Today's server meeting started out pretty normal and ended up in a small work session and trying to figure out how to modify some code for sub inventory. We started out and John reported on his yearend tax forms (W-2's and 1099's). We got a small report on the changes to the email server that Wayne is working on. We are excited to see if we can gain some more control and/or options there. We then switched over to the bug in the sub inventory stuff. This doesn't happen all of the time, but when it does, it throws an error.

Anyways, we spent the next hour or so going over different levels of pre-validation, in-line validation, and post-validation. Sometimes the problems can be stopped before they ever get going, some can be stopped once an error or 0 record count is hit, and some need to be soft or conditional errors that don't do something but allow for other things to still take place. For example, the soft errors may need to happen in a complex loop or nested loop. You may not want to just stop on a dime. You may want the loop to finish and then show problems or gracefully roll past any problems or issues.

Anyways, it got pretty deep. We will end up going in and looking at each page and decide what the best option will be. There is no quick solution for this issue. We may have to check and change code in hundreds of pages. Each one has a different task or function or flow. The quick answer is to just change it at the method level, but this same method is used in tons of places and in different ways. We will just have to chip away at it and/or watch the logs and only fix the pages that show up on the error logs. Once again, it only happens every once in a while and not all pages that use this method have problems. We have to track this one down using manual processes and time reading over the code before the method calls, during the call, and after the call.

To sum it up, it all deals with general error handling and making it more consistent. We are in a state of constant refinement. We will just keep working to make it better and better and smoother and smoother. I will note that we never would have found this error if it weren't for Wayne's monitoring and error logs. Basically, it allows us to look under the covers and catch things that a normal user would never report to us.

After everybody else left, Cory was talking about testing and making sure that we are only pushing up good code. Wayne told her that we push up code everyday, multiple times a day. That is totally true. She only knows about it if it goes bad or has an error or bug in it. Anyways, we'll have to keep pushing on things and we'll find some good balance points between testing, pushing, and progression. Making headway... yee haw!