Re: [HC Evolution] Re: libole2 -> gnomevfs backend?



    Michael>    Hmm; herein lies an interesting quandry. The real
    Michael> reason the merge was not done was licensing, it was
    Michael> thought that leaving libole2 GPL was a good idea [ after
    Michael> some discussion ], although libole2's design should lend
    Michael> to rather simple vfs-ing.

  Uhm, I am not sure what you mean here.  Do you mean that adding a
libole2 backend to gnome-vfs was not done because libole2 is GPL and
gnome-vfs is LGPL?

  Actually, gnome-vfs still has a problem with modules based on GPL
libraries, because it has no license protection.  This means that you
can link a proprietary application with gnome-vfs, and then gnome-vfs
might load a GPL module, thus violating the license.

  But this is actually rather easy to fix and will have to be added
anyway (it has been in the gnome-vfs to-do list for a loong time).  (I
would do this right away, but I have no time ATM.)

  Actually I do think keeping libole2 GPL is a good idea myself.

    Michael>    If I was going to choose a robust in-file filing
    Michael> system to store all my vital mail in I would not
    Michael> automaticaly choose libole2 :-) Although it has had a
    Michael> load of testing there are still a few skeletons in the
    Michael> closet ( eg. seeking beyond the end of a stream doesn't
    Michael> extend it etc. ).

  What about efs then?

    Michael>    Possibly it is a good idea, certainly it shouldn't be
    Michael> too tough though quite why it is any better than
    Michael> eg. tar.gz or zip format is beyond me.

  I think the problem with `tar.gz' and `zip' is that they are not
going to be fast.  It's a real pain to update them.

  But, incidentally, I must say am not very fond of this
filesystem-in-a-file idea myself.  Why not just using a directory?

-- 
Ettore




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]