[GHC] #14398: Fail to install haskell platform on Windows

[GHC]

#14398: Fail to install haskell platform on Windows
-------------------------------------+-------------------------------------
        Reporter:  KAAAsS            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  GHCi              |              Version:  8.2.1
      Resolution:                    |             Keywords:
Operating System:  Windows           |         Architecture:  x86_64
 Type of failure:  Installing GHC    |  (amd64)
  failed                             |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #14168            |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by PaladinOfHonour):

 Replying to [comment:8 Phyx-]:
 > Thanks! the trace shows nothing really wrong, I think 8.2 has a memory
 corruption somewhere that's causing invalid characters to manifest in the
 strings. I'll have to figure out a way to track it down to verify if it's
 actually fixed or just happens to work in 8.4.

 This will be a very late reply to this thread, however, I encountered the
 exact same internal error issue. I run Windows 10 64-bit with Japanese as
 the OS language.
 After troubleshooting, and ruling out the similar #15021 chcp ticket,
 upgrading from 8.2 to 8.4 did do the job and resolved the error.
 As I noticed the error hasn't been closed yet and it's exact origin not
 quite yet discovered, I thought I'd give some extra information.
 Certain Japanese fonts map the ¥ sign to the \ , causing paths to take the
 visual form of C:¥Users¥User etc. This is most often merely visual,
 however, can cause problems within certain programs.
 Furthermore, OS specific directories might be in Japanese (user = ユーザー
 ), however, this should once again be merely a visual change.

 As the original poster was running a Chinese system, I suspect the error
 is caused by either of the above, or similar, effects.

 Thanks for the suggestion of upgrading and if there's still an interest in
 the bug I hope a second error report might elucidate the problem a bit.

 Kind Regards,

 Pally

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14398#comment:9>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler