March 2019 - OvmsDev - Open Vehicles

Locations and scripts
by ovms 03 Dec '19

03 Dec '19

Hi I would like to run scripts based on location. The scripts work fine, but the location service stops working. I guess the location service requires CAN1 to be working? Since CAN1 stops receiving after a short period of time, the location service also stops. Is there a way to bypass this so the location service does not depend on CAN1 working? Kind regards, Stein Arne Sordal

7 61

Dexter'web API
by Tim Jackson 28 Sep '19

28 Sep '19

Hi there all, I have 2 of the latest units, and I am running a Nissan Leaf and e-NV200.. I have 2 questions 1. Is there an API for dexter-web.de <http://dexter-web.de/> site? 2. Is there any updated documentation on the current API on TMC? Thanks Tim

2 2

BUG? in v3.2.0-1 in vehicle_mitsubishi - flood of BMS-error msg
by Michael Eric Menk 11 Jul '19

11 Jul '19

There is a bug in MIEV code, that bombs the web GUI and phone APP with BMS errors. I suspect that this is because the Peugeot and Citroen branded MIEVs have a 80 cell battery pack vs MIEVs 88cells. Sample of a message: > Voltage: SD=1mV OK > Temperature: SD=1.4C !6:+3.1C !16:+3.1C !18:+3.1C The missing cells could also be used to set the voltage gauge. There are two LEV50-4 missing in the 80 cell pack, number 06 and 12. If a flagg is set while looping the CAN massages, then the flag can be used inside OvmsVehicleMitsubishi::GetDashboardConfig to set correct voltage. The float charge voltage of the 80 cell pack is 328.4 volt

3 15

3G network
by Stein Arne Sordal 02 Jul '19

02 Jul '19

Hi I guess this question is basically for Mark, In Norway the telecom providers have decided to turn off almost all of the 3G network soon and focus on 4G only. Any plan for 4G modem? Kind regards, Stein Arne Sordal

6 12

Can buses stop after some time
by Tom Parker 01 Jun '19

01 Jun '19

Hi all, I synced up with master about a week ago and since then I've seen both can busses stop working. I still see the 12v battery metric changing, but everything that comes from the car stops. Rebooting the module with "module reset" does not seem to fix it, while make app-flash monitor does fix it. I haven't tried make monitor on it's own. Is anyone else seeing behavior like this? Sorry for the vague bug report. I'll spend some time later this weekend to try to gather more information.

7 45

Garbage in /store
by Stephen Casner 27 May '19

27 May '19

When I do "vfs ls /store" I get 128 lines like the following: -2146205 01-Jan-1970 00:00 ????????.??? After that comes valid content: [DIR] 01-Jan-1980 00:00 ovms_config/ 83 01-Jan-1980 00:00 .htpasswd The config facility appears to be intact. Does the garbage indicate that the /store filesystem is scrogged and needs to be rebuilt? -- Steve

3 11

Android App v3.12.2 in Play Store
by Michael Balzer 07 Apr '19

07 Apr '19

Whoa. Our first public App update since …drumroll… June 2014 :) I didn't get any bug reports for my 3.12.2 beta release, so I just finished some last details and uploaded it. May take some hours to appear in the Play Store. As change history has grown a bit beyond the allowed 500 chars for the release info, I've added a history page to the Wiki: https://github.com/openvehicles/Open-Vehicle-Android/wiki/History I'm afraid it's just the condensed gitlog though. I hope users won't get confused, sure could need some refinement. Regards, Michael -- Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal Fon 02333 / 833 5735 * Handy 0176 / 206 989 26

10 47

Fasttech stock status
by Michael Balzer 01 Apr '19

01 Apr '19

Mark, any idea when the modules will be in stock again? Until a few days ago it was just the EU kit, so users could assemble the base module + modem, but now the base module is also sold out. Thanks, Michael -- Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal Fon 02333 / 833 5735 * Handy 0176 / 206 989 26

3 4

Hologram spacebridge (IP tunnel)
by Michael Balzer 31 Mar '19

31 Mar '19

FYI: after recognizing it's free I just tried the hologram spacebridge tunneling. Both http (+ws) and ssh work via the tunnel, they're just very slow. I think the bad speed is mostly due to the very slow PPP implementation. The web framework will run into timeouts on many page requests, but once loaded is in fact usable. Executing commands via http /api/execute works nicely. The ssh connection works even better, just feels like good'ol GPRS. Regards, Michael -- Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal Fon 02333 / 833 5735 * Handy 0176 / 206 989 26

1 0

Revised RegisterCommand arg defaults
by Stephen Casner 16 Mar '19

16 Mar '19

I have made a widespread commit to flip the sense of the default value for the 'bool secure' argument in RegisterCommand() to be true rather than false. Very few instances of RegisterCommand() took advantage of the argument defaults because 'secure' defaulted to false when most instances needed true. In addition, the default for the 'max' arg was INT_MAX whereas most instances should have max = 0, so that default is also changed. And lastly, a new default value NULL was added for the 'execute' function pointer. It was not necessary to change any of the calls to RegisterCommand() other than those that used the secure = false default. However, I decided to go ahead and review all the instances in the code and change them to make maximal use of the arg defaults so that the existing code can serve as a useful pattern for the addition of new commands. In particular, for secure, non-terminal commands, such as the top-level "framework" commands, the model should simply be: RegisterCommand("name", "Title"); For secure, terminal (sub)commands that don't require any additional parameters, the model should be: RegisterCommand("name", "Title", execute); This model also applies if the command has children but the command itself wants to execute a default operation if no subcommand is specified. It is incorrect to specify min = 0, max = 1 to indicate an optional subcommand; that is indicated by the presence of the execute function at the same time as a non-empty children array. The default usage string will list all the children. Any command with required or optional parameters should provide a usage string hinting about the parameters in addition to specifying the min and max number of parameters: RegisterCommand("name", "Title", execute, "usage", min, max); I found several instances of commands with parameters that left the usage string empty. If you recognize some of these as yours, please add hints for the parameters in the usage string: cmd_ecu->RegisterCommand("list","Show OBDII ECU pid list",obd2ecu_list, "", 0, 1); cmd_re->RegisterCommand("list","List RE records",re_list, "", 0, 1); cmd_dbc->RegisterCommand("list","List RE DBC records",re_dbc_list, "", 0, 1); cmd_discover_list->RegisterCommand("changed","List changed records",re_list_changed, "", 0, 1); cmd_discover_list->RegisterCommand("discovered","List discovered records",re_list_discovered, "", 0, 1); cmd_test->RegisterCommand("strverscmp", "Test strverscmp function", test_strverscmp, "", 2, 2); The following one includes a usage string, but since the parameter is optional (indicated by min = 0), the usage should be "[<url>]", so I made that change: cmd_otaflash->RegisterCommand("http","OTA flash http",ota_flash_http,"<url>",0,1,true); The iMiEV and Kia Soul command hierarchies marked the top-level command as secure but all the subcommands as not secure. This does not make sense as the subcommands could not be reached if not enabled. Furthermore, a subcommand to unlock the doors should be secure indeed. This commit changes all the subcommands to be secure. Similarly, the 'status' subcommands for 'server v2' and 'server v3' were marked not secure, apparently with the intention of letting the status be examined without enabling, but the top-level commands were marked secure so the status command could not be reached. The status subcommands are now marked secure. -- Steve

1 0