[Hackage] #746: Cabal should complain if you give an upperbound on Cabal-Version

Hackage cvs-ghc at haskell.org
Tue Oct 12 10:19:10 EDT 2010


#746: Cabal should complain if you give an upperbound on Cabal-Version
----------------------------+-----------------------------------------------
  Reporter:  igloo          |        Owner:      
      Type:  defect         |       Status:  new 
  Priority:  normal         |    Milestone:      
 Component:  Cabal library  |      Version:  HEAD
  Severity:  normal         |     Keywords:      
Difficulty:  unknown        |   Ghcversion:      
  Platform:                 |  
----------------------------+-----------------------------------------------

Comment(by duncan):

 I am thinking of ignoring upper bounds entirely. Also, if the package says
 that it uses a more recent Cabal spec version then to reject version
 ranges not of the form ">= x.y".

 That is, this would be allowed:
 {{{
 cabal-version: >= 1.6 && < 1.9
 }}}
 We would just ignore the upper bound.

 But this would be rejected:
 {{{
 cabal-version: >= 1.10 && < 1.11
 }}}

 The rationale is that the first one is saying that it's following Cabal
 spec version 1.6 which did allow arbitrary version ranges. As of Cabal
 spec version 1.10 we do not allow arbitrary version ranges.

-- 
Ticket URL: <http://hackage.haskell.org/trac/hackage/ticket/746#comment:1>
Hackage <http://haskell.org/cabal/>
Hackage: Cabal and related projects



More information about the cabal-devel mailing list