> Once the code will work (or "mostly" work), and once I will be able to > actually understand what I am trying to implement, I am going to remove > this layer. Because yes, I think most probably you are right, the reality > is simpler. (however: _perhaps_ jctl stops can add some complications, > not sure).
Sure, that is fine. I didn't have a fully correct implementation before, so I don't claim to be sure about what holes might be in what I've considered. I'm just sure we can find and consider all the underlying complexities together so we become sure what is really required. > Perhaps I am wrong, but I think it is much simpler to simplify the working > code, compared to improving/complicating the code which doesn't work. Well, sure, except perhaps when the "easy to understand" path to "working" gets hairier and hairier with new caveats introduced by the abstraction. > So, in short: ->ev_array was added just to make some progress. Not sure > this was the wise choice though. Understood. My chief metric remains "overall progress" and so whatever works best for you is the way to go about it. The flurry of hairy patches just got me suspicious of whether this path was really the efficient one for you, but it's up to you. Thanks, Roland
