810 not always obeying redirects from websites
We appear to be getting an issue on the 810 where it is ignoring redirects (i.e. CGI.redirect in perl) when the page doing the redirect takes say 2 seconds plus from being requested to delivering the redirect. This does not affect any desktop browsers we have tested and we have tested most of them. So it would appear to be a bug in the 810 browser. It just sits there for ever trying to get the page - apache logs indicate it gets the page that issues the redirect but never requests the page it is redirected to. If we have no delay in the redirect it works, putting a sleep(2) in the redirect page and it stops working. Any ideas? I would imaging other people are having intermittent problems with the many web applications out there using redirects. Simon -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.516 / Virus Database: 269.20.7/1284 - Release Date: 17/02/2008 14:39 ___ maemo-developers mailing list maemo-developers@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-developers
RE: 810 not always obeying redirects from websites
Just done some more tests and the problem seems to go away on G wireless network and only be present on a B wireless network - so maybe it is related to wireless rather than browser. Has anyone else had these symptoms? I don't have a site to test with setup it is on an internal network. Simon At 10:22 18/02/2008, [EMAIL PROTECTED] wrote: Simon wrote: We appear to be getting an issue on the 810 where it is ignoring redirects (i.e. CGI.redirect in perl) when the page doing the redirect takes say 2 seconds plus from being requested to delivering the redirect. This does not affect any desktop browsers we have tested and we have tested most of them. So it would appear to be a bug in the 810 browser. It just sits there for ever trying to get the page - apache logs indicate it gets the page that issues the redirect but never requests the page it is redirected to. If we have no delay in the redirect it works, putting a sleep(2) in the redirect page and it stops working. Any ideas? I would imaging other people are having intermittent problems with the many web applications out there using redirects. Could you please provide a public url for testing purposes? (preferably one for each case described) -- No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.5.516 / Virus Database: 269.20.7/1284 - Release Date: 17/02/2008 14:39 -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.516 / Virus Database: 269.20.7/1284 - Release Date: 17/02/2008 14:39 ___ maemo-developers mailing list maemo-developers@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-developers
Re: Kiosk Mode Intranet Web Client
Here is a program and source that will make nokia 770 go full screen and a script to start the browser. i.e. run the start browser script then tell it to go fullscreen. Call it in a shell script with this preceding the program export DISPLAY=:0.0 so it knows what display you are talking about You can add the script to the inittab so it get called on boot (bear in mind the nokia really only boots when it is fully powered off. When it is powered off) and charging it is actually on (you can ssh into it and everything, its screen is just off) There is no real way of coping with a user taking it our of full screen or a different APP being in front when you run the full screen program (which will make the app on top full screen). We have noticed several issue with the nokia power on all the time - scripting it to reboot every night is a good idea. Typical issues are loss of network (even though still connected), memory leaks, browser window unexpectedly closing. Note none of this has been tested on newer nokias than the 770. The 810 uses a different browser and may well have the command line kiosk mode (full screen) switch that is so noticeably lacking in opera on the 770. Simon At 12:21 18/02/2008, F. Warnecke wrote: Hello Community, I'm new to this list and couldn't find this topic in the archive. We have Nokia 770s in use as web-clients for an intranet application. The tablets are mount stationary on the users desks and connected to power permanently. Their only job is to connect to a pre-defined WLAN and view a pre-defined website - nothing else. We realized that the tablets slow down in general after some days as they are turned on and connected to the WLAN permanently. Any idea about the reason? I also have this idea about a reduced system that starts in a kind of kiosk mode: auto-connect to this one and only WLAN, showing the webbrowser in full screen with a pre-defined home-url, with no possibility to leave the browser. Any other built-in applications (mail, chat, calculator, pdf-reader, media-player, etc) are removed from the system. I hope for more stability and speed with this. With an image of such a reduced system I also only have to flash it to all the new clients instead of configuring them by hand. Does anyone know such a solution? Another thing I can't find a solution for is to turn of the sleep-mode completely and avoid the screen to turn of the light. Any ideas for this? Thanks a lot. ___ maemo-developers mailing list maemo-developers@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-developers -- No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.5.516 / Virus Database: 269.20.7/1284 - Release Date: 