Howto set up a chroot build environment for deployment?

Per Arnold Blåsmo pablaasmo at atmel.com
Wed Oct 21 13:09:41 UTC 2009


Hi, and thanks for your answer.


On 21. okt. 2009 18:54:44, Erik van Pienbroek <erik at vanpienbroek.nl> wrote:
> Hi Per Arnold,
>
> For my own open source projects I've created some small scripts to
> generate Win32 installers. These scripts build the project in a separate
> prefix and uses a (custom-made) .nsi file and mingw32-nsis to generate
> an installer. When the installer is created, the scripts transfer the
> files using scp to my webserver where users can download it. This all is
> hooked up as a post-commit-script in my Subversion server so that a new
> installer will be created and published every time a Subversion commit
> has been done.
>
>   
This is basically also how I have done it today. But I use the x86-cross
build script from mingw to build a mingw cross environment first, and
the build my packages in that environment. Then I pack the files I need
into a zip file that is transfered to a window xp machin where
installshield is used to build an installer.

I have not tried yet to use Fedoras own mingw environment. That is
basically what I want, but in a clean setup each time.
> These scripts are very specific to my open source projects so they
> probably won't be of any interest to you, but it should be quite easy
> to create such a script yourself. It should be sufficient to create a
> script which does a ./configure, make, make install, some cp's
> (optional, for dependencies) and a 'makensis' call to generate an
> installer.
>
>   
I have not tried nsis yet, but would like to try that. I guess I have
some trial and error in front of me :-)

Regards
Per A.





More information about the mingw mailing list