tag:blogger.com,1999:blog-9138509940389651166.post5751919970180462954..comments2023-07-30T09:24:25.865+01:00Comments on selfelected: Defensive programming and Fail at onceLosManoshttp://www.blogger.com/profile/08166031267431172778noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-9138509940389651166.post-757521148538501242008-07-11T07:19:00.000+01:002008-07-11T07:19:00.000+01:00Completely agree. I checked out the article before...Completely agree. I checked out the article before continuing reading your post and after seeing the example I started thinking if this is really what you expect the program to do "back stage" for the input that is being given. Yeah, maybe "it ain't crashing" but in case the input is given from a machine (calling your example of the tubes) in my opinion, the process must stop and alert the operator that the SENDER is given data that the application is unable to process. If your architecture and logic is based on a certain range of inputs specified before then it shouldn't be certainly your bug, but production must be warned of this, not putting makeup to your code crashers.Anonymoushttps://www.blogger.com/profile/06622183180011708499noreply@blogger.com