Hello
I try to install Cisco AnyConnect on my Touch Pro (FRX07)
but when it start, I have error: failed to connect to VPN service.
Can you say me if you can run it, and how?
I want diag my problem, but log are not very explicite
Thank's for your help
I join the apk I use.
and my debug log
seems it's buz the bundle still not supports VPN
likounin said:
seems it's buz the bundle still not supports VPN
Click to expand...
Click to collapse
We've never supported VPN's. If there's some way to make them work, we will gladly implement the code. As always, patches welcome.
arrrghhh said:
We've never supported VPN's. If there's some way to make them work, we will gladly implement the code. As always, patches welcome.
Click to expand...
Click to collapse
ok, thank's. I understand now.
I am new on Android, but I known Linux
What patch may we do? kernel, daemon, ...
ejantzen said:
ok, thank's. I understand now.
I am new on Android, but I known Linux
What patch may we do? kernel, daemon, ...
Click to expand...
Click to collapse
First, if I knew it would already be patched lol. Second, I don't have any VPN to test with - so you're going to have to tell us what needs patched. It seems like a userland issue, but I honestly don't know. Perhas some kernel module is missing? Again, I don't know for sure. I don't have any way to test it, and I think that's the problem most have - they have no way of testing VPN, and I think most probably use OpenVPN if it's in their home. At work, it's probably mostly PPTP.
So if you find something to patch, please submit it. If you can't... well then we just won't be able to support VPN's. Pretty minor feature, but I'm sure it would be nice for some.
arrrghhh said:
It seems like a userland issue, but I honestly don't know. Perhas some kernel module is missing? Again, I don't know for sure.
Click to expand...
Click to collapse
ok, thank's
but on the debig file, we can't see where is the problem?
ejantzen said:
ok, thank's
but on the debig file, we can't see where is the problem?
Click to expand...
Click to collapse
If by "we" you mean me, no. I'm no expert on the matter of deciphering logs and determining what exactly needs to be fixed. I can usually pinpoint problems, but pinpointing exactly which element needs fixing... I'm pretty new to.
In my horrible estimation, these three lines are where it starts going badly:
Code:
09-04 10:27:28.770 I/vpnagent( 1847): Function: scanAvailableModule File: PluginLoader.cpp Line: 745 Scanning plugin libvpncommoncrypt.so
09-04 10:27:28.770 W/vpnagent( 1847): Function: STGetProcAddress File: HModuleMgr.cpp Line: 263 Invoked Function: dlsym Return Code: 0 (0x00000000) Description: Symbol not found:
09-04 10:27:28.770 E/vpnagent( 1847): Function: scanAvailableModule File: PluginLoader.cpp Line: 783 Invoked Function: CHModuleMgr::STGetProcAddress Return Code: -33554427 (0xFE000005) Description: GLOBAL_ERROR_NULL_POINTER
The "Scanning plugin libvpncommoncrypt.so" portion really interests me. Does this lib exist? Have you been able to find perhaps another version of it? I know very little about VPN's... never had the desire to use one on a phone. I don't even use 'em on computers really... my office has a VPN that requires a PC, and is far from a 'traditional' VPN.
Looking further thru the log, it seems the app is just going thru libs. It would appear we're either missing these libs, or the libs we have aren't playing nicely. I keep seeing the same "Symbol not found" errors... I'm betting we simply don't have the right libs for this type of VPN app.
thank's
I will try in this direction.
Related
Hi I just got SynCE 0.90 working fine with kernel 2.6.x
In case you didn't know synCE is the linux (open source) version of Activesync
homepage at http://synce.sourceforge.net/
I used this with pocket PC 2003SE on blueangel
It also works fine with synce-kde, just like activesync
here is the updated snippet from the wiki:
I ran it on a 2.6.9 kernel and I added this to my ipaq.h:
(look for this)
#define HTC_VENDOR_ID 0x0bb4
#define HTC_WALLABY_ID 0x00ce
#define HTC_HIMALAYA_ID 0x0a02
(new part here)
#define HTC_BLUEANGEL_ID 0x0a05
(end new part)
And this to my ipaq.c (/usr/src/linux/drivers/usb/serial/)
(look for this)
{ USB_DEVICE(HTC_VENDOR_ID, HTC_WALLABY_ID) },
{ USB_DEVICE(HTC_VENDOR_ID, HTC_HIMALAYA_ID) },
(new part here)
{ USB_DEVICE(HTC_VENDOR_ID, HTC_BLUEANGEL_ID) },
(end new part)
In the ipaq_id_table to be precise.
hope this helps (and the wiki is updated when online)
Barry
Uau!!!
Do you know is everything is syncable? (calendar, contact,..)
cannot make it work under suse 9.1 kernel 2.6.8.x
I've been trying to make Syncce under Suse 9.1 with kernel 2.6.8.x work. But no luck. Any suggestions or tips?
synce can sync with kde using Multisync. http://synce.sourceforge.net/synce/multisync_guide.php
I havn't used it myself, but I read good things.
As for the problem with Suse 9.1 with kernel 2.6.8.x you would have to be more specific.
did you modprobe usbserial and ipaq after modifyig the source, rebuilding the kernel, and rebooting?
If you did all of the above and you still have problems then post your dmesg output with the above loaded and the device pluged in.
cool! is this bluangel specific or could work also with xda-mini, ...?
