"Steve Simon" <st...@quintile.net> writes:

> This is the exact scenario I had when I ran fossil + venti with ephemerial
> snapshots enabled, are you sure you don't have them on?

Quite sure.  :) That's why I ended up recovering the data from Linux.

>       fossil/conf /dev/sd??/fossil | grep snaptime

I have just two lines: fsys main config, fsys main open.

Incidentally, that fossil/conf drove me nuts(!) until I finally
discovered it.  I tried switching the drive from /dev/sdC1 to /dev/sdC0.
I couldn't figure out why it kept insisting on mounting /dev/sdC1 when I
changed EVERYTHING in plan9.ini to refer to /dev/sdC0.  (Some day, I'll
get around to asking why someone thought that hiding the device name on
the device itself was a good idea. ??)

Reading about Plan 9, I was quite excited to install it.  I was quite
excited when I first booted and ran it, too.  But I distinctly felt my
heart sink a little the first time it hung.  Since then, I've browsed
some of the OS source code and, having done that, I came to understand
why the system was so buggy.  The core applications appear to be written
in a style of C programming reminiscent of the dawn of UNIX.  While the
operating system architecture is BEAUTIFULLY designed (with the
exception, perhaps of that fossil/conf gotcha!), the C code used to
implement it doesn't seem to take advantage of any of the programming
paradigms that have emerged in the intervening 30 years...

Getting Plan 9 code to crash is almost too easy:

term% mkdir trashdir && cd trashdir && mkdir x
term% touch `{i=0; while (test $i -lt 128) { echo -n abcdefghijklmnop; i=`{echo 
$i+1|hoc} } }
term% cp abc* abc* x
# watch the cp executable suicide
# now, make SURE there's nothing in this rio window that you want to keep...
term% rm abc*
# watch the rio window go bye bye!

I'm not someone to complain without also offering solutions, though.
I'm in the process of writing some C macros that might help clean up the
source code, ensure intended bounary conditions, improve some
interfaces, etc.  I already have some working code, but it's still very
experimental.

-- 
+---------------------------------------------------------------+
|E-Mail: smi...@zenzebra.mv.com             PGP key ID: BC549F8B|
|Fingerprint: 9329 DB4A 30F5 6EDA D2BA  3489 DAB7 555A BC54 9F8B|
+---------------------------------------------------------------+

Reply via email to