax error checking? Unexpected end-of-line is the most common result from attempting to call within an if statement. Or, we get copy-nand? on the 1.5 or fs-update? on the 1.0 when the command exists in the source - whether it will actually get 'called' or not ,based on the variable containing the machine type.. evaluate or eval is a word that expects a string descriptor on the stack, and then executes the string as if it were typed. : eval ( adr len -- ) ... ; For example: ok 8 . eval 8 ok or : install-xo-1 copy-nand u:\fs.img eval ; The string can be assembled from pieces rather than from literals. You may find an example of that in the power log collector on the wiki, which assembles filenames. If there is a possibility that the evaluated command may fail, you should catch the exception and handle it. Use catch for that. Good reference for catch and throw: http://www.complang.tuwien.ac.at/anton/euroforth/ef98/milendorf98.pdf For example: : install-xo-1 copy-nand u:\fs.img ( adr len ) ['] eval( adr len 'eval ) catch ( ??? ??? exception# | 0 ) if ( ??? ??? ) 2drop( ) . copy-nand failed, press any key key drop then( ) ; You might also place the exception handler higher up. We also have $fs-update in later XO-1.5 and XO-1.75 versions, so that eval is not needed. There is no $copy-nand . -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Developer XO laptop loan or buy - Speakeasy project
On Tue, Jun 12, 2012 at 9:15 PM, Lester Leong lester.ble...@gmail.comwrote: As for Javascript, how? Javascript can't handle backends without some significant running around - everything's gotta be database driven. I think you need to look again at modern Javascript/HTML5 toolkits. There are databases. There are routers. You don't need anything else. --scott -- ( http://cscott.net ) ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Developer XO laptop loan or buy - Speakeasy project
This thread has gone pretty long and deep into the bikeshedding realm. I am surprised nobody posted the answer to the original question of getting an XO for a developer: http://wiki.laptop.org/go/Contributors_program However Lester thinks the programming should go, if it is going to go on XOs, that's how to get one. Of course there are ways to program and test on other hardware -- use Fedora with Sugar desktop or jhbuild. m On Wed, Jun 13, 2012 at 8:52 AM, C. Scott Ananian csc...@laptop.org wrote: On Tue, Jun 12, 2012 at 9:15 PM, Lester Leong lester.ble...@gmail.com wrote: As for Javascript, how? Javascript can't handle backends without some significant running around - everything's gotta be database driven. I think you need to look again at modern Javascript/HTML5 toolkits. There are databases. There are routers. You don't need anything else. --scott -- ( http://cscott.net ) ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel -- martin.langh...@gmail.com mar...@laptop.org -- Software Architect - OLPC - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: olpc.fth question
Building upon Jerry's message, you may be interested in our One Education USB (formerly called XO-AU USB): https://dev.laptop.org.au/projects/xo-au-usb/ The idea is to have a single USB stick with many tools that may be needed in the field. It is designed for use by (non-technical) teachers to manage their classroom deployments. You can download a working version from http://mirror.aarnet.edu.au/pub/olpc-au/XO/OE-USB/1.1/ The base version contains no OS, but the xo15 version contains OLPC Australia's latest XO-1.5 image. To use, extract the zip file directly to the root of a USB drive. Then insert into a developer-unlocked XO-1.5 and boot. You should get a boot menu from the stick. Sridhar On 13 June 2012 21:23, Kevin Gordon kgordon...@gmail.com wrote: Jerry, James and Martin: Adam and I thank you all ... a lot We are now 100% operational using 1 USB stick to update all versions of XO. We will add some more exception handling and 1.75 specifics to the procedures once we return to Canada, but the combination of OOB 4.1 and the olpc.fth boot are making the frequent process of updating/enhancing things while here in Kenya just fly!!! Cheers, KG On Tue, Jun 12, 2012 at 8:00 PM, James Cameron qu...@laptop.org wrote: On Tue, Jun 12, 2012 at 11:08:29AM -0400, Kevin Gordon wrote: Disclaimer: Newbie Forth question :-) Always welcome. We are trying to create a consolidated unsecured update stick. I worked on a secured update drive last week, so the techniques are on my mind. [...] So for those coming from a non-Forth background, we have hit a road block. Is there perhaps a way to store a 'possible' command into a variable then execute that 'variable' as a command, thereby perhaps bypassing any of the apparent syntax error checking? Unexpected end-of-line is the most common result from attempting to call within an if statement. Or, we get copy-nand? on the 1.5 or fs-update? on the 1.0 when the command exists in the source - whether it will actually get 'called' or not ,based on the variable containing the machine type.. evaluate or eval is a word that expects a string descriptor on the stack, and then executes the string as if it were typed. : eval ( adr len -- ) ... ; For example: ok 8 . eval 8 ok or : install-xo-1 copy-nand u:\fs.img eval ; The string can be assembled from pieces rather than from literals. You may find an example of that in the power log collector on the wiki, which assembles filenames. If there is a possibility that the evaluated command may fail, you should catch the exception and handle it. Use catch for that. Good reference for catch and throw: http://www.complang.tuwien.ac.at/anton/euroforth/ef98/milendorf98.pdf For example: : install-xo-1 copy-nand u:\fs.img ( adr len ) ['] eval ( adr len 'eval ) catch ( ??? ??? exception# | 0 ) if ( ??? ??? ) 2drop ( ) . copy-nand failed, press any key key drop then ( ) ; You might also place the exception handler higher up. We also have $fs-update in later XO-1.5 and XO-1.75 versions, so that eval is not needed. There is no $copy-nand . -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Developer XO laptop loan or buy - Speakeasy project
