check hfsplus in linux
After a hard reset of my server i found that several afp shares are lost, but i can easily get into the folder by ssh to the server. weird, then i realize it might be the erroneous .appledouble file, so i try to remove it. Wth, the file system turns into read-only, then i checked the /etc/mtab, it clearly indicate the mount option is rw. So i googled it, then i found out the file system need to be checked before it can be mounted as rw after a unclean mount. So all i need is hfsprog package, indeed the ubuntu can tell you to install this package if you try to run fsck.hfsplus when it was not installed. after install, do sudo fsck.hfsplus /dev/sdxxx, All done!
install rt2870 official driver under ubuntu
the whole process can be easily achieved following the readme file inside the tar archive, except for the only edition of common/usb_dev.c, to add the supported device id of you card, in my case the planex one, you can get this id by lsusb command.
Fluid-GTask
If you are trying to run GTask as a webapp using thirdparty webapp engine, than you would probably wonder what is the address to use for GTask as google didn’t clearly provide the independent gtask urls, ok, after a little bit of digging on the internet, I found this one: https://mail.google.com/tasks/canvas, looks neat.
vim plugins
solve bluetooth dongle stop working after resume from s3 sleep
indeed this may related to inability to power off/on usb dongle properly, becasue this devise was suppose to work constantly on, but s3 sleep will power off all the peripherals thus lead to malfunction of bluetooth dongle. the best way so far is to use hibernate (thanks to chameleon, hibernate mode 1 works perfectly), maybe s2 would also work but that is not so eco-friendly.
By turning the hackitosh box into hibernation will cause another problem, which is quite annoying: the mds will start endless working after resume. I think this related to the external mechanism brought to the hibernation process by chameleon, to be more specific the hibernation will use certain amount of disk, this is not aware by the os, so it assume there is a large chunk of data has been renew, so it called for new mds indexing. the best way is to disable spotlight on all the disks(by exclude all of them in privacy tab which is under spotlight preference.)