[racket] Code compiled for version ~a, not ~a

From: Jukka Tuominen (jukka.tuominen at finndesign.fi)
Date: Sun Sep 18 02:20:52 EDT 2011

I think I have had similar errors, but these were outside drracket. I can't recall exactly, but I think they were either commandline raco compilations or commanline gracket executions. I had to remove all compiled directoried manually an rerun the command to bypass the situation. That I recall clearly :)

So, either automatic recompilation, being able to force recompilaton, or interactive question which way to go, would have been nicer. The last option shouldn't be the only one, since human presence cannot be guaranteed.

br, jukka
  -----Original Message-----
  From: users-bounces at racket-lang.org [mailto:users-bounces at racket-lang.org]On Behalf Of Robby Findler
  Sent: 18 September 2011 04:40
  To: Guillaume Marceau
  Cc: users at racket-lang.org
  Subject: Re: [racket] Code compiled for version ~a, not ~a


  There could definitely be a bug where something is not comparing paths properly somehow. If you can give me more help to reproduce the bug that would be a big help. 

  To start I think just sending me the logging tool directory (and saying where it was (collects, link?)) would probably help me make progress with the specific thing below. 

  Robby

  On Saturday, September 17, 2011, Guillaume Marceau <gmarceau at gmail.com> wrote:
  >> Are these files in the collects tree or elsewhere?
  >
  > I have separate collects trees for each version.
  >
  > Occasionally, I do get the error in the collect tree for my git repo.
  >
  > Earlier today I did,
  >
  >  git pull
  >  <compile src/workspc10>
  >  ./raco setup -D
  >  ./DrRacket
  >
  > and received the version error for
  > collects/planet/private/compiled/resolver.zo, which suggests there
  > might be a bug in raco.
  >
  >
  >> drracket should be taking care of this for you (unless you turn off the
  >> automatic compilation option).
  >
  > Sounds like the behavior I see is a bug then.
  >
  > Could it be the code that checks whether a directory is a subdirectory
  > of another breaks under Windows?
  >
  > There is a different bug under Window I meant to report that might be
  > related. If I set PLTCOLLECTS with forward slashes, a directory
  > inclusion test somewhere returns #f instead of #t and I get the error:
  >
  >    $ echo $PLTCOLLECTS
  >    c:\documents\collects;
  >    $ export PLTCOLLECTS='c:/documents/collects;'
  >    $ /c/Program\ Files/Racket-5.1.3/raco setup -l logging-tool
  >    omitted-paths: given directory path is not in any collection root:
  > #<path:c:\documents\collects\= Archive>
  >
  > If I use backslashes, the error goes away. DrRacket is not affected.
  > It boots correctly with either forward or backslashes.
  > 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20110918/91acfa39/attachment.html>

Posted on the users mailing list.