Searched around, some said to remove the face case to press and hold the reset button but for my camera there was no case to remove but there was a small cover what a gear icon saying ‘SETTINGS’ under it:

It can be removed with a phillips screwdriver then we can long press the small floating button/lever on the right of where it says ‘RESET’. Then the cam would say something like reset success and shortly it will finally find the wifi dvr and blink blue, where you can finally see and add the device normally in the app.

Searched around, found this: OdinV4 for Linux - worked like a charm! Full credits to the authors of the project, please star their project!

Just make sure you put all the files you get from samfirmware in their respective param/arguments to the odin binary - I wasnt sure which arg for the HOME_CSC_OXM_S926BOXMACYJ7_MQB102650869_REV00_user_low_ship_MULTI_CERT.tar.md5 file but tried -u and it worked and now phone is working fine with android 16.

Having used Nanoc and Jekyll for probably a decade, I was looking for easy ways to put stuff from phone without a computer while being a static website. I was amazed to find out how easy it was to make any website via github pages (that you can link to your domain name) free and simple just by forking via the instructions described in Academic Pages github repo’s readme - without a computer just through a phone nowadays, the pre-configured github actions handle the auto build so seamless! Really recommended.

Of course, since I had to migrate old posts/data, I had to use my computer (with Emacs of course) to edit and migrate old stuff.

The integrated github pages/actions/defaults/examples made a very good base to start from.

Truly appreciate and would like to hereby thank all the authors of academic pages and jekyll for their outstanding work!

Having used pytest, nose, nose2 for many months and faced random issues like some tests would fail when running in a set (not concurrently) together with all other tests but pass if ran alone (whereas the target use case was to be run in its own process alone), and also random issues in getting tests to continue from the last failed one (in larger projects with 300+ integration tests) - I was looking for a solution that can run tests concurrently, in their own processes, and can continue from the last failed (not having to start again from the first test). Tried pytest, nose, nose2 but they dont seem to fit this need although one of them had a continue from last failed feature but randomly it would start all again..

What if we can use the tried in true ‘GNU Make’ as used in most C projects like the Linux kernel sources to run python tests? It can run concurrently like make -j4, it continues from last failed just by running make, start over with make clean then make - yes it worked perfectly for my needs! Just make sure you handle __main__ to run your test function in each test_<test name>.py file so that each test could be run with python test_<test name>.py.

To use this, just use copy this Makefile from my github repo into your python project folder then run make: https://github.com/ykasidit/python_make_test_example/blob/master/Makefile

Finally, below video shows above config in action:

char cmd_res_line[256];
char total_cmd_res[25600];


//WARNING! NO HANDLING FOR VERY_LONG COMMAND READS YET
char* exec_get_out(char* cmd) {

    FILE* pipe = popen(cmd, "r");

    if (!pipe) 
      return NULL;

    total_cmd_res[0] = 0;

    while(!feof(pipe)) {
        if(fgets(cmd_res_line, 256, pipe) != NULL)
	  {
	    //TODO: add handling for long command reads...
	    strcat(total_cmd_res,cmd_res_line);
	  }
    }
    pclose(pipe);
    return total_cmd_res;
}

...

char imsi[16];
imsi[15]=0;
char* res = exec_get_out("service call iphonesubinfo 3");  
if (strlen(res)>210) {
	imsi[0] = res[75];
	imsi[1] = res[77];
	imsi[2] = res[79];
	imsi[3] = res[81];
	imsi[4] = res[136];
	imsi[5] = res[138];
	imsi[6] = res[140];
	imsi[7] = res[142];
	imsi[8] = res[144];
	imsi[9] = res[146];
	imsi[10] = res[148];
	imsi[11] = res[150];
	imsi[12] = res[205];
	imsi[13] = res[207];
	imsi[14] = res[209];
	printf("imsi: %s len %d\r\n",imsi,strlen(imsi));
}

This code uses 2 methods, should work on most single-sim GSM/WCDMA/LTE phones.

#include <sys/system_properties.h>

	//returns the string length of the value.
	int ir = __system_property_get("ro.gsm.imei", imei_start);           
	
      if(ir > 0)
	{
	  imei_start[15]=0;//strz end      
	  printf("method1 got imei %s len %d\r\n",imei_start,strlen(imei_start));
	  strcpy(g_imei,imei_start);
	}
      else
	{
	  printf("method1 imei failed - trying method2\r\n");
	  //old dumpsys imei getter
	  char* res = exec_get_out("dumpsys iphonesubinfo");  
	  const char* imei_start_match = "ID = ";
	  int imei_start_match_len = strlen(imei_start_match);
	  char* imei_start = strstr(res,imei_start_match);
	  if(imei_start && strlen(imei_start)>=15+imei_start_match_len)
	    {
	      imei_start += imei_start_match_len;
	      imei_start[15] = 0;
	      printf("method2 IMEI [%s] len %d\r\n",imei_start,strlen(imei_start));
	      strcpy(g_imei,imei_start);
	    }
	}      

Simply open an adb shell and enter:

1. ./sqlite3 /data/data/com.android.providers.settings/databases/settings.db "update secure set value='2' where name='preferred_network_mode'"
   

</pre>

1. after “su”, <pre>killall com.android.phone</pre>

Done.

The step 1 changes the settings database - but it is not reflected in the settings app or known/reloaded by com.android.phone (which I understand that it does the real forcing…) yet - we can reboot but that takes too long and unneccesary - I found that killing the process of com.android.phone would cause some watchdog to re-start it (step 2) - and re-load our new settings into effect (probably by calling setpreferrednetworktype with our new values - I think we can’t call that internal java function directly due to signature/key restrictions).

I’d like to try to find some time to make an app to do/automate this - it helps rooted phones with special firmwares have that missing option to force directly - very beneficial for programmatical/automated forcing - this is an alternative to the manual/GUI method using the ##4636## > phone info > Setpreferred network type method.