Downloading and Installing Orca

Orca Already Comes with the Operating System!

As of GNOME 2.16, Orca is a part of the GNOME platform. As a result, Orca is already provided by default on a number of operating system distributions, including Open Solaris, Fedora, and Ubuntu.

Specific instructions for downloading and installing operating systems that have been used by Orca users can be found at the following links:

• Downloading/Installing OpenSolaris

• Downloading/Installing Fedora Core 5

• Downloading/Installing Fedora Core 9

• Downloading/Installing Ubuntu Dapper

• Downloading/Installing Ubuntu Edgy

• Downloading/Installing Ubuntu Feisty

• Downloading/Installing Ubuntu Gutsy

• Downloading/Installing Ubuntu Hardy

• Downloading/Installing Ubuntu Jaunty

• Downloading/Installing Ubuntu Karmic

• Downloading/Installing GRML

Installing from Source

Orca is under continual development, so people may want to experiment with the latest sources. For example, you might want to experiment with the latest support for Firefox 3.0.

The latest stable sources can always be found at the GNOME ftp site. The brave can always obtain the latest 'trunk' sources from the 'orca' module in the GNOME (git) source code repository.

WARNING WARNING WARNING: If you work from the bleeding edge, you will have access to all the latest and greatest stuff. Please keep in mind what you are doing, though. The bleeding edge is named appropriately.

Once you have obtained the Orca sources you wish to work with, you can build the sources and install them on your machine.

Setting Up Your Build Environment

The setup for the build environment differs between operating system distributions. Once you have the build environment set up, however, the procedure for building from source is the same.

OpenSolaris Build Environment

The build environment for OpenSolaris is known as the "Common Build Environment" (CBE). You need to obtain the CBE from the JDS Download Center and unpack it. In a terminal window, you then need to cd to the resulting directory and run the cbe_install script as the root user. As you run the cbe-install script, it may ask you a number of questions similar to the following:

Checking for required packages...
Package SUNWi2cs is required for building JDS
Would you like to install it now? [yes]: 

It is safe (and recommended by us) to answer 'no' to all these questions. There are a little over 30 of them. You may also be prompted with the following:

Installing the JDS Common Build Environment (CBE) version 1.5.1
The recommended C compiler(s) for this version of the JDS CBE is/are:
        Sun Studio 11 (Venus) (cc: Sun C 5.8 Patch 121016-02 2006/03/31)
        Sun Studio 10 (Vulcan) (cc: Sun C 5.7 Patch 117837-04 2005/05/11)
Enter the path to the C compiler [/usr/bin/cc]: 

Selecting the default of /usr/bin/cc is fine, and you should answer 'yes' to the following question:

/usr/bin/cc is not the recommended compiler
"cc -V" should print one of the following:
       cc: Sun C 5.8 Patch 121016-02 2006/03/31
       cc: Sun C 5.7 Patch 117837-04 2005/05/11
Would you like to use this compiler anyway? [no]: 

Finally, you should install the optional packages by answering 'yes' to the following question:

This version of the CBE comes with some optional utilities.
These are not required for building the JDS code, nevertheless
they are useful to have on your system.
Please see the README file for the list of optional utilities.
Would you like to install the optional packages? [yes]: 

The install will chug along for a little while and will eventually complete successfully. At this point CBE will be installed.

Once you've installed CBE, you need to make sure the development tools are in your path. The best way to do this to make sure you are in a terminal window with "bash" as your shell. Then, execute the following command. You need to do this each time you want build anything using the tools provided by CBE (punctuation matters with this command and you should type it exactly as written):

. /opt/jdsbld/bin/env.sh

You may find it useful to embed this in your bash profile. For example, you can add the following to ~/.bash_profile:

if [ -f ~/.bashrc ]; then
   source ~/.bashrc
fi

And then the following to your ~/.bashrc:

source /opt/jdsbld/bin/env.sh
export PATH=~/Desktop/firefox:/usr/local/bin:/usr/openwin/bin:/usr/openwin/demo:/usr/demo/jds/bin:$PATH

Ubuntu Build Environment

Execute the following command to set up the build environment for Orca on Ubuntu. Punctuation matters with these commands and you should type them exactly as written.

sudo apt-get build-dep gnome-orca
sudo apt-get install git-core gnome-common automake1.9

Fedora Core 6 or 7 Build Environment

(Contributed from community member Scott Berry -- thanks Scott!!!).

