Why an Error Browser?

I have now presented the new Error Browser by building DBDOC Version 10.5 at six Australian DBDOC sites this past four weeks. The response has been universally ecstatic.

Severe errors were relatively easy to work with before Error Browser. The messages highlighted things that were clearly wrong, whether or not they were actual problems. Every site with DBDOC worked over the error messages, because we find significant things that can bite.

However, we find many things that are "wrong" and that might cause problems, but that are likely to just be cosmetic. Looking for needles in haystacks (and they are there) could not be justified. Actual errors thus were missed. Furthermore, conscientious DCS people end up with a whole mess of things to be ignored for reasons including the following:

• Logic, graphics and tags simply not used - by far the most usual reason for an otherwise significant error to be ignored.
• Logic cannot be fixed without a shutdown - making notes on the CLD or CAD sheet and documenting problems that cannot be fixed without a cold reload is often necessary.
• Issue is cosmetic - the error stems from a systematic approach like using a macro with functionality that is not needed in all instances.

Until now, the error messages we presented the DBDOC users came back each rebuild. Until the system configuration, graphics or tags could be changed, they were seen again and again.

 

The real strength of Error Browser, however, shows up in the messages we have traditionally called "CHECK" ones. They simply could not reasonably be examined and analyzed before. Now they can.

 

Error Browser makes it a whole new ball game. Sometimes all the messages in a particular class can be hidden as unworthy of further attention. The easy ability to walk through hundreds of messages can make it possible to find the few that need attention. Mark them all as hidden, walk through, star the ones that need attention, in fact, and unhide them. Clients reduced hundreds of mostly innocuous messages to a few nocuous ones (is nocuous the opposite of innocuous, I wonder).

 

Error Browser was first requested by a client in 2002. It is now here. My experience in Australia, with good DBDOC clients, is that it is not too late. The INFI 90 sites without DBDOC need it more than ever, because system integrity now is achievable on top of fast, effective, read-only fault-finding and trouble-shooting (Australians do fault-finding).