Do u use synce with usb or bluetooth? tnx
you will ned to get the device id for the mini to make it work, but thats easy
the blueangel is 0x0a05, you see this when you use the tool usbview
click on ipaq, and then look for
Product Id: 0a05 (this will say a different device id)
change the above code, and replace the device id of the mini.
I have successfully used both bluetooth and usb to sync my pda.
check out http://synce.sourceforge.net/
could you make a howto:
ive been trying everything
here is what i have
i have the Audiovox XV6600 w/cam (default rom) got it like 1 month ago.
im running slack 10.2 with the test26.s kernel u can pick upon install (its 2.6.13)
i leave pda in regular mode and connect it through usb after i am in kde desktop
then i switch the pda mode to wmodem.
with kppp (dial up app) i select software xon/xoff mode
i select script
i try and find my modem and put it on dev/tty0 or /dev/tty1 and querry it
usually it does querry but all the commands come back blank
i try to log in and it says modem ready and then it goes to "initializing modem" and its not froze or hung but it just stays there
note: i do not see the 3 little lights in wmodem flash when i querry or try to connect.
with 2.6.13 i dont think you have to edit the ipaq.h or ipaq.c files in /usr/src/linux/drivers/usb/serial
when i compile a kernel what option makes the ipaq module load?
ok thanks
anyone gonna try and help a fellow linuxer out?
m16-maniac said:
anyone gonna try and help a fellow linuxer out?
Click to expand...
Click to collapse
Hi,
not so much at once. We are already running Linux
on Blueangel now
You don't need to recompile anything. Just
modprobe ipaq debug=1 vendor=XXXX product=YYYY
with vendor and product numbers.
Works here even on amd64 linux.
Can tell you the steps necessary to do from
the command line.
Don't know about KDE though.
cr2 said:
m16-maniac said:
anyone gonna try and help a fellow linuxer out?
Click to expand...
Click to collapse
Hi,
not so much at once. We are already running Linux
on Blueangel now
You don't need to recompile anything. Just
modprobe ipaq debug=1 vendor=XXXX product=YYYY
with vendor and product numbers.
Works here even on amd64 linux.
Can tell you the steps necessary to do from
the command line.
Don't know about KDE though.
Click to expand...
Click to collapse
well linux detects the modem with np and it automatically loads the ipaq module. i can verify this by doing the 'lsmod' command and its in there
im using the HARRIER model but should work close to the same.
ive tried to configure with pppsetup and also kppp from kde and havent had to much luck so far.
i know i have a symlink i create that goes like this
/dev/modem = /dev/ttyUSB0
with verizon i dont think it uses pap/chap as you can logon with just the #777 and no userass
i dont think modem uses hardware control or modem compression but i cant connect to verizon without it but when i look in my "WINXP" modem view log it says no hardware and no modem compression
anyhow ive tried with minicom and kppp and when i try to connect i see the ppc (wmodem) lights flashing on and off and i see in the log viewer a bunch of garbage
so i think this is a config prob
anyhelp will be greatly appreciated
m16-maniac said:
well linux detects the modem with np and it automatically loads the ipaq module. i can verify this by doing the 'lsmod' command and its in there
Click to expand...
Click to collapse
You can lookup vendor and product in
the 'lsusb' output, then do 'rmmod ipaq' and
modprobe ipaq vendor=XXXX product=YYYY
if this is the problem.
cr2 said:
m16-maniac said:
well linux detects the modem with np and it automatically loads the ipaq module. i can verify this by doing the 'lsmod' command and its in there
Click to expand...
Click to collapse
You can lookup vendor and product in
the 'lsusb' output, then do 'rmmod ipaq' and
modprobe ipaq vendor=XXXX product=YYYY
if this is the problem.
Click to expand...
Click to collapse
i dont think that is the problem cause as isay its picled up ok upon boot (as i edited the ipaq.c file then compiled the module to do that already
my problem i believe is in the pppsetup program that creats files called "options", "pppscript", "papscript" and "chapscript" i believe
in those files its doing commands wmodem does not understand i think
i am going to try as u suggested and see what happens but i dont think its gonna do anything different as i can already communicate with the modem just my isp sends garbage (i think) or the modem sends garbage im not sure which one it is
cr2 said:
m16-maniac said:
well linux detects the modem with np and it automatically loads the ipaq module. i can verify this by doing the 'lsmod' command and its in there
Click to expand...
Click to collapse
You can lookup vendor and product in
the 'lsusb' output, then do 'rmmod ipaq' and
modprobe ipaq vendor=XXXX product=YYYY
if this is the problem.
Click to expand...
Click to collapse
ok did what u suggested and it is the same info as it reports when i boot up with ppc plugged in (ur way and my way = same end result) so that not the problem
Okay so me and a few helpful people from the #htc-linux and #linwizard channels on freenode have built a version of linwizards kernel for the wizard to run on the excalibur! At first we had no framebuffer so no screen at all but thanks to tmzt, the framebuffer is fixed. Also myself, tmzt, and ali1234 made a (semi) working keymap. All the buttons work just a few tweaks need to be made.
**UPDATE**Thanks to some brilliant work by starwarrior we now can load android and play with it, the phone functions do not currently work, more details on post 76. Have fun.
I have included all you need to run the linwizard kernel on your excalibur. It is as safe as a soft reset; nothing permanant is done to your phone. You will have to remove the battery to reboot and get back to your normal windows enviroment.