17/02/2008 14:39 gcc *.c -o FullScreen -I/usr/X11R6/include -L/usr/X11R6/lib -lX11 -lXtst FullScreen Description: Binary data // Copyright Fitronics Ltd. // Define the parts of keysym.h we want to use #define XK_MISCELLANY // Include appropriate headers #include X11/extensions/XTest.h #include X11/Xlib.h #include X11/keysym.h #include string.h // Define boolean values #define TRUE1 #define FALSE 0 // Function Prototypes int main(int argc, char *argv[]); // Main function int main(int argc, char *argv[]) { // Declare vars used in the function char name[2] = :0;// Name of display to Display *dpy = XOpenDisplay(NULL); // Handle of the above display unsigned int keysym = XK_F6;// Keysym for Full Screen (F6 = Full Screen on N770) unsigned int keycode; // Keycode of the above int ret_val;// Value returned from a function call // Was a display returned? if (dpy == NULL) { printf(No display returned. Exiting.\n); return 1; } // Get the keycode for the given keysym keycode = XKeysymToKeycode(dpy, keysym); // Run the Full screen command /* Args: - Display Handle - Key Code - Pressed (true) or Releaed (false) - Milliseconds to pause before performing */ ret_val = XTestFakeKeyEvent(dpy, keycode, TRUE, CurrentTime); ret_val = XTestFakeKeyEvent(dpy, keycode, FALSE, CurrentTime); // Close the display and exit the program XCloseDisplay(dpy); return 0; } // Leave new line at the end start_browser Description: MS-Word document No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.516 / Virus Database: 269.20.7/1284 - Release Date: 17/02/2008 14:39 ___ maemo-developers mailing list maemo-developers@maemo.org https://lists.maemo.org/mailman/listinfo/maemo-developers
Re: Launch maemo browser in fullscreen mode
It does work, I've tried it. Not specific to the browser of course, can send any key to any screen, but in this case you send F6 when the browser is the front window to make it full screen. Anyone got any tips on changing X windows (i.e. removing window manager) so just the browser starts up with no other Apps? It would make the above work so much better as a kiosk. (nokia 770) Regards Simon At 18:00 21/03/2007, Johan Bilien wrote: On Tue, Mar 20, 2007, Frantisek Dufka wrote: Tomàs Jiménez Lozano wrote: There is really no way to maximize an application in general (and maemo browser in particular) emulating the HW Key press? Yes this brute force approach could work too. You could use uinput kernel driver (used with serial bluetooth keyboards) to feed proper F-key to the kernel input subsystem to emulate such key press. Could be quite easy but you need kernel module not present by default on the device. Using the XTest extension should work, something like XTestFakeKeyEvent (display, keycode, TRUE, CurrentTime); XTestFakeKeyEvent (display, keycode, FALSE, CurrentTime); -- Johan Bilien [EMAIL PROTECTED] ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers -- No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.15/728 - Release Date: 20/03/2007 08:07 -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.17/730 - Release Date: 22/03/2007 07:44 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: Launch maemo browser in fullscreen mode
Just in case anyone else tries the below - your device might go into a reboot loop. Presumably need to disable the watchdog thingy. Regards Simon At 10:24 23/03/2007, Johan Bilien wrote: On Fri, Mar 23, 2007, Simon Moore wrote: It does work, I've tried it. Not specific to the browser of course, can send any key to any screen, but in this case you send F6 when the browser is the front window to make it full screen. Anyone got any tips on changing X windows (i.e. removing window manager) so just the browser starts up with no other Apps? It would make the above work so much better as a kiosk. (nokia 770) You could use the browser as your desktop process. Replace PROG=/usr/bin/maemo_af_desktop with PROG=/usr/bin/browser in /etc/osso-af-init/maemo-af-desktop.sh. -- No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.17/730 - Release Date: 22/03/2007 07:44 -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.17/730 - Release Date: 22/03/2007 07:44 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
RE: Launch maemo browser in fullscreen mode
It is possible to compile a small App that sends the keyboard function key (F6 I think) to x windows that is equivalent to the hardware fullscreen button being pressed (someone pointed me in that direction a while ago on the list). Far from the best solution but it can work. The downside is it is hard to be sure which application is in front (I believe there is a util that can tell an app to go to the front, but not exactly fool proof). I am told it is possible to disable the hardware key but I am not quite sure how you would do this (change the key mappings file?). If you can't do that they can take it out of fullscreen, they can also press the home button. The browser crashes an enormous amount (one person mentions it is linked to the number of additional windows you open, most others think it is memory related), so you need a way to detect it crashed and reopen it. (I managed to do this in Rh9 really easily, so if browser or x-windows crash it restarts, can't seem to crack it on the nokia). Sleep mode (and also power off when charger plugged in) don't tend cause a proper boot up (it is already up) so startup scripts to start