NOTE: it is highly recommended that you never use root as your regular username. Instead, use your username and su in to your root account or you could inadvertently do some serious damage.

To install from git on Fedora Core 6 or 7 you must have all your ducks in a row. The first thing you want to do is ensure you have all packages. From experience so far this just requires gnome-common and intltool. Depending on which package manager you prefer (apt or yum) you may do the following:

Execute the following commands for apt (do not include the quotes when doing this):

apt-get install git gnome-common intltool

Apt will read the packages and if it finds the packages it will probably ask you if you would like to install the package. Just hit y for yes or n for no. This will download the packages and then commit the changes.

Yum is very close to the same syntax, only you use the following command (again do not use the quotes):

yum install git gnome-common intltool

This will in turn try to find the packages you are looking for. If found successfully then you will want to say y again for yes and allow it to download and commit the changes.

OpenSUSE 10.3 Build Environment

When OpenSUSE 10.3 has been installed from DVD, or you have elected to use the online repositories while installing from the downloaded CD, almost all the software you will need for installing Orca from source will already be available to you. The only package left to install is python-devel. Use the following command at the terminal while logged in as root.

yast -i python-devel

This will launch the text version of yast, download the required package and install it as well as any other missing dependencies. If all goes well, you will be taken back to the command prompt.

Installing the Latest AT-SPI Infrastructure (atk and at-spi)

AT-SPI stands for assistive technology service provider interface, which is the accessibility infrastructure for GNOME. Like Orca, the AT-SPI is frequently updated to include performance improvements such as Collections. To benefit from this work (and often for Orca to even work), you need to obtain/build/install the latest AT-SPI sources from the AT-SPI git trunk. In addition, atk, , which is another portion of the infrastructure, contains important bug fixes; it's necessary to obtain/build/install them, too. The general order in which things need to be built is as follows: atk, then at-spi, followed by orca.

The following commands provide the sequence of steps you need to follow to download AT-SPI sources from the git repository, build them, and then install them. Punctuation matters with these commands and you should type them exactly as written, with the exception of <<directoryholdingatspiregistry>>. Instead of <<directoryholdingatspiregistry>>, you need to specify the directory where the at-spi-registryd command is installed on your operating system:

• Open Solaris: <<directoryholdingatspiregistry>>=/usr/lib

• Ubuntu: <<directoryholdingatspiregistry>>=/usr/lib/at-spi

• Slackware: <<directoryholdingatspiregistry>>=/usr/libexec

NOTE: Orca is officially only supported with the version of GNOME it has been developed for. For example, Orca v2.22.x is supported only on GNOME 2.22. The team regularly builds Orca on previous versions of GNOME, however. For example, while developing Orca v2.22.0, the Orca team often used the GNOME 2.20.x desktop. Doing the following generally works, but you must do so at your own risk. We do not take responsibility for anyone trashing their setup. To determine which release of gnome you are running, run gnome-about --verssion.

# The following is for Ubuntu users:
sudo apt-get build-dep libatk1.0-0 at-spi
sudo apt-get install libgconf2-dev

# Get, build, and install atk
git clone git://git.gnome.org/atk atk
cd atk
# optionally checkout the stable branch for your system.
# git checkout -b origin/gnome-2-20
./autogen.sh --prefix=/usr
make
sudo make install
cd ..

# Get, build, and install at-spi.  Replace the
# <<directoryholdingatspiregistry>> portion following
# the directions above.
git clone git://git.gnome.org/at-spi at-spi
cd at-spi
# optionally checkout the stable branch for your system.
# git checkout -b origin/gnome-2-20
./autogen.sh --prefix=/usr --libexecdir=<<directoryholdingatspiregistry>>
make
sudo make install
cd ..