Currently there is no support for the following and are a work in progress:
X
Sd card access
GSM usage (calls, sms, gprs)
Anyone with this type of experience or anyone who wants to help at all please email me at gentoon at gmail dot com.
Thanks to everyone working on the linwizard project and everyone who has helped do far.
http://www.sendspace.com/file/gfme0e
To run all you have to do is extract run the haret executible and click run. You will see some text letting yo know its loading the kernel, then some garbae as the framebuffer adjusts... then a login screen just enter root and nothing for the password field. Then there you go a working linux enviroment!
I know little to nothing about Linux but I am always up for something new. I have a "spare" Excalibur- and in another week I will have 2 spares so if you need anything specific tested let me know
Ummm so do you have the program available or no...
Know a place for him to host it?
gospeed.racer said:
Know a place for him to host it?
Click to expand...
Click to collapse
http://www.mediafire.com/
stylez said:
http://www.mediafire.com/
Click to expand...
Click to collapse
I was just going to say that... Mediafire is amazing...
sorry about that forgot how long uploading over a gprs can take, im visiting family and have no pc, so i have been doing everything over my dash. I added a link to the first post.
Okay operamini doesnt want to let me edit my post so sorry about the double post, here is the link:
http://www.sendspace.com/file/gfme0e
also you wil probably have to do a soft reset right before you run haret to avoid it freezing. Have fun.
Is it possible to see a screenshot of what it looks like on the excalibur.
Thanks
IDK for some reason I am hesitant to try this...
The main project is located at http://linwizard.sourceforge.net you can see some screenshots there of it working on the wizard. Of course our port doesnt support everything they have working on the wizard yet, but with some moe help we could.
BruteSource said:
The main project is located at http://linwizard.sourceforge.net you can see some screenshots there of it working on the wizard. Of course our port doesnt support everything they have working on the wizard yet, but with some moe help we could.
Click to expand...
Click to collapse
Much respect for putting this together.
let me know how i can help.
ptheabstract said:
let me know how i can help.
Click to expand...
Click to collapse
Well do you know anything about linux? I can get the kernel source uploaded. We need a better keymap first of all( we need \ mapped, also control and esc are mapped but dont seem to be working), then someone who knows something about X, it would be nice to have a GUI. Right now I would just like some people to test it on their excaliburs. Also all of the credit for the original kernel goes to the linwizard project. So thanks guys.
Hi,
Here are the patched files for Excalibur. Remove the .txt extension and copy them over the corresponding files in linwizard kernel tree.
Im going to be adding a detailed to do list on the main page. Once i get some real developers rounded up we are going to make this an official project on sf.
hey..i have seen this day in my dreams..lol
anyways i have ubuntu on of my pc's and love it...so i don't know too much about linux but i would definitely be willing to help!
dope biz...ill start looking into porting a keymap
....after midterms lol
ptheabstract said:
dope biz...ill start looking into porting a keymap
....after midterms lol
Click to expand...
Click to collapse
haha sucks I have midterms as well... sucks so bad...
Has anyone installed and gotten a to a terminal succesfully besides me yet? I could use some more testers.
Glad to see someone wants to help on he keymap.. right now everything works just need to map things like \ ctrl esc... etc
Hello everyone,
I am happy to inform all those folks who waited for a full Linux operating system on the HTC Kaiser PDA that today a little bit modified version of Zubuntu finally booted successfully on my HTC Kaiser.
It is a really in the alpha-state but it works, let's see what do we have:
TouchScreen: works perfectly with calibration
Screen: looks good. but programs doesn't support Kaiser's small resolution. You can fix this by using INTERPOLATION, but this makes picture quality a lot worse. To try it out edit the default.txt, find the mddi.height and mddi.weight values and change to for example 480 and 360.
Keyboard: works exactly as expected, some fine tuning with the keyboard map will be done
ScreenRotating
Not tested/not working:
WiFi (MY #1 ISSUE, TRYING TO FIX): I extracted the wifi section from an Android rootfs, successfully modprobed wlan.ko, dmesg shows that the driver is loaded OK, but:
Code:
iwconfig
tiwlan0: no wireless extensions
ifconfig tiwlan0 up
ifconfig: ioctl (SIOCDIFADDR): Can't assign requested address
TI's WLAN_LOADER program didn't work either, so now I am moving to the other driver (wl12xx)
Bluetooth (Driver is there, no BT programs installed, couldn't test it)
3G
Cameras
Phone (NO dialer, nothing, however perhaps gsm-libs can help us firing it up)
Many many thanks goes for all these people form the #htc-linux IRC channel: randomblame, Markinus, ali1234, Ondalf and bombenbach for their outstanding support and help!
I would like to say thank you especially to phh for his awesome help and patience with me!
Click on the image for more pictures!
Stay tuned for: hopefully working WiFi, screnshots, YouTube-demonstration.
If you are interested, want to say thank you, want to help, or have an idea how to fix some things please write into this topic.