the browser don't get fired. Ideally I'd want to uninstall all non useful applications (e.g. games, email etc). Has anyone written anything to do that? I have not found any way to cut everything back to just the desktop and the web browser (i.e. remove the window manager) if someone can provide more info on that then I can probably knock up a proper kiosk mode for the Nokia and release it back and then we could have a basic kiosk and everyone can chip in to fix the other issues. There are so many fab web apps for a kiosk tablet - e.g. Museum guide (so much better than the talk in your ear ones). Kind Regards Simon At 07:15 20/03/2007, [EMAIL PROTECTED] wrote: That's right. The fullscreen browser launching is not implemented but it is planned in future (not too soon). Br, --jakub There is no way to open the opera browser into fullscreen, i spent quite a bit of time trying various solutions and even tried implementing the open embedded browser via dbus, however these calls are deprecated. The only hope is to use a different browser solution as a base. Andrew Tomàs Jiménez Lozano wrote: As a first step to achieve some kind of kiosk mode I am trying to open maemo browser at startup in fullscreen mode. I've found that, dislike desktop opera browser, maemo browser has no command line switch option (--fullscreen) to do this. I am trying now to launch browser via dbus rpc capabilities. I've found some osso-browser-interface.h document where it says the only methods available for remote calling are for opening a new window or opening a url. There is another collection of methods but they seem to be intended for embedded browser windows within a main application window. One of this embedded-only methods allows you to send a HW Key code to the browser, that is just what I am trying to do. Disgracefully I assume it only works for this embedded browsers. Does anyone know if there is a way to open maemo browser in fullscreen? Thanks in advance Tomàs Jiménez oranginalab ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers -- No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.14/727 - Release Date: 19/03/2007 11:49 -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.14/727 - Release Date: 19/03/2007 11:49 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: Device Soft Reset + App autostart
I've noticed the 770 when connected to it's charger and a USB ethernet adapter is actually still on when it has been powered down via the power button in a semi booted state - enough that I can ssh in and the eth0 has been brought up and I can x11vnc to forward the battery charging screen. However I cannot find any way to then tell the system to power up from that state (without manually pressing the power button), really useful to us for remote debugging software. I am guessing there is some command you can run that will do it. I tried the normal linux things like telinit, shutdown and halt and while they seem to do what they say on the tin none of them actually get it to boot up / reboot with the screen on as per normal. Regards Simon At 17:32 04/03/2007, Frantisek Dufka wrote: Acadia Secure Networks wrote: All, it would be useful to have a GUI capability to tell the N800 OS to reboot without having to power down. This is in fact implemented, but disabled by default. You can uncomment reboot item in /etc/systemui/systemui.xml (effective after next reboot). Then you see it as additional 'Reboot' item in device mode dialog. Frantisek ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers -- No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.8/716 - Release Date: 09/03/2007 18:53 -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.10/720 - Release Date: 12/03/2007 19:19 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Failure Rate of 770s update. now 17%
Current failure rate within the first three months of purchase is 17.2% (sample size 29, 5 failed) We have tested a seven more devices since my last email that have been at a site for 3 months (including flashing them to latest version, enabling host mode and flashing a custom kernel), all worked fine and we have one more failure at another site (version 2, no host mode or custom kernel). Of the failures 4 are the classic screen failure reported by many and 1 was dead out of the box (continuous reboot). We only have one 800 which is still working fine. I have started logging serial numbers to see if there is a higher incident of failure with certain ones. 3 of the failures have nearly identical serial numbers (last digit is different), two with similar numbers last number different are still operational. The other two failed units I don't have the serial numbers of. Leading me to the suspicion that batches of 770s may be faulty - assuming 770s in the same batch have very close serial numbers of course not enough data to be conclusive. I would give out the serial numbers of the ones that have failed but Nokia are using the serial numbers as a sort of password to their firmware so I guess they would prefer me not to. However if anyone wants to send me their failed unit serial numbers I can see if there is any correlation with serial numbers we have had fail. Regards Simon No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.10/720 - Release Date: 12/03/2007 19:19 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] New list spin-off?
