Re: gvfs: fails to mount certain removable devices

[David Zeuthen <zeuthen gmail com>]

Hey,

Sorry for not getting back to you earlier. As you've probably learned by now the GIO/GVfs code bases are pretty complicated; first of all there's the abstract interfaces (in GIO) / concrete, private implementations (in GVfs) thing. And as if that wasn't complicated enough there's out-of-process volume monitor daemons, out-of-process mount daemons ("gvfs backends"), FUSE stuff, mount shadowing / adoption / proxy volume monitors [1]. So, yeah, all the code is pretty complicated (granted, it does solve a complicated and hard problem), and, to a degree, suffers from organic growth, extending interfaces in API/ABI-compatible manners, code living out-of-process and other wear+tear you usually don't see in newly written code (porting from libdbus to GDBus for example).

Also, you probably don't want to hear this, but the gdu/udisks1 bits of GVfs are deprecated (as of GNOME 3.4, released spring 2012). Users should instead be using the udisks2 bits in GVfs - this is also where new development happens and where developers can help users with bugs. IOW, you are not likely to find a lot of interest from upstream developers in working on the GDU bugs. Yes, aspects of this reality sucks but that's just the way it is.

The good thing is that the udisks2 bits are simpler, more robust and better documented. In particular, there's this useful document

 http://git.gnome.org/browse/gvfs/tree/monitor/udisks2/what-is-shown.txt

explaining some aspects of what's going on and why.

In closing, you should probably use bugzilla to report the bugs.

Hope this helps.

Thanks,

David

[1] : http://developer.gnome.org/gio/unstable/GMount.html#g-mount-is-shadowed