Also if you like my work, please consider a little a donation for me
After all, I am pretty sure now you want to try this on your own too, so here is the link for the Google-Code website. Download all the files and read the Wiki to know what to do.
http://code.google.com/p/kaiser-buntu/http://code.google.com/p/kaiser-buntu/
Link to RootFS:
http://www.megaupload.com/?d=ANRC94ON
Mirror (thanks to jmcbrother):
http://rapidshare.com/files/371631240/ubuntu_rootfs.rar
Mirror #2 (thanks to hoatienii):
http://www.mediafire.com/?uvytadoykyy
Want to be updated? Check out the project's Twitter page!
http://www.twitter.com/kaiserbuntu
Thank you,
DOMy
RESERVED for future additions
Reserved for future additions
great work . please upload link to test
Sweet that is great news! I can't wait to try it out!
Me wants the alfa link
This is so great!
Maybe we can find a solution at my hackersspace.
Great Work... Luv Linux.... Ported Android on Windows Mobile Device... Eager to port Ubuntu on Kaiser...
can you please share the Linux Beta link ...
Makes me almost wanna install Ubuntu on my PC.
Dukenukemx said:
Makes me almost wanna install Ubuntu on my PC.
Click to expand...
Click to collapse
me 2
oh……thank you……I wanna test……wait ……
Would it be destructive to flash it to nand at this point in the project?
I wouldn't count on it even remotely fitting on the NAND. Besides, you'll probably have to have half the apps on an SD card anyway...
how about a build that will work with the new nomorootfs installers so hose of us like me can try it without going back to winmo, Would like to install to ext2.img files or sdcard partitions.
And is this remotely close to the MER Linux build that progress seems to be halted on I was able to get it to work
working XWindow?
I wouldn't recommend installing it to NAND (it doesn't work so good, for example I didn't have time to test power/suspend).
Xwindows works perfectly (except the resolution issue).
DOMy
resolution issue. can you please explain that? (whats wrong in resolution) before i flash back to ,WM and try it
Michga said:
resolution issue. can you please explain that? (whats wrong in resolution) before i flash back to ,WM and try it
Click to expand...
Click to collapse
The problem is that the rootfs is not for the Kaiser and the Xserver uses some hardcoded resolution (that means there is no config file for it) so circa half of the windows are too big for the screen. It is not a big problem because you can move them around so it is easy to get to one of the corners and resize them so you can see the whole thing at once.
Fixing this is on my To-Do list
sorry for the dumb question but...
will this support phone???
no mention in the OP.
I assume no. (for now anyway)
mnjm9b said:
sorry for the dumb question but...
will this support phone???
no mention in the OP.
I assume no. (for now anyway)
Click to expand...
Click to collapse
Driver support is there, but AFAIK no frontend yet.
DOMy
mod: I found this: http://packages.ubuntu.com/hardy/comm/gsm-utils and this http://manpages.ubuntu.com/manpages/gutsy/man8/gsmctl.8.html perhaps I will try it later (or someone can try it and report)
This intrigues me muchly, but why ubuntu/debian based?
There are other, (and imho simpler), bases to build on, slackware springs to mind, and there are some tiny distros out there, which may also be worth looking at, to say nothing of the likes of Familiar, an interesting candidate for a non-android htc-linux.
I'm attempting to access the camera and its features through Microsoft.Phone.Media.Extended.dll, but am only coming up with COM errors. Any help would be appreciated.
PhotoCamera cam = new PhotoCamera(CameraSource.PrimaryCamera);
when I go into debugging mode and inspect "cam" it shows that most of its properties are null. Does anybody know how to initialize this?
You can only access the camera from the photo Chooser task:
http://msdn.microsoft.com/en-us/library/microsoft.phone.tasks.cameracapturetask(v=VS.92).aspx
Sent from my HTC HD2 using XDA App
Blade0rz said:
You can only access the camera from the photo Chooser task
Click to expand...
Click to collapse
Using the official SDK APIs yes, but I'm talking about Microsoft.Phone.Media.Extended.dll
There are more methods inside of this dll than is normally available to developers.
Have you granted yourself the necessary permissions? You need to add ID_CAP_CAMERA to gain access to the PhotoCamera class.
kaspur said:
I'm attempting to access the camera and its features through Microsoft.Phone.Media.Extended.dll, but am only coming up with COM errors. Any help would be appreciated.
Click to expand...
Click to collapse
How did you get it load to solution? I getting error:
Code:
Could not load file or assembly 'Microsoft.Phone.Media.Extended, Version=7.0.0.0, Culture=neutral, PublicKeyToken=24eec0d8c86cda1e' or one of its dependencies. Strong name validation failed. (Exception from HRESULT: 0x8013141A)
You need to get the file from phone image. When you got it, you can reference it in your windows phone project.
To allows compiling your project you will need to remove the strong name check on media.extended.dll (in a visual studio command prompt sn -Vr pathto.dll)
Restart Visual Studio and you should be able to build your project
sn -Vr just whitelists the assembly on your PC. Why are your assemblies broken to begin with? All my GAC libraries work without errors (when referencing them). I suggest you work on extracting clean files.
With regards to COM -- what are the exceptions?
titaye said:
You need to get the file from phone image. When you got it, you can reference it in your windows phone project.
To allows compiling your project you will need to remove the strong name check on media.extended.dll (in a visual studio command prompt sn -Vr pathto.dll)
Restart Visual Studio and you should be able to build your project
Click to expand...
