Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Already on GitHub? Sign in to your account

depsolver error messages have become incomprehensible #1005

Closed
bos opened this Issue Aug 15, 2012 · 4 comments

Comments

Projects
None yet
4 participants
Contributor

bos commented Aug 15, 2012

Trying to build criterion HEAD using GHC 7.6.1 RC this morning, I am confronted by this output from cabal-install's depsolver.

I have absolutely no idea what it means, how to identify what's broken, or how to fix it. This is really bad.

cabal: Could not resolve dependencies:
next goal: criterion (user goal)
rejecting: criterion-0.6.0.1 (global constraint requires ==0.6.0.0)
trying: criterion-0.6.0.0
trying: parsec-3.1.3 (dependency of criterion-0.6.0.0)
next goal: text (dependency of parsec-3.1.3)
rejecting: text-0.11.2.3/installed-157... (conflict: vector-algorithms =>
bytestring==0.9.2.1/installed-375..., text =>
bytestring==0.10.0.0/installed-35d...)
trying: text-0.11.2.2
rejecting: text-0.11.2.2:-integer-simple (conflict: base =>
integer-gmp==0.5.0.0/installed-56b..., text-0.11.2.2:integer-simple =>
integer-gmp>=0.2 && <0.5)
trying: text-0.11.2.2:+integer-simple
next goal: integer-simple (dependency of text-0.11.2.2:+integer-simple)
Contributor

bos commented Aug 15, 2012

What I'm doing to reproduce this:

git clone git://github.com/bos/criterion.git
cd criterion
cabal install -w /usr/local/ghc/7.6.1/bin/ghc
Contributor

bos commented Aug 15, 2012

I eventually figured out that vector-algorithms is the culprit. But wow, that was painful!

Contributor

kosmikus commented Aug 16, 2012

Hi Bryan. Sorry to hear you dislike the error messages of the new solver. Improving the error messages was one of the main reasons for the rewrite. If you think they've become worse, then we should do something to make the messages more understandable for you. Independently, your log looks like there might be an actual bug in the message generation though. I'll try to look into it.

@ghost ghost assigned kosmikus Aug 16, 2012

jsl commented Feb 28, 2015

I imagine the error messages have improved considerably since this was reported. I'm also not sure exactly what HEAD of the criterion library was when this was reported so this is tricky to reproduce. I propose closing, and opening other tickets if there are specific issues with the solver.

@ttuegel ttuegel closed this Feb 28, 2015

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment