> > I understand that you want to separate the steps and build a consensus, > i think we are spending a lot of time on the raw part, we should decide a > syntax and move to the real question, do we offer support for alignment > directly in the language or not ?
I prefer to move directly on to the real question, and then discuss syntax :) Among others reasons, picking a syntax often implicitly constrains the solution, before you know what the right questions are. Of course, there is the ongoing challenge that, when discussing questions and possible answers, one has to use _something_ to illustrate what you mean, and of course then people will want to discuss the syntax. And we’re going to resist that urge, even though this feature has a higher syntax quotient than most. I agree that raw-ness was a distraction in the first round, which is why I placed it at (2) in my list of steps. And right now, we’re still at 1 / 1a. In the first round, we got wrapped around the axle with raw-ness so early, we didn’t even stop to think about the bigger problem, multi-line. (It is tempting, as I mentioned before, to consider multi-line to be “just” a special case of raw, and while that’s a possible outcome, there are some really good reasons to consider it on its own first.). Jim is working on some organized thoughts for 1a, stay tuned.
