2015-08-29 Set my masqmail package in Debian to RFA (request for adoption): https://bugs.debian.org/797301 Is goto considered harmful? In my eyes, goto is a language tool as any other: It has power and problems, and mastering a language means learning in which situation which tool is appropriate and what to care for when using a specific tool. I do use goto in certain situations because in these I think it makes the code simpler to understand than not using it. This following change shows a blind-avoidance approach to goto: -goto finish; +break; /* to finish */ The code is worse afterwards. At the small level, the former statement was clearer and more definitive. A keyword and an iden- tifier are treated agains a keyword and a comment. That's a bad deal. If what you want the code to do is ``goto finish'', then write just that. Encoding the clear message into the generic ``break'' with the help of the comment and the surrounding con- trol structures is information blur. Nothing is more helpful for understanding code than obvious statements like ``goto finish'' or ``if (have_items)'', for instance. (They require good identif- iers, of course. Second most useful are idioms, but they require knowledge of the particular language, whereas clear statements do not.) At the larger level (not visible here), this change made finding the `break' target become distracting when reading the code. Dijkstra was all right when he went to fight the goto disease, but that was in times when they were programming Basic and For- tran and Cobol. These languages had a lack of structural program- ming language features and also the mind view of programmers was more goto than structural. The problem is not having a goto statement in the language and using it wisely from time to time, but it is having nothing but goto or using it blindly *all* of the time. Goto is a truly sharp knive ... sometimes, that's just what does the job best. http://marmaro.de/lue/ markus schnalke