|On Fri, 2004-09-24 at 11:47, Bram Van_Oosterhout wrote:
> > I recently found that there are updates for wvware on
> > http://sourceforge.net/projects/wvware/
> > I installed wv 1.0.2. It required some fiddling, but nothing too hard.
> > The good news is that I now parse 90%. Only 1 out of the sample of 11
> > fails (with Signal 11)
> > The bad news is that the conversion is very slow. The 10 documents took 9
> > hours to convert to HTML. They are all a similar format. Tables, graphics
> > and text. 3 to 5 pages.
> > I intend to move to a faster processor and a Linux implementation later in
> > the year. I expect that to be much faster.
> > In the mean time, is there anyone with better experience with respect to
> > the cygwin greenstone/wvware implementation?
That sounds like a bug in wvWare... in my experience some word documents
make it go into an infinite loop (this is under linux) so maybe under
cygwin it got stuck in a loop until the OS decided to kill the process
for some reason.
wv does much better with .DOC files from older versions of Microsoft
Office, presumably because the format is simpler/more understood.
Microsoft do not make the file specifications public, so the wv people
need to reverse-engineer the format.
I don't really have an answer for your problem. Under linux you could
add a 'ulimit -t 60' command or something to the converter so that it
isn't allowed more than 60 seconds of cpu...