A minor problem w/ access to my flash drive

[Alexey Morozov]

Pozsar Balazs writes:

>>I really don't talk about card readers and other 'fashionable' stuff. I 
>>guess there should ways to automatically distinquish between card 
>>readers and 'plain flash drives'. Am I wrong?
>>    
>>
>Apart from usb id database, there no way to distinguish unfortunately :(
>  
>
:-(

>>Yes, I wrote one, but w/ exception that hald now doesn't try to detect 
>>volume fstype, so no fs-specific options are applied... Could you 
>>suggest smth?
>>    
>>
>
>I think the iocharset=utf8 vfat mount option should be a kernel mount 
>option, and not hal should provide it.
>
>Look at the .fdi files, I suppose it is possible to merge the filesystem 
>attributes also which should work well for your own case.
>  
>
Well, iocharset is certainly a kernel mount option, and it's written to
fstab. But unfortunately not all filesystems actually support it and it
makes sense only for certain types of filesystems. So /an agent/ should
first detect the filesystem (as hald does w/
storage.media_check_enabled=true) and form appropriate kernel options
mount line. But when storage.media_check_enabled=false hald doesn't even
try to get the fstype and writes 'generic' options line: auto
noatime,sync,exec,noauto,user which is suitable for [most] filesystems,
but not completely convenient for fat (and several other FSes). If I
turn media_check_enabled on, I got vfat-specific options including
iocharset and the polling comes back;-)

So I guess, there should be two different keys, smth like
'initial_media_check' to correctly form mount options and
'repeat_media_checks' to actually check if the media is still there...
Maybe initial_media_check should be done automatically w/o visible user
configuration option.

Alexey


_______________________________________________
hal mailing list
hal at lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/hal