I have an update, but not really an answer as to what the exact problem is/was.
I looked into the driver issue first today. I discovered that Windows insisted on loading version 7.0.0.1 of the Google USB driver, even if I told Windows that I wanted the newer 8.0.0.0 version. My understanding is that even the 7.0.0.1 driver should work, but my working Vista system was using the latest 8.0.0.0 driver so I wanted my not working Windows 7 to also use that version. When I used the Windows GUI to update the driver, it would say that a reboot was required. I would reboot, plug my Nexus 4 in, and it would go right back to using the 7.0.0.1 version. So I completely removed/deleted 7.0.0.1 from Windows, and then had to convince Windows that the 8.0.0.0 version was compatible with my Nexus 4, then it would install and be used. Unfortunately, that didn't solve my problem. In my original posting about my problem, I failed to mention that I'm running adb via the Cygwin bash. I have checked and rechecked that I have my PATH set correctly. I know that when the adb server is not running the first run of "adb devices" will start the server. adb.exe is also the server, but run with magic command line option to make it the server. So the adb I run from the command line runs the second adb which is the server. I even checked to make sure that the second adb was in fact the adb I wanted it to be. It was. I continued to try all sorts of combinations of things. So somewhere in here I ran "adb devices" from a straight up DOS command line. Boom, it worked. Once the adb server was running via the DOS command line adb, then a bash run adb would work fine. So on my system, the adb server must be started from a DOS command line for some reason. I broke out some of the Microsoft sysinternals (procexp, tcpview) and looked at the adb server started via bash and compared it to an adb run via DOS. This is not the sort of thing I do very often so I was only able to do a cursory what's different. I couldn't see what was different between the two. I looked at what files they had open, what ports they opened, what their environment variables they were passed, their permissions. I just couldn't see the difference. The environment variables where where the most differences were, but I couldn't find a difference that would explain the differing behavior. I even started a DOS cmd line from withing a bash. That also didn't work. So my status as of now is that I don't know what the problem was/is, but I'm able to get my work done (for now). I say for now because I obviously have some sort of problem with my environment and that sort of things usually comes back to bit you. As a workaround I have a reset_adb.bat file that I run from a desktop shortcut. Once I do that the adb server is properly running, and I'm able to perform an "adb devices" from a bash window and it works. I have not fully tested all the Eclipse functionality yet, I'm hopeful that it'll be fine once the adb server is running properly. I'm moving on but would love to know what the real problem is/was. I'll reiterate this, my environment was working fine before I installed a bunch of Windows patches last week. I certainly know better than to assume that the root bug is Microsoft's. I could easily believe that they fixed some security related hole and that exposed a bug in adb, bash, or something in my setup. Anyway, maybe this will help someone someday. :-) Lee On Saturday, October 12, 2013 6:16:04 PM UTC-7, Lee wrote: > > I have a new problem with my Android development environment. In it's > simplest form, I have a Windows 7 PC and a stock unrooted Nexus 4. I have > been developing via this environment for at least six months. I'm using > the ADT bundle. I've on R 22.2.1 SDK tools. My ADB is 1.0.31. Basically, > from what I can tell, I've got the latest everything, except not the latest > Android Studio stuff. I'm primarily work with my Nexus 4, but every once > and a while, I try it out on a Nexus One and a Nexus 7. > > This environment has been working well form me for at least the last six > months. On Thursday 10/10/13, I installed all this months Microsoft patch > Tuesday updates. My system was working fine on Wednesday, I installed the > updates on Thursday. On Friday, I got back to work and was not able to > connect adb to my Nexus 4. > > Obviously I have looked online for a solution. The thing that is actually > surprising me here is that I fully expected other people to have this same > problem. But I haven't seen a single sign that anyone else is encountering > this problem. There are tons and tons and tons of people having problems > with adb over the years, but that's all old stuff with old well known > solutions. > > I've tried so many combinations of things. Reboots. Different USB > cables. Different USB ports. Different PC drivers. I've fully > uninstalled the Google USB drivers and reinstalled them. > > When I attempt to turn debugging on in my N4, I first turn developer > options on, then I enable USB debugging. I hear the PC beep indicating > that it needs to do some driver stuff. From what I can tell, that is all > working. When I run the old 'adb devices', I get the serial number of the > N4 followed by "unauthorized". On the device, the authorize dialog never > appears. I used to have this PC set as remembered by the device. I've tried > the "Revoke USB debugging authorizations" many times. > > I have tried every combination of enabling developer options, MTP vs PTP > modes, and setting these before and after the device is physically > connected to the computer. > > Here are a few key this works and that doesn't work facts. My primary PC > is Windows 7. ADB will work with my Nexus One. It will not work with my > Nexus 4 and Nexus 7 (both are running stock unrooted 4.3). This makes me > think the problem is in the authorization part of things vs a basic > connectivity thing. > > I have access to a Vista machine. I applied all the Microsoft patches for > this month to that Vista machine, and it does not have this problem. Vista > fine, Windows 7 not fine. > > fastboot is able to communicate with the N4. > > I was going to roll back the Microsoft updates, but I do not have a system > restore point old enough available. > > I'm open to trying anything I've already tried over again, but I really > think this is a new problem. Although, since I'm the only one who seems to > have this, who knows. > > I am completely making this up. It's like adb doesn't know who the PC is > anymore, but doesn't know it doesn't know, so when it sends the 'do you > authorize me' command to the N4, it's sending an invalid fingerprint, and > the N4 silently ignores it. > > Anyway, thank you for any help. Lee > > -- You received this message because you are subscribed to the Google Groups "Android Developers" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/android-developers?hl=en --- You received this message because you are subscribed to the Google Groups "Android Developers" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
