Installation
Notes on installing and running
the Tapestry release
There are several steps in installation and setup which must be completed
for Tapestry to successfully run.
Setting up supporting software (JDK and Jikes):
- Note: this only needs to be completed once for each machines Tapestry
needs to run on
- Download and install the IBM JDK 1.3. This is available for all
platforms at the IBM
website.
- For Debian: install alien (apt-get update; apt-get install alien
), and then (as root) do alien IBMJava2-SDK-1.3-10.0.i386.rpm; dpkg
-i ibmjava2-sdk_1.3-10_i386.deb.
- For SuSE, Redhat and others: install the RPM
- For FreeBSD: copy the .tgz file to /usr/ports/distfiles/
. Then (as root):
cd /usr/ports/java/linux-ibm-jdk13/; make all install; and
follow the instructions.
- Install the Jikes Java compiler, which is available via the IBM Jikes
website
here.
- On Debian, this can be done with a simple apt-get update; apt-get
install jikes (from testing).
- For SuSE, Redhat and others: install the Jikes RPM
- For FreeBSD: (as root) cd /usr/ports/java/jikes/; make all
install
Unpacking and installing the Tapestry distribution:
- Note: this needs to be done for each user who wants to run an instance
of Tapestry
- Add the JDK to your classpath.
Linux: export OSTORE_JAVAHOME=/opt/IBMJava2-13
FreeBSD: export OSTORE_JAVAHOME=/usr/local/linux-ibm-jdk1.3.0
- Add the JDK to your path: export PATH=$OSTORE_JAVAHOME/bin:$PATH
- Pick a directory to store your Tapestry source code, for instance "
/home/ravenben/tapestry"
- Unpack the tar.gz or zip file into the Tapestry directory
- Add the resulting pond directory (/home/ravenben/tapestry/pond
) to your CLASSPATH environmental variable. This directory will
be the root directory from which all your java class names will reference.
- Download the Cryptix JCE library from
Cryptix
, and add the jar files to your CLASSPATH variable.
- The Sandstorm code will already be inside the pond directory under "
mdw". For more information on Sandstorm/SEDA, see the SEDA
website
. Compile the code and add the full path of the pond/mdw/lib directory
to your LD_LIBRARY_PATH variable. To compile the code:
Linux: cd ~/mdw; CLASSPATH=~/:. make.
FreeBSD: Compile NBIO on a linux machine, as specified above. Then copy
~/mdw/nbio/jni/libNBIO.so and ~/mdw/util/jni/libMDWUtil.so
to the FreeBSD machine. Check that the ~/mdw/lib/libNBIO.so and
~/mdw/lib/libMDWUtil.so links are set properly.
- The Berkeley DB java libraries should already be installed under pond/db-2.7.7.
- You should now be ready to compile and test the Tapestry code. Do
cd /home/ravenben/tapestry/pond/ostore; make
- You can run the regression test(s) to verify your installation. The
tests are found in the regression subdirectory under pond. The tests
can be run as
cd /home/ravenben/tapestry/pond/regression; ./run-tests
Ben Y. Zhao
ravenben@cs.berkeley.edu
April 5, 2002