#1518: Parameter parser and CCTK_ParameterSet interpret leading zeros in numbers differently ---------------------+------------------------------------------------------ Reporter: rhaas | Owner: sbrandt Type: defect | Status: review Priority: minor | Milestone: Component: Cactus | Version: development version Resolution: | Keywords: ---------------------+------------------------------------------------------
Comment (by hinder):
No, I don't. The ET tests are not exhaustive, and while they are useful for detecting some problems, they should not be used as evidence that something should not be supported. Cactus is a framework; the Einstein Toolkit is just one user of that framework. Many people use the Cactus flesh with thorns which are not in the Einstein Toolkit. Just because something is not used in the toolkit, it doesn't mean it can be changed or removed. If we want to remove the Util_Expression API, we should discuss this first, weighing up the pros and cons, evaluating how likely it is that people will be affected, announcing the change well in advance (e.g. at least one release), and getting a consensus from the Cactus developers that this is a good idea. It's possible that nobody uses it; it is, after all, a rather obscure feature. But the Cactus flesh has a large number of users, and a decision like this should not be taken lightly. Would you like to propose that this flesh API is changed, and give arguments for why this is good and/or OK? My own preference would be to implement the old functionality (down to the return types and semantics) in terms of the new, or to provide a new API in addition to the old one, deprecating the old one, if wrapping the new API with the old is not considered feasible.