Encounter a strange problem to load PyDrive in Google DataLab:
--> 141 @_helpers.positional(3)
142 def run_flow(flow, storage, flags=None, http=None):
143 """Core code for a command-line application.
AttributeError: 'module' object has no attribute 'positional'
Of course, no apparent results and solutions in google. But after a few trying, there is a workaround to load it properly: change the kernel type and change it back. What a weird problem!
Just compelled to upgrade to OSX High Sierra, because XCode 9.3 is only available for 10.13 to support development over iOS 11.3! The system seems quite the same except for the new ridiculous Chinese Translation of UI.
As usual, a Time Machine backup was made before the upgrade. But since NAS disk is quite full, some history backups need to be deleted after everything runs OK. Well, the new finder is frozen during the Time Machine deleting process and un-escapable from that UI. After a hard reboot and some googling, the following steps might help:
- connect to target Time Machine Sparse Bundle
- ensure that /Volumes/Time Machine Backup/Backups.backupdb shows
- sudo tmutil delete /Volumes/Time Machine Backup/Backups.backupdb/YOUR-MAC/SOME-DATES
- eject Sparse Bundle
- sudo hdiutil compact TARGET-SPARSE-BUNDLE
Just encounter strange high cpu problem of kernel_task in Sierra on my rMBP. At first, it seems like some software problem, especially when opening atom editor, which causes long opening time. After closing atom, it does lower the cpu for a period, but running other task heavy software afterwards will also boost the kernel_task cpu percentage to 80%. It may suggest some underlying hardware problem there.
Comments in this post give the reason: temperature and fan control. Since it’s summer and super hot, the osx on my mbp is reducing cpu speed for temperature compensation, causing kernel task to take more cpu to consume. Resetting SMC in this article brings the fan control back and fixes the kernel_task cpu problem.
First thought is some bluetooth problem, tried everything related but not working at all. After a few trial and inspection, it seems that the thumb button is stuck in between causing the mouse pointer not able to move smoothly.
Disabling the thumb side button in Logitech Options won’t help. It sounds like a hardware interference not a software driver setting.
This Inkandescent’s post and this operation guide will help fixing it. The pull-up workaround in the previous thread doesn’t work very well, so buying the T5 screw driver and loosing the thumb button board screws is the final solution.
This blog has detail instructions to install USB WinToGo for Mac. But things won’t go that smooth on pre-2013 rMBP.
Firstly, diskpart from Windows cmd within Virtual Machine seems to have limitations, especially for Win8.1. Partitioning can work sometime but formatting will probably fail. Therefore, diskutil in Mac is the solution for disk handling. But another problem occurs. FAT32 partition is tried but found to cause errors during imagex transfer due to large file size restrictions. But disktuil doesn’t support NTFS. At last, NTFS partition is formatted with diskutil with the help of Paragon-NTFS 14 in El Capitan.
Secondly, “bcdboot /f ALL” will create EFI boot mode for usb drive, under which Win2Go can be successfully installed but instantly turns into black screen when installing bootcamp/nvidia drivers. It’s because that pre-2013 MBP is not fully compatible with UEFI crashing the graphics/audio drivers under EFI boot. The solution is to boot with CSM-BIOS mode. That means when booting MBP with alt, “Windows” is the only menu choice not “EFI”. Clearing the boot partition and “bcdboot /f BIOS” for BIOS boot mode section. However, the “none system disk” error shows when booting the “Windows” option.
Finally, this article sheds the light. “bootsect /nt60 x: /force /mbr” to force using the win8.1 bootmgr will boot your MBP into Win8.1, with all bootcamp drivers ready to be installed.
Enjoy the Win2Go, thanks for the almighty Google and a whole day of rebooting!
My windows 2003 vm in vmware fusion turns into a black screen after an OSX crash. The guest system seems bootable but starts with total blackness. Fortunately safe mode works. Event log shows it’s unable to load vmx_fb.dll, which is the graphics driver of vmware tools.
The paradox is that vmware tool is unable to install in safe mode, which will always fail even though MSI service can be enabled with this method. Then manually deleting vmx_fb.dll and vmx_mode.dll in system32 folder also not works.
It seems the only way is to roll back vm to 2-month-ago snapshot losing all the intermediate stuffs. But this thread is the rescue. Additional vmx_svga.sys in drivers directory should also be deleted. Then the guest 2003 system can be started with the default vga driver.
Now enter the vm, reinstall vmware tools, and make a new snapshot for such emergency case.
Just encountered a mysterious problem after upgraded to 10.11.4 that every time MBP reboots and logins, my trackpad settings like tap-to-click are lost. It’s mysterious because during the very beginning time after login it worked as normal, but a few minutes later trackpad settings were reset automatically. It also affected the Mac AirPlay to AppleTV which would not work except for the first few minutes after reboot.
Googling the airplay problem leads to login items. But removing all of my existing login items doesn’t fix the problem. (Of course, the trackpad plist is the first try which also doesn’t work.)
And the console log during possible resetting trackpad has messages like “kernel: Sandbox: CommCenter(3359) deny(1) file-read-metadata /Library/Developer” and “otherbsd: Could not submit LoginItem job com.XXX.XXXYYY: 119: Service is disabled”;
More googling doesn’t find much useful info. Similar keyboard reset problem exists but has no fix yet. When analyzing the details of the console logs, one interesting thing is that there’re two login window app started msg around that time, when the failed login items are all at the second login window session and possibly kill my trackpad settings and airplay.
Then why on earth exists another login window session? The culprit is _xcsbuildd. As 10.11.4 upgraded Mac Server, with XCode server service included, which will run as user _xcsbuildd at background. The upgraded service’s second login session breaks some login items which seems indeed killing airplay and resetting the trackpad settings.
Simply disable xcode service in Server.app will not remove _xcsbuildd related stuffs. Use this methods to totally get rid of it.
Finally reboot and enjoy the hassle-free trackpad and airplay again.
One of my relative’s iPad2 got stuck in the iTunes logo interface and unusable for a while. Just tried to get it restored when connected to iTunes via USB. However, it always auto rebooted after progress bar appeared under the Apple logo.
Tried DFU mode and restore still fails with “Unknown Error 9”. This thread googled suggests the unbelievable culprit: the latest iOS 9.2.1 is incompatible to restore iPad2.
Since older iOS version cannot be signed and updated by Apple now, latest iOS 9.3 beta 4 works fine to restore iPad2 successfully.
My iPhone6+ got stuck when uploading some photos automatically to iCloud Photo Library. It’s quite possible if the phone’s available free space is lower than 400MB when using optimizing size option, which might be required by the mysterious blackbox photo upload process. But after cleaning up disk space, it still paused at the “uploading xx photos” phase.
Tried everything googled: reboot phone / disable-enable iCloud Photo Library / delete videos / logoff iCloud account, nothing worked.
Finally, forget about the iCloud synchronization. Disable Photo Library, transmit these photos to Mac and manually upload photos to iCloud.com, then delete all existing photos from the phone. And at last re-enable the iCloud Photo Library to download all photos again to the iPhone.
Hope it will sync next time.
Just encounter a strange problem of key chain access. When prompting for access to credentials, the dialog can only be dismissed by Deny. In the console, “Ignoring user action since the dialog has received events from an untrusted source” was logged.
According to this post, 10.11.1 contains a security patch to prevent synthetic clicks on keychain dialogs. Try disabling all apps in the privacy settings of accessibility-mac-control, but not works.
Therefore manually quitting such apps one by one will be the trick to check for the culprit. It founds that BetterTouchTool is the problem.
Update: 10.11.5 will correctly prompt for administrator password and shows the blocking app.