Follow this audio walk through for step by step instructions for this process.

Obtaining the Orca Sources, Building and Installing

WARNING WARNING WARNING: If you work from the bleeding edge, you will have access to all the latest and greatest stuff. Please keep in mind what you are doing, though. The bleeding edge is named appropriately.

The latest development sources can be obtained from the 'orca' module in the GNOME git source code repository. You can download the sources using the git clone command as shown below. Once you have downloaded the sources from the git repository, the Orca sources will be in a new directory named orca. If you want to update your sources from the latest git repository, you can cd to the orca directory and enter git pull. This will pull the latest sources from the git repository.

After you've downloaded the sources from the GNOME git (git) repository, you need to cd to the resulting directory and execute the commands needed to build and install Orca (NOTE: on Solaris, sudo may not exist, so you'll need to do an su - root before executing the make install command).

The following commands provide the sequence of steps you need to follow to download Orca sources from the git repository, build them, and then install them. Punctuation matters with these commands and you should type them exactly as written.

git clone git://git.gnome.org/orca orca
cd orca
# optionally checkout the stable branch for your system.
# git checkout -b origin/gnome-2-20
./autogen.sh
make
sudo make install

Let's break this down a bit so you know what you are doing here.

1. git is a repository of bleeding edge sources for Orca so you will only want to get these if you are not feint of heart like us computer junkies. Grin!

2. git clone git://git.gnome.org/orca orca allows you to get the bleeding edge. git is set up in directories and there are many on GNOME. Because GNOME has many packages that work with it. Pay attention to what you type or you could inadvertently get the wrong git repository! Once you have typed this command, all the sources are downloaded to the orca directory.

3. Note: if git fails to pull from the main repository, you can always attempt to use the mirror. Instead of git clone git://git.gnome.org/orca orca use git clone  git://git-mirror.gnome.org/git/orca orca to pull orca sources from the mirror repository.

4. To get into the orca directory, you type cd orca. If you ever want to update the sources from git, you can cd to this directory in the future and type git pull. Any new files and updated files will be downloaded to your machine.

5. NOTE: if you want the latest stable sources, you can always grab them from a stable branch. For example, the Orca team continually checks in stable code to the gnome-2-20 branch for GNOME 2.20.x releases and the gnome-2-22 branch for GNOME 2.22.x releases. To get the gnome-2-20 branch, cd into the orca directory and use the following command to checkout the 2-20 branch git checkout -b origin/gnome-2-20 after you clone the main orca entry. Once in the orca directory git branch -r will display all available branches for checkout. Notice that once you checkout a sepearate branch, you must update with git pull . instead of just git pull (notice the addition of a decimal point).

6. ./autogen.sh enables all the compilers and decides how the program will be assembled in order for you to use it.

7. make actually goes in and out of directories getting everything set up to install for us.

8. If the make succeeded, sudo make install installs Orca for you under /usr/local, and the Orca executable will be available under /usr/local/bin/orca.

At this point, Orca should be installed on your machine and you are now ready to move onto the next step: configuration and use. Note that on some systems (e.g., Fedora) it doesn't change the desktop shortcut for Orca so you need to log in to GNOME and then hit Alt+F2 to get to the run dialog box. Then, type orca and press Return.

NOTE: The above commands install Orca under /usr/local/bin/orca. As long as your PATH environment variable includes /usr/local/bin ahead of /usr/bin, you should pick up the Orca you just built and installed. If your PATH environment variable doesn't include /usr/local/bin ahead of /usr/bin, you may end up using an old Orca. You can tell which version of Orca you are using by typing orca --version. You can also tell which version of the Orca sources you pulled from the git repository by typing git status | head -1 when you are in the orca source directory that you pulled from git using the command above.

Speech, Braille, and Magnification

Orca does not ship with a speech synthesizer, braille drivers, or a magnifier. Instead, Orca relies upon system services to be installed on the machine. On Open Solaris and Linux, these services typically come for free (just like Orca). Please refer to the following pages for more information:

• Speech

• Braille

• Magnification