Click to collapse
Thanks! I remove a strong name (now I have no exceptions at design time) but the app doesn't load. Stack trace:
Code:
'taskhost.exe' (Managed): Loaded 'mscorlib.dll'
'taskhost.exe' (Managed): Loaded 'System.Windows.RuntimeHost.dll'
'taskhost.exe' (Managed): Loaded 'System.dll'
'taskhost.exe' (Managed): Loaded 'System.Windows.dll'
'taskhost.exe' (Managed): Loaded 'System.Core.dll'
'taskhost.exe' (Managed): Loaded 'System.Xml.dll'
'taskhost.exe' (Managed): Loaded '\Applications\Install\AF30D7A5-21BB-480F-94CC-658A0DCE171A\Install\CameraTest.dll', Symbols loaded.
'taskhost.exe' (Managed): Loaded 'Microsoft.Phone.Media.Extended.dll'
'taskhost.exe' (Managed): Loaded 'System.SR.dll'
The thread '<No Name>' (0x166f19be) has exited with code 0 (0x0).
The thread '<No Name>' (0xecf0ade) has exited with code 0 (0x0).
The program '[371728818] taskhost.exe: Managed' has exited with code 0 (0x0).
I've already added <Capability Name="ID_CAP_CAMERA"/>, no luck. Seems like I'm missing some important part. Unfortunately guys who already had camera working doesn't like to share experience with community...
WithinRafael said:
sn -Vr just whitelists the assembly on your PC. Why are your assemblies broken to begin with? All my GAC libraries work without errors (when referencing them). I suggest you work on extracting clean files.
Click to expand...
Click to collapse
Could you please post your dll's, test project and some instructions? I'm really curious to get video camera working.
I'm able to get this working. The trick seems to be to create a CameraVisualizer, and then SetSource to your camera object. The camera is then shown in the CameraVisualizer. (This is as far as I've got)
Here's a sample: dl.dropbox.com/u/12359/Flashlight.zip
You'll need to follow the instructions here to add Microsoft.Phone.Media.Extended: thounsell.co.uk/2010/11/avoiding-reflection-adding-the-interopservices-library-to-the-wp7-sdk/
Thank you, perfect, get it works!
sensboston said:
Thank you, perfect, get it works!
Click to expand...
Click to collapse
microsoft boots it after a zune sync...tells you to uninstall it... any ideas?
kawgirlval69 said:
microsoft boots it after a zune sync...tells you to uninstall it... any ideas?
Click to expand...
Click to collapse
run the unlocker again and it will work again .
ya.. thanks... found that out... was looking for a more "permanent" fix... still the nicest and most functional flashlight... better than that white screen crap hands down.. even having to reunlock every time..
kawgirlval69 said:
still the nicest and most functional flashlight...
Click to expand...
Click to collapse
Thanks, but it's not so functional and a little buggy (was written in couple of hours, just for academic purposes). HTC have a real nice flashlight, cool looking and functional but runs on HTC handsets only... They directly (probably using native code calls) manipulating LED.
Will see, may be in future we'll have more expanded API (I hope so!)...
Just a note, in case you didn't notice, or other people don't know, that we're leaking videos when using this as a flashlight. I'm pretty sure that videos and thumbnails are created in IsolatedStorage for the application. Uninstall/reinstall should clear them, but we should really delete them as soon as they're created. It'd be nice if there was an API that didn't actually record, but we'd need to see whats in PhotoLibEnt.dll to figure that out.
sensboston said:
Thanks, but it's not so functional and a little buggy (was written in couple of hours, just for academic purposes). HTC have a real nice flashlight, cool looking and functional but runs on HTC handsets only... They directly (probably using native code calls) manipulating LED.
Will see, may be in future we'll have more expanded API (I hope so!)...
Click to expand...
Click to collapse
lol i guess im not picky... it would be fine for me ... again only dislike is the having to reunlock after a zune sync... the other glitches are minor to me.. i mean it does whats needed... you tap..it opens..tap on it comes... tap off it goes.. that = functional to me... the cert thing is all i see bad..
You don't need to reinstall the applications each sync. Just re-unlock it before opening the apps, and if you've already opened the dialog to be greeted with the prompt to uninstall - just hit the Windows Key to cancel and go back to the Start screen.
sensboston said:
Thank you, perfect, get it works!
Click to expand...
Click to collapse
excellent! Thanks so much!
titaye said:
You need to get the file from phone image. When you got it, you can reference it in your windows phone project.
To allows compiling your project you will need to remove the strong name check on media.extended.dll (in a visual studio command prompt sn -Vr pathto.dll)
Restart Visual Studio and you should be able to build your project
Click to expand...
Click to collapse
Where I can find a visual studio command prompt in Express 2010 for Windows Phone?
I hope all this is just a little bit interesting for some developers because I can't do any further investigations on ANDROID because my device is broken. At the moment it's on the way and I hope I will get a change of my broken HD2. I have summarized some more or less insteresting facts and I hope someone of you will think about all this informations and can work with them.
First at all: I'm not a developer but I have done some search on ANDROID and logged and logged and logged it again. All this is just a collection of facts and results of compare log files.
When you disassemble some Files within Android, you will see that something will not work because it can't. Let me start with the init.rc what is missing or might be wrong:
1. MISSING FOR BLUETOOTH:
Code:
# create mountpoints
mkdir /mnt_data/download 0777 root system
2. WRONG?
Code:
chmod 0666 /dev/ttyHS1
chown root radio /proc/cmdline
chmod 666 /proc/cmdline
There is no ttyHS1 device within the ANDROID. Can't find it.