Scott - could you point me in the right direction as far as a good JS/HTML5 framework? I've only ever done node.js to interface with DB's, and that's on the server side. I'm assuming you're talking about application frameworks. As far as obtaining an XO, I may just end up purchasing an XO unit from ebay, although I only saw XO-1's there. That may be good enough to get going right away, as the it may take several weeks to ship when going through the Contributors program. If anyone knows of a good place to purchase XO-1.5 units, I'm all ears. Thanks! Lester On Wed, Jun 13, 2012 at 8:58 AM, Martin Langhoff martin.langh...@gmail.com wrote: This thread has gone pretty long and deep into the bikeshedding realm. I am surprised nobody posted the answer to the original question of getting an XO for a developer: http://wiki.laptop.org/go/Contributors_program However Lester thinks the programming should go, if it is going to go on XOs, that's how to get one. Of course there are ways to program and test on other hardware -- use Fedora with Sugar desktop or jhbuild. m On Wed, Jun 13, 2012 at 8:52 AM, C. Scott Ananian csc...@laptop.org wrote: On Tue, Jun 12, 2012 at 9:15 PM, Lester Leong lester.ble...@gmail.com wrote: As for Javascript, how? Javascript can't handle backends without some significant running around - everything's gotta be database driven. I think you need to look again at modern Javascript/HTML5 toolkits. There are databases. There are routers. You don't need anything else. --scott -- ( http://cscott.net ) ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel -- martin.langh...@gmail.com mar...@laptop.org -- Software Architect - OLPC - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Developer XO laptop loan or buy - Speakeasy project
On 2012-06-13, at 17:37, Lester Leong wrote: Scott - could you point me in the right direction as far as a good JS/HTML5 framework? I've only ever done node.js to interface with DB's, and that's on the server side. I'm assuming you're talking about application frameworks. The Lively Kernel is a JS/HTML5 environment very much in the spirit of Sugar: http://www.lively-kernel.org/ It allows you to examine and extend the code right inside your web browser. Works on iPad, too. As far as obtaining an XO, I may just end up purchasing an XO unit from ebay, although I only saw XO-1's there. That may be good enough to get going right away, as the it may take several weeks to ship when going through the Contributors program. As others pointed out, you can download Sugar today and get going right away: http://wiki.sugarlabs.org/go/VirtualBox - Bert - ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Developer XO laptop loan or buy - Speakeasy project
On Wed, Jun 13, 2012 at 11:37 AM, Lester Leong lester.ble...@gmail.com wrote: Scott - could you point me in the right direction as far as a good JS/HTML5 framework? Keep in mind that _today_ XOs don't ship with a workable JS runtime environment other than the webbrowser. We may do so in the future, as JS is getting more and more interesting. But unclear when that specific bit of future will land. You wrote a wikipage that talks about use in a deployment, and it sounds like it'd be needed soon. If so, Python is your friend. m -- martin.langh...@gmail.com mar...@laptop.org -- Software Architect - OLPC - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Outdoor Light Sensor
We are looking for a better place for the outdoor light sensor in a future laptop, where we have a chance to make minor changes in the mainframe tooling. The problem with the current location is: interference from LEDs (noticeably the storage LED, with which it shares a package and light-guide) and interference from the display backlight, which shines through the back of the display and can easily be brighter than room lighting in the current setup. Suggestions ? wad ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Outdoor Light Sensor
On Wed, Jun 13, 2012 at 9:27 PM, John Watlington w...@laptop.org wrote: We are looking for a better place for the outdoor light sensor in a future laptop, where we have a chance to make minor changes in the mainframe tooling. The problem with the current location is: interference from LEDs (noticeably the storage LED, with which it shares a package and light-guide) and interference from the display backlight, which shines through the back of the display and can easily be brighter than room lighting in the current setup. Suggestions ? wad Below one of the speakers? Or could a sensor even be suspended in front of a speaker? --Fred ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Outdoor Light Sensor
The rear facing storage LED is of less use, since when storage latency is important to a user they are typically also operating the unit from the front. There is no display backlight leakage that I can see through the rear facing light pipes. However, this mounting point would capture light behind the laptop, instead of light on the front. This might also require two drivers for the storage LEDs instead of one. Otherwise, having disassembled the front face and thought through the mainframe volumes ... the only other place I can think of for a light pipe is ... Below the left lone USB socket, using the same front panel structure as for the microphone LED. It looks to be about the same distance as the microphone LED is from the microphone mounting centre. However, the forward facing surface of the motherboard is certainly bathed in backlight. We might also do both, wiring the sensors in parallel. -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Outdoor Light Sensor
