Did you know that MathWorks publishes information about product bugs on its web site? And that you can be automatically notified with updates regarding specific bugs, or regarding specific products?

A few weeks ago, Joan posted a blog comment about strange behavior in the deconvwnr function. I investigated and determined that there was a bug in some code that was intended to avoid divide-by-zero warnings. I fixed the problem and then published a bug report, including a work-around, to our support site. (The fix was too late to be included in the upcoming R2008b release, so it will be in R2009a.)

The bug report includes a summary of the problem, which product releases have the problem, whether the problem has been fixed in a shipped release, and whether there is a work-around. In the case of the deconvwnr problem, there is a work-around, and the report includes instructions for downloading new versions of two M-files.

To search the online bug reports, start at the MathWorks home page, then click on Support, and then click on View and Report Bugs. From this page you can search the bug reports, look at bug reports for specific products, set up a watch list, and more. You can sign up to receive e-mail whenever a specific bug report is updated, or you can subscribe to RSS feed, customized to the list of products you are interested in.

There is also a "Frequently Asked Questions" page about bug reports that explains, for example, what an "OPEN" status means, or why some bugs aren't shown in the bug reports.

We are continually trying to improve the quality of our products and therefore reduce the need for published bug reports. In my area, there are fewer unfixed bugs in the development builds of the Image Processing Toolbox and the Mapping Toolbox than at any previous time in the history of these products. But any number of unfixed bugs is too many, and we're trying to do better for you.

Note: Most of the time, you should report product bugs directly to MathWorks technical support rather trying to post them as comments on this or other MATLAB Central blogs. My usual practice is to delete any comments that aren't directly relevant to the topic of the blog post. Joan's comment was relevant to the topic, so I let it through.