[Ben Fowler <wapdev@xxxxxxxxxxxx>]

At 08:16 AM 6/27/00, Guy Harris wrote:
> On Sun, Jun 25, 2000 at 10:32:06PM +0100, Ben Fowler wrote:
> > Unless there is an obvious way to get that routine to bail out, it
> > might be a good idea to have file_quit_cmd_cb() stuff a suitable
> > USER_CANCELLED_READ value in the capture_file structure,
> > and have the read routine check this whenever it receives
> > control back from the event loop.
>
> I've done that (actually, there's now a state variable in there, which
> is either:
>
>         no capture file is open;
>
>         the file is open and in the process of being read;
>
>         the file is open, and in the process of being read, but the user
>         asked that the read be terminated;
>
>         the file is open and has been completely read).

Many thanks for your double plus good and speedy response.
This is exactly what is required. I have checked out a new
version of the sources and ethereal know works well for me.

Three virtual cheers for Guy, Gilbert and all the others!

> The changes also allow Ethereal to, if you do File->Quit while a live
> "update the screen as packets arrive" capture, kill off the child
> process (which the code does with a "kill(<pid>, SIGTERM)" in UNIX; no
> code is in there yet to do that on Win32) before exiting, so that it
> doesn't leave the child process hanging around.
>
> > Here is a patch which converts the global cf to cfile.
>
> Checked in.


Ben.

--
Leedsnet - The information resource for Leeds and the West Riding
< URL:http://www.leedsnet.com/mobile/ >