gusl: (Default)
[personal profile] gusl
I really really need some better debugging tools for Lisp/Slime.

Instead of having to back up working code all the time, I should be able to change one or two things, and align the two run-time call-trees to see what's different.

I am often asking "What part of my code will react negatively to this change?", and tracing this through testing is very time-consuming. trace gives me way too much to look at. I wonder if diff'ing the two traces would help.

I'd also like a stepper with variable watch. Right now, I'm doing prints, that get commented in and out.

Also, my code needs to be better designed. Implementing what should be a trivial change is proving to be a lot of work. Much of this is legacy code from the 2-month-old Lisper... I was so stupid back then... I feel much more confident as an 8-month old Lisper. Although the design also needed more thought, period... much of the code was evolved "genetically", where my tests were playing the role of natural selection.

---

Also, I'd like to publicly bitch about emacs not letting me paste text from other windows (even if they're emacs windows).

(no subject)

Date: 2005-09-18 05:53 pm (UTC)
From: [identity profile] williamallthing.livejournal.com
this doesn't exactly address the call tree question, but i've been having a great time in ruby developing software + tests thereof at the same time. the support in ruby for doing this is fantastic, and you end up both explicitly thinking about the weird fringe cases and how you want the thing to behave, and developing a large test corpus against which to vet every change. good times.

as far as emacs goes, i find -nw (text-mode only) makes my life much easier. letting the xterm handle cutting and pasting also means it goes to where my text cursor is rather than wherever the fuck my mouse cursor ends up when i click it.

(no subject)

Date: 2005-09-18 06:01 pm (UTC)
From: [identity profile] gustavolacerda.livejournal.com
Does it automatically perform all the tests in the corpus, and tell you which ones failed?

Coming to think of it, this shouldn't be very hard for me to implement... if only I figure out how to make Lisp handle errors.

February 2020

S M T W T F S
      1
2345678
9101112131415
16171819202122
23242526272829

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags