Working Tomtom Nav On Angstrom

[speculatrix]

This really is exciting!  I agree with Tovarish, a good GPS application would be a killer app (for me anyway     )

Have you thought about adding a link to your page for sponcership?  I think that this is a really worth while project and would support it.  Its probably fairly unlikely but you could even point out the size of the Japanese Z market to TomTom and see if they'd also sponcer you.... 
[div align=\"right\"][a href=\"index.php?act=findpost&pid=117791\"][{POST_SNAPBACK}][/a][/div]

The snag here is that I am treading a thin line with TT's s/w rights... because I own a GO 300 I figure that I am simply seeking to allow me to run the software on an alternative device.

As I understand it, the maps and stuff are all locked to the device hardware (it tries to read a specific section of flash to get the device ID), possibly even encrypted using a key unique to the device. Now, I am created "look-alike" devices/files which will make the ttn think it's running on a GO 300, *but* it will be running on a copy of *my* device. If I then released a file system dump, it'd mean everyone would be using my license key which would be a Bad Thing.

If I get it running, I can tell people how to do it, and if they have a GO, they can use their own licenses! And if they want to donate a beer, that's fine.

If I can turn round to TT and say "look, it's really easy, why not release it for the Zaurus, it will run with very little modification", then maybe they will - if enough Z users ask! Maybe we should ask them anyway - or maybe there's no way they could lock it down? Does each Z have a unique ID somehow? From vague memory, the Picsel browser is somehow locked to the memory card they supply it on?

[speculatrix]

p.s. there's also the matter of time, which I am short of, and getting help from the OpenTom people as they can tell me more about the "missing" devices and /proc entries which ttn is looking for. I'm currently waiting for an answer on this.

if there's anyone out there with a TomTom GO, who is interested, join in the discussion.

[ltrm]

The snag here is that I am treading a thin line with TT's s/w rights... because I own a GO 300 I figure that I am simply seeking to allow me to run the software on an alternative device.

As I understand it, the maps and stuff are all locked to the device hardware (it tries to read a specific section of flash to get the device ID), possibly even encrypted using a key unique to the device. Now, I am created "look-alike" devices/files which will make the ttn think it's running on a GO 300, *but* it will be running on a copy of *my* device. If I then released a file system dump, it'd mean everyone would be using my license key which would be a Bad Thing.

I thought that something like this might be the case.  The best scenario for me would be if the Z did have something (processor id?)  that could be used as a key so that I could just buy the closed source map data from TT and use it with the open source software.  I fully understand TT wanting to keep control of that.

Since they do sell software only packages for other platforms I guess there is an outside chance they might go for just selling the maps, unsupported, on SD if they could still DRM it somehow.

As for time scales, I'm not too fussy about that.  Just knowing it was possible would be enough for me to start pestering TT  

Failing getting a legal working copy on my 860, I think that I'm going to have to get a symbian phone and run GPS software on there.  Seems like a shame to miss out on the Zaurus screen though....

[speculatrix]

p.s. there's also the matter of time, which I am short of, and getting help from the OpenTom people as they can tell me more about the "missing" devices and /proc entries which ttn is looking for. I'm currently waiting for an answer on this.

if there's anyone out there with a TomTom GO, who is interested, join in the discussion.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=117796\"][{POST_SNAPBACK}][/a][/div]

well, only one person came back to me from the opentom lists and it was a bit vague.. perhaps  noone's tried exploring the default TT linux environment.

so, after a lot of head scratching, I've been having an interesting* time with bluetooth. I now have a quick and dirty hack script which listens for an rfcomm serial connection and provides an interactive terminal on it. This will allow me to connect to the TTG and run a few commands and hopefully determine what the special devices and proc files are for. "All" I've got to do is roll my own TTG 300 boot image!

Hopefully, these special files are going to have obvious meanings (hmm, I guess I should put "od" or something into my boot image so I can make a hex dump if necessary!).

*bluetooth commands on linux have a particular nasty failure mode - they usually just ignore what you entered rather than reporting an error... try "hcitool scan" then "sdptool scan" instead of "sdptool browse" by mistake for example.

[Tarquin]

Before I start getting too excited about this, is there any chance of it running on an SL5500?  I have been looking for a mapping system for the UK, and this seems like a great idea...

[speculatrix]

Before I start getting too excited about this, is there any chance of it running on an SL5500?  I have been looking for a mapping system for the UK, and this seems like a great idea...
[div align=\"right\"][a href=\"index.php?act=findpost&pid=120095\"][{POST_SNAPBACK}][/a][/div]

I would say there's a reasonable chance of running on an SL5500 - the TTG is only QVGA display, and has (IIRC) the same memory size or even less than the 5500. So long as you have a serial GPS (which appears as a /dev/ serial device) and a 128MB memory card for storing maps etc, I can't see why not.

