Conversation

Varun
@typesanitizer
Drinking game where you take a shot every time someone says "you don't really need a debugger because <approach that I use which requires compiling code one or more times> is almost always better" without admitting that their language's debugger is awful/non-existent.
11
138
Jon Purdy
@evincarofautumn
I agree Haskell needs a good debugger, so I usually don’t mention that I haven’t needed a debugger in Haskell—at least not like I did in C. The ghci debugger and call stack support are implementations of non-native *debug models*; I want to see a first-principles native one.
1
3
Jon Purdy
@evincarofautumn
A while back I was playing with implementations of lazy evaluation, and the process of debugging lazy runtimes led me to think about debugging lazy programs, and what a debug model even is. I think it’s a *space*, in which a debugger is a vehicle for navigating a computation.
1
3
Brendan Zabarauskas
@brendanzab
Replying to
@brendanzab
@evincarofautumn
and
@typesanitizer
Oh, I just remembered that there is something similar available for Haskell: twitter.com/TaReven/status – not sure how complete it is compared to what you can find in Mercury though.
Quote Tweet
Taren
@TaReven
·
Replying to @brendanzab @csaba_hruska and @MercuryLang
Haskell has Hoed which hasn't bitrotted yet. It has TH support and it has declarative debugging but not data provenance hackage.haskell.org/package/Hoed-0
Twitter Web App
1
Like