This is only marginally related to Paradox in any of its forms, but related
to programming in general. On an annoyingly regular basis, I see the
following on MySpace:

Sorry! an unexpected error has occurred.
This error has been forwarded to MySpace's technical group.

Thinking about the programming books I've read over the years, the ones
which deal with good programming habits in any language, the idea of an
"unknown error" strikes me as absurd, at least for any application* that's
been around for a while. So, when I see this on MySpace, my first reaction
is "sloppy and they don't give a damn as long as the advertising revenue
keeps pouring in". Am I being harsh? Partially correct? Opinions wanted.
Assuming I'm trying to access a feature that's not new, is there really any
such thing as an unknown error?

* I'm choosing to consider web sites to be applications at the moment.

As you've suggested, numerous whole BOOKS have been written on
high-quality programming... In a "normal" commercial program, perhaps
half of the code is dealing with "errors", and these are only the
"normal" errors that are expected.

In the PC arena, for many different reasons, a vast array of errors can
arise that are not usually handled within the app code, and the
programmer will hope the operating environment, hardware drivers, etc,
will handle them adequately.

As a "Hello World" example:
Input "A?: ",A
Input "B?: ",B
Print "A+B=",A+B

What if:
- the sum of A+B exceeds the arithmetic capacity available, just as
if one typed 2 huge numbers on a 2-dollar calculator?
- the user types "Hi, There!" as one of the values to be added?
- the Keyboard is stuck, and the program just sits there waiting for "A"?
- the hardware to display the result is not functioning properly?
- ...etc, etc, etc...
This tiny program could turn into a monster, if all "Known" errors were
to be handled adequately. No to mention RAM limits, Disk limits, LAN
issues, comms troubles, timing problems, failing hardware, bugs in
drivers and operating systems, hardware bugs..... I recall the
"Bug-list" for a very early Pentium CPU design amounted to 40 pages.

And... another huge issue is trouble arising within error-handlers
themselves... The code is trying to trap all relevant info for later
debugging, and this code breaks, and loses the info of the original
error, and then all hell breaks loose...

....Or, maybe I mis-understand...

- Mike

"Michael Kennedy" <Info (AT) KennedySoftware (DOT) ie> wrote

I *know* I misunderstand, especially since MySpace was sold to Rupert
Murdoch's company for $580 million, a budget that would seem large enough to
enable them to hire people to make sure that the most basic features work
all the time.

I'm not a professional programmer, but I've learned enough here, and from
reading to know that a particular error is only unknown once, unless it's
not fixed.