-
Notifications
You must be signed in to change notification settings - Fork 710
Missing files in package #3093
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
It's possible that Cabal is using a different database. Is this in a sandbox? Can you post the output with |
I was not using a sandbox. After nuking .cabal and .ghc for the second time, I managed to install cpphs (and all my other packages) without any errors. I think what I did differently that time was that I installed cpphs before any other packages. I'll see if I can find a way to reproduce this - it shouldn't be too hard, as it is not the first time that these sorts of errors have popped up. Incidentally I do have the -v output:
Thanks for the help! |
Could also be a bug in |
We ran into this with the hackage builders too for some reason. @davean ended up just kicking the box over and then it got working again, but we didn't track down the exact source of issues... |
Then it stopped again. If you have a proper solution I'm all ears. On Mon, Feb 1, 2016 at 5:33 PM, gbaz [email protected] wrote:
|
Happened again :(
Again, nothing detected by ghc-pkg and I can import Data.DList just fine in ghci. |
hey @vladfi1 - are you able to tell me a bit more about how you reproduced this error? |
My current hypothesis is that this is due to missing dynamic libs. |
@silky Unfortunately I wasn't able to reproduce the error. |
I have reason to believe, but not absolute confirmation, that this is related to cabal sandboxes being found up the directory tree. At least for the builders. That package database is used, but its paths not translated, so it looks for packages, that it isn't supposed to have, in the wrong place. I need to build a lot more packages to confirm this has solved the problem for me though. |
I have doubts that this is the case. Unless you have a |
My reasons for believing this is that:
There are of course reasons the hashes might match but not be coming from Since removing the sandbox I've been able to build over 200 packages -davean On Thu, Feb 11, 2016 at 9:33 PM, Mikhail Glushenkov <
|
Would be interested to hear about how exactly you were using sandboxing. |
I used a cabal sandbox to build https://github.com/haskell/hackage-server From the root directory of said git repo, I ran the builder script. The "--package-db=clear --package-db=global --package-db=" ++ packageDb (See -davean On Fri, Feb 12, 2016 at 3:56 AM, Mikhail Glushenkov <
|
Oh, and GHC is installed via hvr's ubuntu packages. There is a slightly On Fri, Feb 12, 2016 at 4:07 AM, davean [email protected] wrote:
|
So if you're running |
no sandbox, using arch, exact same issue still :/ |
@xerus2000 could you post the sequence of commands that leads to that error and the error itself? |
ok this is going to be the usual haskell on archlinux issue where their default install is busted since it provides only dynamic builds of the libs. Here's their documentation, but I wouldn't go the route they suggest: https://wiki.archlinux.org/index.php/haskell Instead, just rip out everything arch provides for cabal and ghc and install good, reasonable versions via ghcup: https://www.haskell.org/ghcup/ |
By the way I think we can close the ticket since it was originally for v1- sandboxes |
yes close it, I indeed had the usual arch issue :/ |
I've started getting some strange errors on a new machine running ghc 7.10.3 and cabal head. When I try to cabal install cpphs (or anything depending on it), I get:
The odd thing is that ghc-pkg check doesn't show any errors, and I can import both Text.Parse and System.Time just fine in ghci.
The text was updated successfully, but these errors were encountered: