fpgroups.com: Tip of the day!

[John Esak]

> Now I am always gun shy when I hit F5 in this case, I have had the
> deadly "Unreferenced memory error" that amounts to a seg violation on
> Unix and lost anything I had done in the table.  It happens often enough
> that I save before hitting F5, but not predictable enough that I have
> been able to define a bug report for Ken.
> 
> It seems to be an issue related to how long I have been working on the
> table.  The more time I have spent working on the logic the more likely
> it is that it will crash.
> 
> So add that to your list of stuff to worry about.
> 
> Nancy
I don't know... my guess is the problem you are talking about is solely
something that happens on Windows versions of *cabe.  I have *never* lost
any data in a processing table except for one time and one time only.  The
bug where you would lose everything you had done if you pressed F5 and then
D and then L for long variables... when you had one declared twice.  That
would bang you out of the table immediately and you lost everything new.
However, this was fixed the day I told FP Tech about it.  Other than that, I
have never seen an "Unreferenced memory..." error on a *nix system. Never in
*cabe, and never in the runtime programs. I have seen lots of SEGV's but
again never in *cabe.

I have seen "memeory" based errors on Windows systems often enough in
earlier 5.x versions. I haven't seen any in 5.6.

So, I wouldn't worry about *cabe losing anything under *nix... just haven't
seen this.  Thank god.

John