I created a program (using DISKSAVE) which reads from standard input and
writes to standard output. The problem is that I want to differentiate between
two sorts of output. The first one is the program output which can be
directed to a file (i.e. under unix: % myprogram < input > output). The other
type include messages concerning the progress of the program which should
appear on the screen.
In LCL4.0 *standard-output* stream is equal to *error-output*. So the output
directed through any of these will make no difference. Even *terminal-io* is
the same. I don't want to direct the error output (or progress messages) to
a file because it is important for the user to see.
Another question: is there a way to decrease the size of the program (Lisp Image)?
Thank you in advance for your help.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Dr. Haitham Lababidi Tel +44 31 650 4862
Department of Chemical Engineering Email ·······@uk.ac.ed
Edinburgh University
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
In article <······················@chemeng.ed.ac.uk> ·······@Chemeng.Ed.Ac.Uk writes:
>In LCL4.0 *standard-output* stream is equal to *error-output*. So the output
>directed through any of these will make no difference. Even *terminal-io* is
>the same. I don't want to direct the error output (or progress messages) to
>a file because it is important for the user to see.
The following Lucid-specific code will connect *error-output* to the Unix
stderr stream:
(defconstand unix-stderr 2)
(with-open-stream (*error-output* (make-lisp-stream :output-handle unix-stderr))
...)
>Another question: is there a way to decrease the size of the program (Lisp Image)?
Use the Delivery Toolkit. It includes a treeshaker that removes unused
code.
--
Barry Margolin
System Manager, Thinking Machines Corp.
······@think.com {uunet,harvard}!think!barmar