3. NOT SURE
rfkill is seems not working. Recommendation on the website:
http://www.gitorious.org/openaos-te...5f360fec4c21c1e06d145a5fe?diffmode=sidebyside
Code:
chmod 0666 /sys/devices/platform/wifi_bt/bt_enable 0
write /sys/devices/platform/wifi_bt/bt_enable 1
but based on ANDROID Developer FAQ it should be:
Code:
chmod 0666 /sys/modules/board_[platform]/parameters/bluetooth_power_on
write /sys/modules/board_[platform]/parameters/bluetooth_power_on 1
and this MUST also done after the change:
Code:
on property:persist.service.bluetoothd.enable=0
write /sys/modules/board_htcleo/parameters/bluetooth_power_on 0
...depend of the settings it also could be:
Code:
write /sys/devices/platform/wifi_bt/bt_enable 0
on property:persist.service.bluetoothd.enable=1
write /sys/modules/board_htcleo/parameters/bluetooth_power_on 1
...depend of the settings it also could be:
Code:
write /sys/devices/platform/wifi_bt/bt_enable 1
Also interesting is the command write which is used a lot of times by the init.rc. I don't know where this command is located. But if you open a SHELL on Android and try to "write" something you will see that this command is not available. Can someone explain how this will work when this command is not available????
4. WRONG!
Code:
chmod 666 /dev/uinput
chmod 666 /etc/bluez/audio.conf
chmod 666 /etc/bluez/hcid.conf
chmod 666 /etc/bluez/input.conf
On the one hand this path is not existing and on the second hand the hcid.conf is also not existing. If you take a look only within /etc you will see that /etc/bluetooth is only existing.
5. BT INITIALIZATION
Code:
service hciattach /system/bin/brcm_patchram_plus --enable_hci –enable_lpm --baudrate 3000000 --patchram /etc/firmware/bcm4329.hcd /dev/ttyHS0
user bluetooth
group bluetooth net_bt_admin
disabled
This is also not working at all. If you try to run this command via the shell nothing will happens. And if you need to load the firmware into the device you need also setup this device by nvram.txt which you can see within the bluetooth.c file. The nvram.txt is important to get a valid MAC Address from the hardware. If you take a look into the source code of the BCM4329 you will see additional informations about this.
The second problem is the value of 3000000. This is to small. If you want to work with head set than a value of 4000000 is minimum required. From my point of view this initialization should be:
Code:
service hciattach /system/bin/logwrapper /system/bin/hciattach -s 57600 /dev/ttyHS0 any 4000000 flow
The value 57600 is the minimum value for initialize bluetooth and it could be that this value will downsize the standby drain. Not sure, just an idea.
Why? Lets take a look into the main.conf or hcid.conf:
Code:
# What value should be assumed for the adapter Powered property when
# SetProperty(Powered, ...) hasn't been called yet. Defaults to true
InitiallyPowered = true
# Remember the previously stored Powered state when initializing adapters
RememberPowered = true
This was the reason why I think the initialization speed "...-s 57000 /dev..." should be lowerd down which might also solve the battery drain issue. Again, not sure - just an idea.
Btw, if you open the ADB shell and try hciconfig, hciattach or hcitools non of this are working because there is no Bluetooth Adapter even if you have BT enabled. This tools are requried for BT analyse, test and command options. You can call each of the files within the Shell but whatever you try to do you it will fail because BT is not enabled. Try to run rfkill which could initiate a reset on the BT device. If you try to rfkill ttyHS0 which is the BT adapter on HD2 device this will not work. From my point of view all this problems are not only related to the kernel.
6. TAKE A LOOK INTO THE init.htcleo.rc
Code:
# Make sure we startup btld before hcid
# Set target address to emulator host loopback IF
# Limit baudrate to 460800 to ensure reliable uart operation
service btld /system/bin/logwrapper /system/bin/btld -hwtun 10.0.2.2 -hb 460800 3000000 -lpm 1
# service btld /system/bin/logwrapper /system/bin/btld -lpm 1 -hb 3000000
Not sure what will run first. The init.rc or the init.htcleo.rc. If the init.rc will run first this could be also a problem. Btw, the first value is 460800. From my point of view this is too much for initialization.
7. WiFi INITIALIZATION
Just an example but if you do a Google search for "BCM4329 nvram.txt" you will get more than 3 links. Let me show here an example:
http://android-wifi-tether.googlecode.com/svn-history/r465/trunk/res/raw/tether_edify
Search for Value 4329 and you will see SAMSUNG Device bla, bla – don't know from top of my head:
Code:
!file_exists("/sdcard/android.tether/bcm4329.bin") && ( module_loaded("dhd") || log(insmod("/lib/modules/dhd.ko", "firmware_path=/system/etc/wifi/bcm4329_mfg.bin nvram_path=/system/etc/wifi/nvram_mfg.txt"), "Loading dhd.ko module<br>(bcm4329_mfg.bin from /system/etc/wifi/)"); );
Hmm... That was the reason why I grabbed the SAMSUNG i9000 from a friend of me and double checked the device. If you take a look into the /etc/wifi you will see the nvram.txt. I'm not sure how this will be loaded into the device but the i9000 has the same BCM4329 Chipset.
