Move /=== to a Dojo-only affordance.

[sigilante]

No description provided.

[tinnus-napbus]

The problem with these is arguably the inverse: that we've lead people to expect our/desk/now due to how it's implemented in the Dojo. As I understand it the basic purpose is to specify the filepath to include in the stacktrace during a compilation failure, and I guess it's exposed to the thing being compiled via % & /=== so that it can either include it in its own error messages or pass it through to things that itself compiles

[sigilante]

This is an interesting point—if /=== usage is the result of scope creep, then we need to decide a good partition if we want to "fix" the situation. Certainly /=== is convenient, and I can see how it's useful in stack traces. But /app usage is the real problem, and the /(scot %p our.bowl) etc. pattern is really pretty wretched.

[midden-fabler]

This is an interesting point—if /=== usage is the result of scope creep, then we need to decide a good partition if we want to "fix" the situation. Certainly /=== is convenient, and I can see how it's useful in stack traces. But /app usage is the real problem, and the /(scot %p our.bowl) etc. pattern is really pretty wretched.

(en-beam [our %base da+now] /foo) is nice, and underused.

[sigilante]

I've been thinking about this, and I think that the better course is to stop teaching /=== then, and instead teach en-beam.