[speculatrix]

BTW, there was recently a positive step on Opentom, someone started documenting the special devices in OT's /proc and /dev filesystems... I expect to have the next experimentation session soon!

[Da_Blitz]

At the moment i have a copy of the SD_Pysical_Spec (siplified version) and the SD_SDIO spec (siplified version) if any one is intrested, they were free to download off the sandisk website but have scince been removed

contact me if you want a copy

[speculatrix]

this update has been a long time coming... having spent a lot of time flying or in airports recently, I was able to apply a succession of nasty hacks to my chrooted environment, and am sad to say it's mixed news.  

Luckily, TomTom's binary spits out useful info as it starts up, so I was able to incrementally fix or kludge things to get ever further: I was able to fool ttn into finding the remote control, think the right things were mounted, finding the right devices, think it was running on real tomtom hardware.

I am now officially stuck... and would welcome any guesses...

malloc_init: Never using mmap for malloc
Checking whether blueaclserver.a is running (bluetooth)
Executing [pidof 'blueaclserver.a' > /var/run/temp.txt]
#### RECEIVED SIGNAL 17 [errno 0, code 1]
#### Sent by process: pid 11193, uid 500
#### Exit status: 1, user time: 0, system time: 2
#### RESUMING
Error returned: 256
succeeded in opening /var/run/temp.txt
bluetooth_available = 0
Opening in MASTER mode
The framebuffer device was opened successfully.
InitScreen: 480x640, 16bpp - portrait - fake: 640x480
The framebuffer device was mapped to 0x40255000 successfully.
Initialized touchscreen
Initialized remote control
Error -9000
Unmounting SD Card...FAILED: Invalid argument
Unmounting flash...FAILED: Invalid argument
Powering off because the game frame couldn't be initialized
Forcefully unmounting /mnt/flash...FAILED: Invalid argument
Forcefully unmounting /mnt/sdcard...FAILED: Invalid argument
wd_set: WARNING: Attempt to set watchdog, but not opened yet!

googling for "game frame" hasn't helped at all  :-(
next step is to play with strace and/or gdb to see what it was trying to do before it barfed.

[Da_Blitz]

I would say that perhaps it has some stuff to do with ioctl on the Frame buffer device that isnt working, game frame could only be the frame buffer in my opinion, peerhaps check the source for modification?

[speculatrix]

strace has worked its magic and I found that the errors I was getting were from the system shutting down... so, armed with trace I was able to fix more and more things.

it now gets a long way through - opening the maps, opening the gps devices, but then segfaults with address 0xfffffffff (dunno how many 'f's)... spitting out a simple backtrace.

Now, I created fake /dev/ and /dev/gps devices by linking to ttyS1, as was the /dev/watchdog device, and the seggy triggers just after writing null bytes to /dev/watchdog (but there is clearly a lot going on here in the ttn binary, so I am guessing that writing to devices isn't the answer).

so, I am again stalled. My next step will have to be to use a debugging to run the code and see what happened. I might post here the strace output and the normal stdout/stderr output for people to see just how far its getting.

hmm. I just wondered if perhaps the fact I am using junk values for things like device ID might cause the map data decryption routines to barf.

I really need now to create my own ttsystem file with a bluetooth login hack so I can be sure my dummy dev & proc files are "real" rather than junk.


tuned for another thrilling episode another time!

Paul

[Mickeyl]

Exciting stuff, thanks for keeping us posted!

[scoutme]

what's the actual status at this time?

[speculatrix]

I had to sell my 860, so I had to start again with my 6000 but it got broken during an abortive upgrade attempt. Once I've got my 6000 working again I am hoping to restart "zomtom" (the secret name I have just invented for zaurus-tomtom project). If you're wondering what I did to my 6000 just follow link in my .sig.

I guess although many people were interested, noone has a tomtom themselves and thus they don't have any "moral" right to extract data from their tomtoms and try and run it on the Z otherwise they'd have been trying to help, right?

Of course, there's nothing to stop people emailing and asking TomTom whether they would consider a native Zaurus port of their software

[koen]

I am hoping to restart "zomtom" (the secret name I have just invented for zaurus-tomtom project).
[div align=\"right\"][a href=\"index.php?act=findpost&pid=134015\"][{POST_SNAPBACK}][/a][/div]

Oh goody, another 'only for zaurus and gcc 2.95.3 and qt/e from 5 years ago' project that would need a complete rewrite if you want to use it on another linux machine....