Perhaps separate the 800 and 770 into separate lists as obviously a lot of post are specific to one or the other? At 12:05 31/01/2007, [EMAIL PROTECTED] wrote: Content-class: urn:content-classes:message Content-Type: multipart/alternative; boundary=_=_NextPart_001_01C74530.2D0EC2B7; x-avg-checked=avg-ok-5DD64D15 This list is growing and sometimes (like this month) is difficult to follow. What about finding a specific development area that could be discussed in a separate list, leveraging the weight of this one? Contributors interested in that area could collaborate better. A list specialized in a specific topic would be also easier to join and follow for the [EMAIL PROTECTED] developer working on that topic. Please have your say. And now some numbers for the joy of statistics: looking at http://maemo.org/pipermail/maemo-developers/http://maemo.org/pipermail/maemo-developers/ we can see that on December there was an average of 8 mails a day, while this month we are getting almost 30 emails. Of course the N800 launch has a lot to do with this increase, but nevertheless the average in 2006 was 12,3 emails/day (17,2 if you don't check this list during the weekend) ;) -- Quim Gil Maemo team ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.5.432 / Virus Database: 268.17.16/660 - Release Date: 30/01/2007 17:04 No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.432 / Virus Database: 268.17.16/660 - Release Date: 30/01/2007 17:04 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] Unreliable Large Network Transfers
Not had any problem with the network as we don't transfer big files. I think Frantisek might be on to something. We defiantly had the problem writing files to the MMC and they don't even have to be big. If you attempt to write data at a significantly faster speed than the MMC can actually do, then it just fills up the buffer, so if you carry on for a while it looks like it uses all the available memory (slowing you down), then in our case the program writing the file slowed down and things reached an equilibrium somewhere around the point of not really being usable. (we were testing a low level driver logging everything to a file). If you want to prove (although perhaps not conclusively) it is a network issue then find some way of transferring the same files but don't store them anywhere (e.g. pipe to /dev/null) so it never hits the MMC bottleneck, or perhaps put a 250MB file on an MMC and send it the other way (although presumably this will be the speed ot the MMC bottleneck from the beginning but see if it gets worse). Shame the drivers are not open source as the others have said, because if you do 'prove' it is the wireless driver there is not a lot that can be done. If it is an MMC bottleneck then perhaps best to find a way to drop the transfer speed down to that of the MMC bottleneck so your device doesn't slow down (also try faster MMC cards with the fast drivers, although I'm not sure they are stable yet). Regards Simon At 07:57 15/12/2006, Frantisek Dufka wrote: Mike Lococo wrote: I'm finding that large (200MB) transfers of data via the wifi network are extremely unreliable. They start out running at a reasonable speed (350KB/sec-600KB/sec), but within a few moments slow to less than 100KB/sec. The rates become very erratic, and the device GUI becomes very unresponsive. Are you sure network is a problem? Are you writing data somewhere? Both MMC card and iternal flash are really slow when writing. I had similar problem when writing to MMC card. First it is fast because data is buffered in RAM but then RAM is filled and as you say 'GUI becomes very unresponsive'. You can use 'top' in osso-xterrm to see cpu usage. If it is mmcqd of something like that then it is MMC writing. Enabling multiblock writes helps (a bit experimental feature). Only when multiblock writes in MMC driver are enabled then wi-fi driver really becomes a bottleneck. It theory it should be the fastest way how to get data into the device (g wlan does normally 2MB/s speeds) but sadly only ~600kb/s is reality with N770 and current driver. I wonder where exactly the bottleneck wih wi-fi is (hardware of software). Frantisek ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers -- No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.15.19/587 - Release Date: 14/12/2006 -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.15.19/587 - Release Date: 14/12/2006 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
[maemo-developers] keyboard simulator
Hi I am interested in writing a script or program (or obtaining from elsewhere an open source one) that can simulate hardware keypresses of a keyboard, in particular those of the hardware keys on the nokia. Can any one point me in the right direction. Regards Simon -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.14.16/552 - Release Date: 26/11/2006 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] keyboard simulator
Thanks, looks like it may do the job. Kind Regards Simon At 11:24 28/11/2006, Santtu Lakkala wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Simon Moore wrote: I am interested in writing a script or program (or obtaining from elsewhere an open source one) that can simulate hardware keypresses of a keyboard, in particular those of the hardware keys on the nokia. XTest is probably what you want, XTestFakeKeyEvent to be more precise. See for example: http://homepage3.nifty.com/tsato/xvkbd/events.html, google is your friend too. =) - -- Santtu Lakkala -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.2.2 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFFbBx2OjErIbnoFXIRAh3PAJ0b8pm7D4qjg/dXblFJNkh9vI9R1wCeL6Vj oKR213GHP/w+v6uScAb1d/M= =vCIw -END PGP SIGNATURE- -- No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.14.16/552 - Release Date: 26/11/2006 -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.14.16/552 - Release Date: 26/11/2006 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
Re: [maemo-developers] FlashPlayer and Opera updates?