Putting the light sensor only on the rear could run into issues if the screen is facing the Sun and the rear of the XO creates a shadow. An ugly yet isolated place to put the OLS might be in the neck of the laptop. But there isn't much room in that area. You would need two openings since the neck doesn't rotate for ebook mode, and therefore the need to monitor the ebook switch or similar to know which direction to trust. It might also be possible to add one or two OLS sensors to the bottom assembly -- one above the keyboard near the top of the slant, the other (if necessary) in the rear above the battery. But fixed sensors in the bottom would not be exposed to light at the same angle as the screen. In any case if we move the OLS away from an obvious hole or light pipe, children are going to have to know not to cover it with stickers and similar. On Wed, Jun 13, 2012 at 10:06 PM, James Cameron qu...@laptop.org wrote: The rear facing storage LED is of less use, since when storage latency is important to a user they are typically also operating the unit from the front. There is no display backlight leakage that I can see through the rear facing light pipes. However, this mounting point would capture light behind the laptop, instead of light on the front. This might also require two drivers for the storage LEDs instead of one. Otherwise, having disassembled the front face and thought through the mainframe volumes ... the only other place I can think of for a light pipe is ... Below the left lone USB socket, using the same front panel structure as for the microphone LED. It looks to be about the same distance as the microphone LED is from the microphone mounting centre. However, the forward facing surface of the motherboard is certainly bathed in backlight. We might also do both, wiring the sensors in parallel. -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Outdoor Light Sensor
Please also take into account the possible influence of the user's hand position. I've seen tablets where the display illumination changes as the user shifts how he holds the unit. mikus ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Outdoor Light Sensor
On Wed, Jun 13, 2012 at 9:27 PM, John Watlington w...@laptop.org wrote: Suggestions ? wad Entirely impractical, but one must wonder about turning the wiggling ears into eye stalks :-) On a more serious note, other than the OLPC folks, the people who run repair centers have probably had more facetime with the guts of the XO than most and seen the worst that kids can do to them. It would be nice to hear from them. cjl ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
[no subject]
http://www.eadvancepayday.com/blog/wp-content/themes/BLANK-Theme/googlemail.html___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Outdoor Light Sensor
On Thu, Jun 14, 2012 at 12:07:36AM -0400, Chris Leonard wrote: Entirely impractical, but one must wonder about turning the wiggling ears into eye stalks :-) I did. Too many parts to change. Besides, the ears can be in any orientation. I also wondered about using the backlight as a sensor. This would require the backlight to be turned off for the sample time, so it would need a much reduced sample time, to avoid perception of flicker, which suggests circuit design with: - gating to turn off the drive current fully, - gating for a faster charge path, rather than rely on a microcontroller pin configured for output, - gating for a faster discharge path with a series resistance, rather than relying on microcontroller input pin current, - merging into the existing drive circuit. But there's one irritating thing about white LEDs used in the backlight, and that is that they are really blue LEDs with a YAG phosphor, and the phosphor glows for enough time to destroy the idea. On a more serious note, other than the OLPC folks, the people who run repair centers have probably had more facetime with the guts of the XO than most and seen the worst that kids can do to them. It would be nice to hear from them. Yes, their involvement on this mailing list would be valuable, so you'll have to let some through without grabbing them for translation duties. ;-} -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Testing wanted: XO-1 USB compatibility Q2F12jb #11931
G'day, Thanks to many testers over the years, we have finally figured out a cause of incompatibility between Open Firmware and some common USB devices, in particular hubs and USB FLASH drives with embedded hubs. We have a fix, thanks to Bert and Mitch, but I'd like to see some wider testing on XO-1. There's a small possibility that there are devices which are affected by the fix. This build of OpenFirmware is a temporary release for testing compatibility with various USB devices during device or hub probe: http://dev.laptop.org/~quozl/q2f12jb.rom Do not use this build on a secured laptop without first verifying that your USB drives work with this build. Previous versions reset the port, then asked the device to set an address before reading the device descriptor. Some devices did not return a device descriptor, resulting in errors or hangs before boot. This test build resets the port, then reads the device descriptor, then asks the device to set an address. Also, any previous device descriptor in memory is destroyed. Insert a device and then use the probe-usb or p2 command to test. Examine the output and make sure the correct devices have been discovered. Restart Open Firmware if you remove any devices, otherwise the devices that were seen continue to be listed. Testing with Linux or Sugar is not necessary, as Linux uses device drivers in the kernel. See http://dev.laptop.org/ticket/11931 for further detail. When you are finished with the testing, you can reflash to Q2F12 or later. Q2F12 is the latest at time of writing. Please report your results, thanks! -- James Cameron http://quozl.linux.org.au/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel