# location to install the JS9 web files: ./configure --with-webdir=[path_to_web_install]
# where to find cfitsio and install binaries, what sort of helper to build: ./configure --with-cfitsio=[path_to_cfitsio] --prefix=[path_to_prog_install] --with-helper=nodejs
The current JS9 source tar file is available at:
https://js9.si.eduThe source tar file will unpack into a js9-[version] directory with the usual tar command, e.g:
tar xfz js9-[version].tar.gzYou also can clone JS9 from GitHub:
git clone https://github.com/ericmandel/js9
file:///Users/me/js9-[version]/js9basics.htmlto see the basic JS9 demo page. You then can drag any FITS image file onto the JS9 display and view it, change contrast/bias, colormaps, scale, create regions, etc. Furthermore, if you load the imexam web page:
file:///Users/me/js9-[version]/js9imexam.htmlyou can move a region around and run browser-based analysis as the region changes.
NB: The file://URI does not work in Google Chrome: Chrome doesn't permit a local HTML file to read other local files. The workaround is to start Chrome with the --allow-file-access-from-files switch:
# Linux: chrome --allow-file-access-from-files # Mac: open /Applications/Google\ Chrome.app --args --allow-file-access-from-filesFirefox and Safari do not have this restriction.
If you just want to run JS9 in this simple way, you are done. However, you might want to edit the js9prefs.js file to set up default values for colormaps, scaling, etc. See: JS9 Site Preferences for a description of the available parameters.
Running JS9 on the desktop simply requires that you install Electron.js, an app building platform that combines Node.js and the Chrome libraries. To install Electron, go to the release page: http://electron.atom.io/releases, choose the latest available stable release, and download the zip file for your platform. On a Mac, Electron.app should be installed in the /Applications or ~/Applications folder. On Linux, the electron program should be placed in your PATH. Note that Electron requires a relatively recent version of Linux: Ubuntu 12.04, Fedora 21, Debian 8, CentOS 7 (not CentOS 6).
Once the Electron app is installed, generate the JS9 quick-start files:
./mkjs9 -q Editing js9Prefs.json for Node.js helper ... Editing js9prefs.js for Node.js helper ... Generating js9 script for JS9 messaging and desktop use ... If you plan to use Electron.app with JS9, consider codesign'ing it: sudo codesign --force --deep --sign - /Applications/Electron.app/Contents/MacOS/Electron/ This will avoid repeated requests to allow incoming connections.The mkjs9 script will create a js9prefs.js file (for the browser) and a js9Prefs.json file (for the JS9 helper), which you can edit to add preferred JS9 properties, as well as a js9 script to start the JS9 app. On a Mac, you probably will want to codesign the Electron.app application to avoid repeated requests about incoming connections (see example above).
Run the js9 script to start the app and load data files:
# the -a switch tells the script to bring up the desktop js9 app js9 -a ~/data/casa.fits &and then use the same script to interact with the JS9 page (or any other JS9-enabled web page):
# without -a, the script sends commands to the JS9 display js9 cmap cool js9 regions circle
For more details, see: Desktop JS9.
Although none of these files are large, it always is worthwhile making downloaded files smaller to shorten the web page load time. One straight-forward way to do this, if you have control of the Apache web server, is to use the Apache rewrite capability to serve gzip'ed files in place of requested (uncompressed) files, if the former exist. Using this technique, your header can remain as shown above, but the Apache server will send gzip'ed versions when available. To do this, you can add code such as the following to the Apache httpd.conf:
Note that the JS9 Makefile contains a rule called install-gzip which will install gzip'ed copies of several JS9 files along side the uncompressed versions:
./configure ... make make install make install-gzip
JS9 supports server-side ("back-end") analysis on FITS data using a server-side helper. This capability allows you to execute virtually any command-line analysis program from JS9. The analysis command is run on the back-end server and results viewed on your web page. You can utilize your own web server as the JS9 back-end helper using CGI calls, or you can run a separate Node.js-based server to process JS9 back-end requests. The server-side analysis capability is especially useful for archive centers, but also can be attractive to individual users who want to integrate their own data analysis programs into JS9.
In addition, JS9 supports command-line messaging between the shell and JS9, pyjs9 Python messaging between the Python and JS9, and also has large-file support via the use of representation files. These capabilities requires the configuration of a Node.js-based server-side helper.
You configure a JS9 helper by adding additional switches to the configure command, e.g.:
# where to find cfitsio and install binaries, what sort of helper to build: ./configure --with-cfitsio=[path_to_cfitsio] --prefix=$HOME --with-helper=nod make make installSee: Installing a Server-side Helper for details.
./configure [your JS9 switches]
and then build the JS9 system using the make command:
The js9prefs.js file and js9Prefs.json file contain various default settings for JS9, e.g. default colormap and scale for image display. Feel free to edit this file to set up your own site-specific parameters. See JS9 Site Preferences for a description of the available parameters.
When the build is completed, you can install the JS9 into your web tree:
The first time you install JS9, your preference files will also be installed. After that, the install command will not install these files in order to avoid overwriting custom files. You can explicitly re-install preference files using:
make install-prefsor simply copy them into the install directory. Clean up the build directory using the command:
If you want to display our test data files in the JS9 demo pages, you must retrieve the JS9 data file tar file from the JS9 web site:
https://js9.si.eduand untar it into the JS9 web install directory. This will create sub-directories containing the image data. These data files also are available on GitHub: