#241 broke synexpand

[gilch]

Synexpand was using the reader's gensym_counter() directly. It probably needs much more thorough tests, being a mini compiler in its own right. I need to think carefully about how to fix this.

I consider synexpand and friends highly experimental and unstable. It might be too complex for the bundled macros, so one possible fix is to remove them altogether. They might end up in a separate repo.

It's been a while, but I think it's only being used for PICK's let'd names, a feature I was considering removing anyway #227.