i don't know if this relates to the code in question, but i have had
trouble with metakit data on shared drives also.  even if the server
is a windows box, in fact even if the server is the same box the app
is running on, using a share path when writing metakit data corrupts
it, while writing to a local path (eg C:\file instead of
\\server\C\file) is fine.  reading from a share is ok, though.

i'm using 2.4.9.2 with C++, but i first had the problem with some
earlier version.

(btw, does anyone know of a 'localhost' equivalent for windows?  it
would be nice if code could have paths like "\\localhost\share\file"
in it, instead of having to figure out the name of the machine it is
running on.)

On Mon, 20 Sep 2004 17:22:55 -0600, William K. Volkman
<[EMAIL PROTECTED]> wrote:
> Humm...seems to be a problem with the VMware/Samba mounted
> file systems.  I moved the '.mk' file to my C: drive and
> now am seeing correct output (at least from this simple
> program).  How bothersome, a whole day wasted chasing this.
> 
> 
> 
> On Mon, 2004-09-20 at 16:51, William K. Volkman wrote:
> > On Mon, 2004-09-20 at 16:41, Brian Kelley wrote:
> > > I'll bet you a beer that I know what you are doing wrong here (because
> > > I've done the same thing :) )
> > >
> > Where are you? I'd like to collect (just kidding ;-)
> >
> > grep -Hn open dbrestore.py
> > dbrestore.py:42:r = open('backup-data.sav','rb')
> >
> > Good thought though.
> >
> > > If you load or save on windows, you MUST supply the binary flag.
> > >
> > > read
> > > file = open("foo", "b")
> > >
> > > write
> > > file = open('foo", "wb")
> > >
> > > Another reason why windows stinks (IMHO).  If you want portable
> > > metakit, whenever you use save or load just add the stupid binary
> > > flag.
> >
> > The attached simple program pukes garbage on the screen
> > and after a couple of runs gives me the
> > SystemError: error return without exception set.
> >
> > Do you have a Windows system that you could test it on?
> --
> William K. Volkman <[EMAIL PROTECTED]>
> 
> _____________________________________________
> Metakit mailing list  -  [EMAIL PROTECTED]
> http://www.equi4.com/mailman/listinfo/metakit
>
_____________________________________________
Metakit mailing list  -  [EMAIL PROTECTED]
http://www.equi4.com/mailman/listinfo/metakit

Reply via email to