Skip to content

Commit

Permalink
Initial commit from IPW v2.1.0
Browse files Browse the repository at this point in the history
  • Loading branch information
Scott Havens committed Mar 5, 2015
0 parents commit a451eda
Show file tree
Hide file tree
Showing 7,331 changed files with 459,908 additions and 0 deletions.
The diff you're trying to view is too large. We only load the first 3000 changed files.
3 changes: 3 additions & 0 deletions Changes
Original file line number Diff line number Diff line change
@@ -0,0 +1,3 @@
1998 Dec 1, J. Domingo
Increased command buffer size for imath from 100 to 1000
(constant CMD_BUF_LEN)
156 changes: 156 additions & 0 deletions Install
Original file line number Diff line number Diff line change
@@ -0,0 +1,156 @@
Configuring the Software
------------------------

1. Change directory into the directory where the IPW software will be built
and where this Install file is located.

2. Run the "configure" script. This script determines the type of machine
and operating system you're on, and sets up some local configuration
files.

./configure


Building the Software
---------------------

You can type "make help" to see what the make targets are and how they
may be used. If you have a problem and the build or install steps fail
to complete without error, you should clean everyting out using the
"make distclean" command, and start over from the beginning.

1. Set the environment variable "IPW" to the IPW directory (the
command below assumes you are in that directory).

*Note: the pwd command is surrounded by "backward" quotes " ` ", which
not only look different, but have a different meaning to the shell.

For Bourne shells (sh, bash):

export IPW=`pwd`

For C shells (csh, tcsh):

setenv IPW `pwd`

It is a good idea to check the IPW environment variable to make sure it is
properly set - if it isn't the build won't be successful. This can be done
by comparing the output from "echo $IPW" to the output from "pwd";
they should be the same.

2. To build the IPW software, run the "make" utility. We recommend that
you redirect the output into the "tee" command. That will allow you
to watch the progress the build process while saving the output to a
file. That file can be handy if an error occurs.

For Bourne shells (sh):

cd $IPW
make 2>&1 | tee Make.out

For C shells (csh):

cd $IPW
make |& tee Make.out


Installing the Software
-----------------------

1. The IPW software is built in the $IPW/src directory structure. In order
to use it, it must be copied ("installed") into other directories.
Historically, IPW commands are installed into the $IPW/bin and $IPW/sbin
directories. To install the software, use the "install" target with the
"make" utility. We again recommend redirecting the output to the "tee"
command, but this time into a different file.

For Bourne shells (sh):

make install 2>&1 | tee Make-install.out

For C shells (csh):

make install |& tee Make-install.out

2. After compiling and installing the IPW software, you can run a
series of tests to check the software. Again, we recommend that you
redirect the output to a file:

For Bourne shells (sh):

make tests 2>&1 | tee Make-tests.out

For C shells (csh):

make tests |& tee Make-tests.out

Errors are indicated by asterisks and a message. Certain programs will
not be tested because their behavior is either not reproducible without
problems (ipw, binexp), or they read from tape (tmpt, dempt), or other
reasons. They will be indicated by the line "Not testing program" which
is not to be construed as an error message.

Some programs may have the appendage "(sp)" added to the "Program passed"
message. This means one of two things:

1. your program evaluates float expressions in float precision rather
than casting to doubles first. In standard K&R C, all floats were
cast to doubles first, but ANSI C specified that arithmetic on float
operands may be done in single precision. This can make some
expressions evaluate faster, but the result will be slightly
different. In some cases your machine will come up with a different
result, and the program will return "Small float diff" which means
the result is off by a small amount which is most likely caused by
small differences in floating point evaluation.

2. Your system is missing a function such as random(), and a
functionally equivalent function has been substituted. The new
function may return different values for the same input, if that
is allowed by the function definition (e.g. random).


Using IPW software
------------------

1. Modify your shell initialization file to set the IPW environment variable,
and to include the IPW directory in your paths for commands and for
man pages.

For example, if the IPW directory is "/usr/local/ipw", then Bourne shell
users should add these lines to their ".profile" file in their HOME
directory:

# This example assumes that PATH and MANPATH are already
# set somewhere else in the file
#
IPW=/usr/local/ipw
export IPW
PATH=`$IPW/path user`:$PATH
MANPATH=`$IPW/path man`:$MANPATH

C shell users should add these lines to their ".cshrc" file in their
HOME directory:

# This example assumes that "path" and MANPATH are already
# set somewhere else in the file
#
setenv IPW /usr/local/ipw
set path=($path `$IPW/path -s user`)
setenv MANPATH ${MANPATH}:`$IPW/path man`

If a user is going to be programming IPW software, they should replace
the argument "user" with "system" in the calls to $IPW/path.


2. To view the HTML documentation for IPW software, open the following
file with a web browser (replacing $IPW with the full path to the IPW
directory):

$IPW/doc/html/intro.html

So, if the IPW directory is /usr/local/ipw, then the file URL would be:

/usr/local/ipw/doc/html/intro.html

Once you've opened this file in the browser, you can bookmark it for
future reference.
Loading

0 comments on commit a451eda

Please sign in to comment.