I have a dilemma - when is a bug worth fixing? This is a hard question to answer. Here is some insight to how we would treat it pre-RTM and to a point, post RTM.
Lets say I have a clear cut bug in the product , however the product is not current (lets say Win2k) and there is a relatively easy workaround. Do we still fix it?
What if the chances of hitting this bug are very great, thus generating support calls? Do we write a KB article and hope folks find it before they call PSS? Even then, we hope the PSS engineer finds it in short order (luck of the draw there sometimes). We can also look at it from a code standpoint and if we determine there is a high chance of regression then not to fix it, but is that always right? What if this is an especially ugly bug – perhaps resulting in a bugcheck? All of these things are items to consider when filing a bug, or even before I begin debugging an issue.
Currently I am wrestling with one such issue and when its resolved (one way or the other) Ill follow up on this thread.. but for now, I just want folks to know that we don’t fix things because we don’t want to – we want to fix most every bug we run into, but everything has a tradeoff.
After a nice meal and relaxing by a fire - I thought I would go back and revisit some blog entries.
This bug was not fixed - the TAM and the customer initially pushed very hard for a fix, but I showed them how it was not really relevant. Maybe another time I will go into the details of the bug.. it had to do with account lockouts for local Win2k accounts.