I am guessing for Minimo you just need to look for a certain keyboard key press and map it to the function to make it go full screen? And didn't I see some little file that gives the keyboard equivalent of the Hardware keys on a per application basis? Let me know if there is anything I can do to help. Regards Simon At 11:42 24/11/2006, Antonio Gomes wrote: That is what I will be working on regarding minimo development for these days. http://tonikitoo.blogspot.com/2006/11/what-is-about-to-come.htmlhttp://tonikitoo.blogspot.com/2006/11/what-is-about-to-come.html thoughts ? On 11/24/06, Simon Moore mailto:[EMAIL PROTECTED][EMAIL PROTECTED] wrote: At 09:46 24/11/2006, Michael Dominic Kostrzewa wrote: I have similiar (opera) problems with the latest 2006 update. I've got the impression that the browser stability has been somehow altered. There is no rule in the crashes but it seems to happen after going through a few pages. Did you have the problem before the latest 2006 update? I have noticed the problem both pre and post latest update. I would be interested in hearing other people's views on Opera's stability. (I think it is very poor) The latest minimo seems much better than it was when I last tested it. I have not had time to rigorously test it though (no support for plugins like flash at the moment). Perhaps Nokia could put some development time into making Minimo port better and compile a flash player for it (version 6 would be fine)? I have no time at the moment to try to hack something together - but it may well be the flash .SO for Opera could be made to work for Minimo, perhaps someone who knows more about it could indicate if it is worth trying to do this, or whether the flash player needed to be compiled with Minimo headers (in which case Nokia will need to do it)? Regards Simon -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.14.14/548 - Release Date: 23/11/2006 ___ maemo-developers mailing list mailto:maemo-developers@maemo.orgmaemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers -- --Antonio Gomes No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.14.14/548 - Release Date: 23/11/2006 No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.14.14/548 - Release Date: 23/11/2006 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
[maemo-developers] FlashPlayer and Opera updates?
I don't suppose there is an updated flashplayer anywhere? 7 or better (or source code to compile one). All I can find is pre-compiled .SO for x86 Linux (7 and 9beta). On that note are there any plans to update Opera in future OS releases? - I checked with Opera and they say only Nokia can release updates to Opera on Nokia 770 (i.e. you can't buy an upgrade). Could really do with version 8.5 or better for more complete AJAX support (Maemo does it, but is not really production ready as a browser, having said that Opera crashes quite a bit too for no apparent reason). Regards Simon -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.14.3/531 - Release Date: 12/11/2006 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers
[maemo-developers] Working Peripherals for Wired Networking and Serial and powering Nokia from Hub
I thought I would post this so that other people might find it via google when they hit similar issues to us trying to figure out which USB peripherals work out of the box (we tried a lot that claimed to work in Linux and don't work in 2.6 kernel with Manufacturers drivers let alone on the Nokia). No doubt others can improve on the things I have said. Top Tip see if the hardware is listed in the source code of the USB modules that are compiled on the Nokia - then spend weeks googling to find out who actually uses that obscure Chip (e.g. pl2303) that everyone says works. If it is in the source code for the 2.16 kernel but not compiled on the nokia you can presumably compile it as a module and insmod it - unfortunately I don't know how to compile just the modules - a how to from someone please? HowToSetupVariousPerihperals You may wish to first look at http://maemo.org/maemowiki/Easy_USB_Host_Hub_setup http://maemo.org/maemowiki/HowDoiBecomeRoot http://maemo.org/maemowiki/HowDoiBecomeRoot2 http://maemo.org/maemowiki/HowTo_SwitchToHostMode In essence you will need to be able to become root and you will need to have used the flasher to enable host mode (then use http://maemo.org/maemowiki/HowTo_SwitchToHostMode to switch back to non host mode whenever you need on a temporary basis) You will probably want to setup some kind of file networking e.g. to a