I received my new Dell with Ubuntu yesterday. In addition to wanting this machine for future Haiku testing, I wanted to build Haiku on Linux. With help from an article on our web-site (http://haiku-os.org/documents/dev/getting_linux_developer_tools) I got things working pretty fast. I was also extremely pleased with how fast Haiku compiled on this new machine. Once I had the cross-tools built, I built the full haiku.image in about 13 minutes. I was quite shocked actually since building takes much, much longer on my old Celeron 1.33 GHz which I run BeOS R5 on. Though the aforementioned article is pretty good, I wanted to document the step-by-step instructions on how to get Haiku compiling on a fresh Ubuntu 7.04 (Feisty Fawn) install. 1. Install the needed packages for building: sudo apt-get install subversion autoconf automake texinfo flex bison gawk build-essential 2. Create the Haiku development directories and get the Haiku build tools source: mkdir develop cd develop mkdir haiku cd haiku svn checkout svn://svn.berlios.de/haiku/buildtools/trunk buildtools 3. Build and install the Haiku Jam: cd buildtools/jam make sudo ./jam0 install cd ../.. 4. Get the Haiku source (this will take a while): For anonymous check-out: svn checkout svn://svn.berlios.de/haiku/haiku/trunk haiku For an existing Haiku developer: svn checkout svn+ssh://developername@xxxxxxxxxxxxxx/svnroot/repos/haiku/trunk haiku If there are any errors when checking out above, just cd in the haiku directory and type "svn up" to get any missed files. This applies when updating the code in the future too of course. 5. Build the cross compiler tools: cd ~/develop/haiku/haiku ./configure --build-cross-tools ../buildtools/ 6. Build the Haiku image: jam haiku-image That is it! Not too bad. Thanks to Ingo and the various other contributors for making the Linux compiling so easy. Now to test an image, there seem to be at least two good options in Linux: 1. qemu. To install in Ubuntu: sudo apt-get install qemu Qemu is also needed when converting the default Haiku image to a VMWare image. To run the Haiku image in qemu, first start in the haiku source directory, then: cd generated qemu -hda haiku.image -serial stdio I found Qemu runs quite a bit faster on this machine than my BeOS box. 2. VMWare Player In Ubuntu this can be installed with Add/Remove Programs or by using the command: sudo apt-get install vmware-player This also requires accepting a license agreement and other installation steps. I received some errors about existing files for the network, but it didn't seem to be a problem in the end. To use qemu-img to convert the image for use in VMWare: cd ~/develop/haiku/haiku/generated/ qemu-img convert -f raw haiku.image -O vmdk haiku.vmdk You will also need this file to run the VM (this should be saved in and run from the same directory as the vmdk file is in): http://haikuhost.com/haiku.vmx There is a third option which I hope we can get working one day. Right now running Haiku in it causes a kernel panic: 3. VirtualBox. This is the new open source virtual machine on the block. Go here to download the right version for your system: http://www.virtualbox.org/wiki/Downloads If you install the Debian package in Ubuntu with the graphical installer, be sure to open the little terminal window in the GUI as you have to accept a license agreement and do other installation steps. You also need to make your Linux user account a member of the vboxusers group. To convert the default Haiku image into something that can be used in VirtualBox, you need to download this file: http://www.virtualbox.org/download/testcase/vditool I put mine in ~/bin. Then make it executable: cd ~/bin chmod +x vditool To convert the Haiku image: cd ~/develop/haiku/haiku/generated/ ~/bin/vditool DD haiku.vdi haiku.image You can then use the VirtualBox GUI to do the rest. It is pretty easy to use. Of course as I said right now Haiku has a kernel panic when being run in this VM. I don't quite know where to start in debugging this, but this might be a good tool to get Haiku working with in the future. Hope all this helps! Regards, Ryan