It turns out that all of the odd issues I was having were resolved by
ditching SRCDIR. I like to build from a separate directory from source
as I tend to leave intermediate files in a tmpfs partition, to save my
SSD some unnecessary thrashing, so I naturally attempted to use SRCDIR.
Anyhow, building from the same directory as the source resolved all of
the issues.
Here's hoping Oleg's cmake branch comes together soon. :)
-Dan
On 14-04-06 08:55 AM, Daniel Leslie wrote:
I'm trying to create a 'portable' distribution of chicken and am
running into a simple issue. Basically, the built-in library search
path isn't always valid, and csc and csi don't appear to pay attention
to LD_LIBRARY_PATH.
Thus, the following:
bin $ ./csi -n
CHICKEN
Error: (string-append) bad argument type - not a string: #f
bin $ echo "(display \"Hello world\")" | ./csc - -o foo
Error: (string-append) bad argument type - not a string: #f
Call history:
<syntax> (##core#begin (display "Hello world"))
<syntax> (display "Hello world") <--
Error: shell command terminated with non-zero exit status 17920:
bin/chicken - -output-file foo.c
Alright, so with csi and csc we can get around this by explicitly
providing a -include-path parameter at the command line. But what of
installing eggs? Chicken-install takes no such parameter, and
invocations of csc from setup scripts are not mutable to accept the
parameter.
It seems to me that the easy fix would be to import a path from some
environment variable by default. Does such an environment variable
exist already?
-Dan
_______________________________________________
Chicken-users mailing list
[email protected]
https://lists.nongnu.org/mailman/listinfo/chicken-users