windows share so you can put files on the device more easily. http://maemo.org/maemowiki/HowToAccessWindowsCIFS - Note does not work with newest 2006 kernel, try samba client instead Everything below assumes you have setup a powered USB hub that provides power back to the Nokia and the Nokia is in USB host mode. Working Devices (out of the box) HUBS (all should work, as long as you figure a way to do the power thing, Y cable method is best) Targus Mini USB 2.0 4-Port Hub comes with Y cable that can do the power thing, but you will need to buy a power supply for it. PCWB you can buy a 2.5 amp 5v 'Digital Camera Adapter' PSU for targus Hub from Maplin for £10 tip positive, so no need to solder D-Link DUB H4 - Hub - 4 ports - Hi-Speed USB 10 cm x 6 cm x 2.6 cm PCWB (has power supply but no Y cable) USB Keyboards (all should work when connected to powered hub) USB to Serial (RS232) anything using the pl2303 chip (the on board serial port uses this chip, so it is already in the kernel and will load the module when you plug in, type dmesg to see it detected) USB to Serial 9 Pin Male Adaptor Maplin code ZP43W (uses pl2303 chip) TRENDnet TU-ET100C USB 10/100 Fast Ethernet Adapter (Apparently not tested by me) A Cyberpower 4-port USB 2.0 powered hub (Apparently not tested by me) dlink dub-e100 (rev A only not the new B, see below) (Apparently not tested by me) Not working out of the box StarTech.com 2 Port USB to RS-232 Serial DB9 Adapter PCWB unsuccessful at making it work in Ubuntu with Manufacturer drivers Keyspan High Speed USB Serial Adapter - Serial adapter - USB - RS-232 - serial - Probably could be made to work by compiling module and insmod on device. Belkin USB to Serial Adaptor F5U109 (mct_u232.o is available for the kernel so probably could be compiled and loaded) dlink dub-e100 (rev B) (newer kernel should support it, can be made to work by compiling the kernel - you will need to update ASIX.C USBNET.C USBNET.H to the 2.6.18 Kernel - or just overwrite them and then apply various patches to ASIX.C and USBNET.C to make the REV B work - details in another mail, as I can do it with the whole kernel but really want to do it with a module.) Various useful cables StarTech.com - USB cable - 4 PIN USB Type A (M) - 5 pin mini-USB Type B (M) - 1.8 m ( USB / Hi-Speed USB ) PCWB StarTech.com - USB gender changer - 4 PIN USB Type A (F) - 4 PIN USB Type A (F) PCWB StarTech.com - USB adapter - 4 PIN USB Type A (F) - 4 PIN USB Type B (M) PCWB 1 meter USB 2.0 Cable for 2.5in USB HDD www.usbtech.co.uk Powering the Nokia from the USB hub (why plug in two 5v PSUs when one will do? and why on earth didn't they design it to power/recharge from it's USB socket anyway?) Obviously this is somewhat destructive but you may have more luck sourcing the exact parts. Heat-Shrinkable Sleeving BF88V - 3.5mm Heatshrink 1m - £1.63 BF89W - 5.0mm Heatshrink 1m - £1.73 (actually maybe some smaller stuff too) 2500mA Digital Camera Power Supply (power the targus hub from PCWB) L69BQ - Digital Camera SMPS - £9.99 USB 2.0 A to B Leads (or you could use PN86T - USB A Plg- £1.17 if ever in stock) L00BT - USB2 A(M)-B(M) 1.5m - £7.99 Cut the USB lead in half, cut your nokia power lead in half (I have tried to source the jack they use and bought everyone that Maplin supply and none fit). Expose the USB inner cables (discard the outer shielding). Put the big heatshrink onto the USB cable and two small ones on the red and black cable before soldering. Now you
Re: [maemo-developers] Patching Nokia 770 Kernel to get DLINK DUB-E100 Rev B to work
Fantastic I'll give just doing the modules a go and patching the source properly. (I take it no Nokia specific changes get messed up with the patch so why haven't they release the code yet, surely they don't need lawyers for minor updates to source code). Every time I tried using make it always gave me errors and didn't work, it only worked with dpkg-buildpackage -rfakeroot -uc -us -sa -d . Ift you typed make, then tried the dpkg-buildpackage if didn't work either. I assumed it was something weird about scratchbox - what is dpkg-buildpackage anyway I have not come across it before, something about scratchbox or more general debian? Your instructions are certainly different from the things I tried, so no doubt will work like a dream :-) Regards Simon -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.409 / Virus Database: 268.14.0/524 - Release Date: 08/11/2006 ___ maemo-developers mailing list maemo-developers@maemo.org https://maemo.org/mailman/listinfo/maemo-developers