[BBLISA] Mapping inode numbers to file names

[John Stoffel]

>>>>> "Edward" == Edward Ned Harvey <bblisa3 at nedharvey.com> writes:

>> From: bblisa-bounces at bblisa.org [mailto:bblisa-bounces at bblisa.org] On
>> Behalf Of John Stoffel
>> 
>> My personal feeling is that you should just move up a few levels and
>> search down from there, and not spend time searching all over the
>> filesystem for something, since I would assume (and we all know what

Edward> More likely: Scrap the search idea.  Note in the man page,
Edward> "Due to present limitations of the filesystem, zhist can only
Edward> find snapshots for objects for which the pathname and filename
Edward> have not changed.  For more information, including
Edward> instructions to perform an exhaustive search, see http://...."

It's a hard hard problem I agree.  As disks get larger, and people get
more forgetful, it's not easy to keep up with indexing stuff.  Esp if
the performance hit is too high and impacts the user's experience.

Edward> The result really needs to be returned instantly.  I can't
Edward> afford to search around random unknown directories of unknown
Edward> sizes.  And the assumption that "most moves are relatively
Edward> local" is of course, user behavior driven.  It's funny you
Edward> should mention, because just 30 minutes ago, I got an email
Edward> from a user saying "I downloaded this tool, can you please
Edward> install it" so I moved it from her home directory to the
Edward> "standard download area" for installers, and ran it.

Heh.  I like it.  Modeling user behavior is all about the exceptions,
right?  And we all know how well computers handle exceptions they
don't expect.  Humans are much more tolerant.  

Maybe the idea of a decaying history of filesystem activities would be
usefull?  You generally (there I go again, generalizing!) look for
stuff you handled recently.  Except when you're digging through your
piles looking for something you touched "about two years ago... I
think" looking for something.  

I dunno... I haven't put alot of thought into this really beyond the
initial question, which we've drifted away from a bit. 

John