Well let’s see, the reason I outlined in my last email is one reason.
Well Robert, I may have missed it, but I don’t find where your last mail referred to Stretch at all. Instead it specifically referred to “snapshots”. Then you linked to a page shows links to BOTH Jessie and Stretch. So I was hoping to get your advice on whether the latest version of Jessie or Stretch was better for wifi.
That would be for you to do.
OK, great. I guess I will try to remove connman altogether in hopes that does not cause the kernel to rename my interfaces as Jacek has mentioned. It really sounds like connman is half baked. I’m having trouble seeing what problem it solves because it sure seems to bring in a lot of new problems.
Ah, where exactly do you see that the Bluetooth can be configured as WiFi? That’s a pretty stupid snake oil article…
Yeaaaa, it may be “stupid snake oil", but if it is you need to take it up with TI, who designed the chip and wrote the all the documents I read.
Here is some supporting information:
https://github.com/beagleboard/beaglebone-blue/blob/master/BeagleBone_Blue_BOM.csv This BOM shows that U7. the wifi chip on the Blue, is a WL1835MODGBMOCR. This is how I found the part number of the wifi chip (unless the BOM is wrong).
http://www.ti.com/lit/ds/symlink/wl1835mod.pdf This is the data sheet for the chip. Look at figure 1-1. It clearly shows that
RF_ANT1 can be connected to either Wifi or BT. Table 4-1 also confirms.
and then you can confirm by looking at this…
http://www.ti.com/lit/ml/swpb012b/swpb012b.pdf Look at the WiLink 8 Block Diagram. It also very clearly shows that Antenna 1 supports both Wifi and BT.
The documents I have found seem to show it is possible to have two wifi interfaces from this chip (at the expense of losing BT), but do not appear to show how to do it.
Does anyone know how I would proceed to switch RF_ANT1 to wifi? Hopefully it isn’t something as complicated as rewriting a driver.