hallas added a comment.
In D20938#459149 <https://phabricator.kde.org/D20938#459149>, @ivan wrote: > I'm torn between two approaches: > > - doing what you have done, maybe with a customization point - `fusermount -u` by default, something else for specific mount types; > - disabling the teardown operation > > The rationale for the second one: > - The users which create fuse mounts from the shell know how to unmount them; > - Users which used a tool to mount something (like Plasma Vault) should use the same tool to unmount (these tools potentially do more than simple unmounting - like calling a destructor in C++ instead of just free :) ); > - Some applications will have their own mounts (for example, akonadi will use encrypted storage for storing indexes of encrypted mails) - you don't want to have those controlled by the user.* > > (*) we will also need to hide these from Places, but that is not important at this point. I am actually leaning mostly towards your second suggestion, disabling the teardown action. I think you described it very good that the user already knows how to use fuse and in the cases where he doesn't (e.g. when using RDP or something) then he probably shouldn't unmount the filesystem anyway. @ngraham - do you have an opinion on this? REPOSITORY R245 Solid REVISION DETAIL https://phabricator.kde.org/D20938 To: hallas, #frameworks, ngraham, elvisangelaccio, broulik, bruns Cc: svuorela, nicolasfella, ivan, kde-frameworks-devel, michaelh, ngraham, bruns
