[racket] Windows build problems on 5.90.0.7

From: Ben Goetter (goetter at mazama.net)
Date: Sat Aug 17 15:39:23 EDT 2013

Thank you.  That's exactly what happened: I missed the submodule nature 
of native-pkgs.

These directions are already in INSTALL.txt, which I did not reread.  I 
didn't look beyond 
http://git.racket-lang.org/intro.html#clone_the_plt_repository.

On 8/17/2013 11:52 AM, Nick Shelley wrote:
> I just ran into this very thing today because I didn't realize the 
> packages were put into a submodule and I'm in the habit of not 
> reading error messages. However, after trying some things and getting 
> this error multiple times, I finally read it carefully. The answer is 
> in your skimmed "something something" text. Just go to the root of the 
> repo and do a 'git submodule init' then 'git submodule update'.
>
> I think it would be helpful if these instructions were added to the 
> install readme (unless it is there and I missed it).
>
> On Saturday, August 17, 2013, Ben Goetter wrote:
>
>     I apologize for the semi-clueless nature of this message.  I know
>     that Racket's been going through a lot of changes in the last few
>     months.  Rip van Winkle here is trying to come back up to speed.
>
>     I did a fresh git clone, then built Racket's core on Windows with
>     VS2008 via the script in racket/racket/src/worksp/build.bat.
>     Apparently this built only the core (thank you, Vince on irc).  I
>     got a working racket.exe with some intrinsics included, but no
>     drracket or the standard libraries (gracket -z complained that it
>     could not find racket/racket/collects/racket/gui/init.rkt).
>
>     At Vince's advice, I tried
>         raco pkg install -i main-distribution
>     to complete my installation, but got a SSL error
>
>         Resolving "main-distribution" via https://pkg.racket-lang.org
>         SSLv23_client_method: implementation not found; no arguments
>     provided
>         ...collects/openssl/mzssl.rkt:397:18
>
>     So then I did something radical: I read the top-level README.txt
>     and followed its directions.
>         nmake win32-in-place
>
>     This rebuilt the core, created a spurious "-p" subdirectory, and
>     finally triggered the complain-no-submodule target:
>         "Native packages are not in the expected subdirectory"
>     something something "git submodule" something
>
>     Generally, what's the new proper practice for building my own
>     Racket on Windows?  Build.bat no longer suffices by itself.
>
>     Thanks,
>     Ben
>     ____________________
>      Racket Users list:
>     http://lists.racket-lang.org/users
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20130817/e6e92485/attachment-0001.html>

Posted on the users mailing list.