it's really reassuring to see that even the author of CIDER can't keep it in his head what the software can do. I'm about 43 years old and I remember tremendous amount of info, but more and more often I find great code I've written just month ago and I have no recollection of writing it. on the flip side, I also do remember writing certain code already before and I just can't find it... :) I think it's an extremely clear sign that this project needs help, but it seems to require very deep knowledge of both Elisp and Clojure. it was also useful to see how CIDER can cause Emacs to get laggy and understand why it happens and how to recover from it. I think these are the most precious moments of a demo, because it teaches us not just how to use a software but also how to develop it. we should see more failure modes, not less! But, you definitely should NOT feel embarrassed about it! This helps gaining future contributors and maintainers. As it's been pointed out, many ppl are not using the profiler or the tracer, because they are not even aware of those features. those seem to be good candidates for extraction into their own packages. (I can imagine that it might not be very feasible from tech point of view though) having all those evaluation options is great but also overwhelming. I wish it would be organized as a matrix of sexp source (top-level, before point, at point, ...) and result destination (overlay, REPL buffer, comment, inspector, clipboard) and result format (pretty-printed or one liner).
it's really reassuring to see that even the author of CIDER can't keep it in his head what the software can do.
I'm about 43 years old and I remember tremendous amount of info, but more and more often I find great code I've written just month ago and I have no recollection of writing it.
on the flip side, I also do remember writing certain code already before and I just can't find it... :)
I think it's an extremely clear sign that this project needs help, but it seems to require very deep knowledge of both Elisp and Clojure.
it was also useful to see how CIDER can cause Emacs to get laggy and understand why it happens and how to recover from it.
I think these are the most precious moments of a demo, because it teaches us not just how to use a software but also how to develop it.
we should see more failure modes, not less!
But, you definitely should NOT feel embarrassed about it!
This helps gaining future contributors and maintainers.
As it's been pointed out, many ppl are not using the profiler or the tracer, because they are not even aware of those features. those seem to be good candidates for extraction into their own packages. (I can imagine that it might not be very feasible from tech point of view though)
having all those evaluation options is great but also overwhelming.
I wish it would be organized as a matrix of sexp source (top-level, before point, at point, ...) and result destination (overlay, REPL buffer, comment, inspector, clipboard) and result format (pretty-printed or one liner).
Thanks for sharing. And thanks for CIDER!
The real show starts at 30:28