Ihor Radchenko <yanta...@posteo.net> writes: > The proper test is not presence of the message, but the correct > conversion of the legacy options in the actual html output.
We do test the results in the HTML output, of course. Still, I would like to test the user messages too. By my definition, "proper" tests verify all requirements. Hence, I ask: Do we want to reliably notify the user? If we do not, we can remove the messages. If we do, then we either test the messages or they will break at some random point in time and generate yet another pointless mail thread. If we do not have any better idea, then I can define a variable that points to the 'message' function by default. The relevant tests can then set the variable to a function that gathers messages in a list. With that, the tests can verify that the list contains exactly one instance of the expected message. Rudy -- "It is no paradox to say that in our most theoretical moods we may be nearest to our most practical applications." -- Alfred North Whitehead, 1861-1947 Rudolf Adamkovič <salu...@me.com> [he/him] Studenohorská 25 84103 Bratislava Slovakia