Now let us take a look into the wifi.c for QUALCOMM devices which is also valid for the HD2:
http://gitorious.org/linux-on-qualc...f8dffe668c0448/libhardware_legacy/wifi/wifi.c
Code:
Line 61: Take a look. Nothing of this is existing on the HD2
Line 71: there is no wlan.ko on the HD2
But now, where the hell the MAC Address is stored for the WIFI Adapter? Just take a look at this location:
Code:
sys/module/board_htcleo/parameters/bdaddress
or use this:
Code:
/system/sysroot/module/board_htcleo/parameters/bdaddress
If you check this file you will see there your current MAC Address of the WIFI device. The stupid MAC for the BT device seems not stored on the device. And this is the point where I guess to need the fu***ing nvram.txt.
I have done a search for some BCM files and found something interesting which is attached as "Broadcom Files.7z." There are some (!) more or less firmware files within (?) and also the famous nvram.txt
For additional informations about WiFi and more just open the wifi_suplicant within the /system/bin directory with a simple text editor and browse down to the end. There you will see also a lot of interesting settings for the WiFi environment.
8. W.t.F. IS THIS STUPID NVRAM.TXT?
Download attached Broadcom.7z and check the txt file within. If you open the file you will see the nearly same content as within the
Code:
/etc/calibration
file.
But this file you can ignore on your HD2 device. Delete it, rename it – do whatever you want and you will see no changes on the device. The more interesting thing is within
Code:
/sys/calibration
Huhhhh...! The same file but less values. But where does it come from??? Currently I don't know.
It seems that the nvram.txt is as same as like a BIOS for the device. The BCM4329 is a BT AND WiFi Chip within one CHIP. It might be wrong but I guess this is the easy explanation for this.
And because of this the thing can't work. First it needs to load the BIOS (nvram.txt) and then it needs the OS for it (BCM4329 Firmware).
Okay here is something by reading and following your examples:
sys/module/board_htcleo/parameters/bdaddress
That lines exists in the current NAND Android roms, but this line:
/system/sysroot/module/board_htcleo/parameters/bdaddress
doesnt exist, not in Rafs rom or in imilka's 0.1 GB sense rom.
But here is my key interest: "bdaddress" is where the BLUETOOTH MAC Address is!
Another interesting thing, in imilka's 0.1 GB Sense rom, I can change this to whatever I want and it Sticks till I reboot, but in Rafs it does not.
Key Question, How do we make this file KEEP the changes we make to it??? I know its a dirty fix but none the less its a FIX!! So anyone got a clue?!??
First of all there is a WRONG that is big like an house
the init.rc "syntax" is not shell syntax.
so as example, command "write" will not work in shell... but only in init.rc
I see also a lot of confusions about stock froyo/gingerbread stuff and sense stuff.
Example: Sense uses it's own stuff for BT so a clean init.rc for sense is really different from the one for a non-sense build.
Also some stuff/commands needs to be changed/replaced to have them working with our hd2 kernel.
I have no time now but really this posts is about 30 different things that cannot be explained all togheter.
My suggestion is to google for "android init syntax" and start from there understanding all the android boot process and the android init syntax.
About the wifi mac address.
The wifi mac address is "read" from the libhardware_legacy.so
normmaly this lib reads the mac address in /etc/calibration
this file is a "kernel" file so you are not able to change it.
You can use a modified libhardware_legacy.so that reads the mac address in /system/etc/calibration so you can change it, as I did on my builds where you can change the wifi mac address.
Unfortunately the modified libs causes other issues like gps not working.
---------------------------------------------------------------------------
About the problem that you cannot write permanetly the changes on stuff inside the folders
/
/etc
/proc
/sys
and so on.. this is because those folders are the "kernel" that normally is read-only
you can make it write enabled but it's not a safe way to proceed...
so is there a chance to fix the issue with the bluetooth mac?!
would be important for me
(can't use my parrot car kit with my wifes hd2 because of the same mac adress )
rafpigna said:
First of all there is a WRONG that is big like an house
the init.rc "syntax" is not shell syntax.
so as example, command "write" will not work in shell... but only in init.rc
I see also a lot of confusions about stock froyo/gingerbread stuff and sense stuff.
Example: Sense uses it's own stuff for BT so a clean init.rc for sense is really different from the one for a non-sense build.
Also some stuff/commands needs to be changed/replaced to have them working with our hd2 kernel.
I have no time now but really this posts is about 30 different things that cannot be explained all togheter.
My suggestion is to google for "android init syntax" and start from there understanding all the android boot process and the android init syntax.
Click to expand...
Click to collapse
Thanks mate! The problem is that I have no clue how to do it but I hope someone outside who has more experience with all this knows what to do.
Btw, I have had also tried to do a simple test with this "write" command by fill a line within /etc/test and also /proc/test which fails. Usually I added some lines into the init.rc to see if this will work.
But I'm not sure if I have had something wrong.
rafpigna said:
About the wifi mac address.
The wifi mac address is "read" from the libhardware_legacy.so
normmaly this lib reads the mac address in /etc/calibration
Click to expand...
Click to collapse
Sorry, but this is not true. Check this:
http://developer.android.com/reference/android/net/wifi/WifiInfo.html
If you check this from Android SDK you will see that the wpa_supplicant is responsible for this. And this would be O.K. if you also check the BCM4329 source code.
