|
PureBytes Links
Trading Reference Links
|
I appreciate the problem reports that are sent to the list, even though
I, or others, may not be able to reproduce them. Some people seem to
believe that the problem is fictional if it cannot be reproduced by
others.
I suggest that there are at least two kinds of problems. The easy kind is
the one that can be reproduced be most people, but not necessarily all.
The hard kind is the one that is not reproducible by many people. These
tend to be dismissed by some as non-problems.
My recent report of a charting problem is an example. For one day, I could
cause the problem to occur at will, It was repeatable many times under a
variety of conditions - new window, new workspace, restart charting,
restart everything (except NT), and others. It looked like a very
repeatable problem.
The next day, with the machine left on overnight, and since then, I am
unable to cause the problem to recur. Now, was that a fictional problem?
Was I imagining that it happened so many times? I think not, but there
are some out there that would still maintain this is the case :-).
Software has become very complex. This increases the probability that some
*slightly* different combination, or sequence, of events can cause a
problem. And, this combination might only occur with the particular
hardware and drivers on that machine. Using the latest "bleeding edge"
bells and whistles often introduces another layer of problems. These
'intermittent' problems are the hardest to fix.
This puts Omega in a tough position. TS2k is a new rewrite, according to
them. Initially, they will have their hands full with the easily
reproducible problems. As these get fixed, more attention can be paid to
the serious intermittent problems.
Back to probability. If there are a large number of intermittent problems,
then the likelyhood of having at least one of them becomes significant.
This means that Omega could work hard to fix problems, yet a
mission-critical product could still be viewed as unacceptable.
By reporting all problems, even intermittent ones (but clearly identified
as such), we can get an idea of the actual stability that one individual
might expect to see. Omega, of course, would like us to believe there are
no problems.
I believe it is a good idea to continue to report problems to the list,
even though it might be considered as 'whining' by some. Reporters might
be labeled as 'non-producers' too - but what you are producing is
information - information that can be very useful in making a decision
about which software to use, or when it is stable enough to use.
What do you think?
Larry
|