Comment 61 for bug 213494

Hi,

Thanks for the help everybody, I see now what a strange bug
this is.

I spent some time this morning reproducing this error, and though
I could trigger it I couldn't find problems in the reading of the history
file. I'll try again, but I don't know whether that will reveal anything.

I don't want to upload the sleep() patch, as that won't fix it for everyone,
and will just tend to slow things down.

The try: patch would be more acceptable, but as the applet can be
added once the session is up and running it seems that it would
only lose things from the history when it doesn't need to.

Writing this made me realise that this could well be caused by reading
from the file from two different threads, and I have just checked and
the applet does indeed use threads, so that would be an avenue worthy
of investigation.

Thanks,

James