First the CHIPSET will be initiated and prepared by the firmware and the nvram.txt and after that the wpa_supplicant do the rest.
How this exactly will work - sorry, too less coding knowledge
A good thing is it to compare the MAC & BT issue with the i9000 devices and Sony X10. The Sony X10 has the same QUALCOM chipset as the HD2.
Here are the files from the original Desire HD 1.8x.
The initrd.zip contains is the original file and the initrd~.7z is the extracted who is interested to read and compare.
see this http://gitorious.org/linux-on-wince...mmit/ce69804632e64b397758c1c9835f06efd0c8cb54
in file arch/arm/mach-msm/board-htcleo.c i see from markinus some changes to file bdaddress.but it is not in the main git tree we use for hd2 kernel
maybe someone kernel developer can see it and make a kernel for testing;
clio94 said:
see this http://gitorious.org/linux-on-wince...mmit/ce69804632e64b397758c1c9835f06efd0c8cb54
in file arch/arm/mach-msm/board-htcleo.c i see from markinus some changes to file bdaddress.but it is not in the main git tree we use for hd2 kernel
maybe someone kernel developer can see it and make a kernel for testing;
Click to expand...
Click to collapse
Looks good, but this is not valid for HD2. This is HTC Tattoo. If you check this code you will see "akm8973".
Damn!
http://nagaraj-embedded.blogspot.com/2011/02/guide-on-adding-wifi-drivers-on-android.html
and
http://www.jetdroid.org/forum/viewtopic.php?f=78&t=456&start=20#p4502
Can someone double check this???? From my point of view this looks quite good...
MrT69 said:
Sorry, but this is not true. Check this:
http://developer.android.com/reference/android/net/wifi/WifiInfo.html
If you check this from Android SDK you will see that the wpa_supplicant is responsible for this. And this would be O.K. if you also check the BCM4329 source code.
First the CHIPSET will be initiated and prepared by the firmware and the nvram.txt and after that the wpa_supplicant do the rest.
How this exactly will work - sorry, too less coding knowledge
A good thing is it to compare the MAC & BT issue with the i9000 devices and Sony X10. The Sony X10 has the same QUALCOM chipset as the HD2.
Click to expand...
Click to collapse
I'm wrong?
Sorry to say, I'm not an andorid guru but maybe I have a little bit more knoledge and trust me in our case, at least with our builds, the mac address is read from the calibration file.
So how is possible that in my builds you can change the wifi mac?
I just take yor desired wifi mac, write it in the /system/etc/calibration file and replace the libhardware_legacy.so that will read from that instead of /etc/calibration
MrT69 said:
Thanks mate! The problem is that I have no clue how to do it but I hope someone outside who has more experience with all this knows what to do.
Btw, I have had also tried to do a simple test with this "write" command by fill a line within /etc/test and also /proc/test which fails. Usually I added some lines into the init.rc to see if this will work.
But I'm not sure if I have had something wrong.
Click to expand...
Click to collapse
Already said... init.rc syntax is not shell syntax!!!
Anyway.. I still miss the purpose of this thread....
Just the first two google search links appearing with a simple "android init.rc" search
http://www.androidenea.com/2009/08/init-process-and-initrc.html
http://elinux.org/Android_Booting
Read that and something will be clearer.
rafpigna said:
I'm wrong?
Sorry to say, I'm not an andorid guru but maybe I have a little bit more knoledge and trust me in our case, at least with our builds, the mac address is read from the calibration file.
Click to expand...
Click to collapse
It might be that I have had something missunterstand but I only want to say by default (!) the MAC Address is not read from the calibration file by ANDROID. Of cause I have had tested your solution and it's working.
So how is possible that in my builds you can change the wifi mac?
I just take yor desired wifi mac, write it in the /system/etc/calibration file and replace the libhardware_legacy.so that will read from that instead of /etc/calibration
Click to expand...
Click to collapse
I know that the MAC Address could be changed within this way. I hope you don't understand it within the wrong way but from my point of view this is only a patch solution. Because of the fact that the BT & WiFi Chip is ONE (!) chip within HD2 there must be an other solution to read and get the MAC address for WiFi AND BT.
Already said... init.rc syntax is not shell syntax!!!
Click to expand...
Click to collapse
I only was asking why and where it was not working. In the mentime I know it because of your feedback. But I have add some lines into the init.rc and the files and changes was not written to the device. That was the reason why I was asking.
Anyway.. I still miss the purpose of this thread....
Click to expand...
Click to collapse
Hope to find some people who will help to fix all this issues and on the second hand I hope to find a solution for the BT and WiFi MAC to get it from the device.
It is proper time to make an open discussion about the 2.2.1 rom and more tweaks closer to HD2 reveal in the community.
MrT69 said:
I know that the MAC Address could be changed within this way. I hope you don't understand it within the wrong way but from my point of view this is only a patch solution. Because of the fact that the BT & WiFi Chip is ONE (!) chip within HD2 there must be an other solution to read and get the MAC address for WiFi AND BT.
Click to expand...
Click to collapse
maybe this is wrong but it is extremely good that in sense roms,you can change at any time wifi mac address.in my city municipality have free wifi hotspot but after sometime reads your wifi mac address and cut the connection for a time.so with dynamic wifi address i can be almost all time everywhere online and this saves me from a slow (because i live in a small greek city and 3g connection is not everywhere) data edge connection.