next message in archive
no next message in thread
previous message in archive
previous message in thread
Index of Subjects
Index of Subjects
On Jun 18, 10:35am, James Fifield wrote:
}
} Sorry for taking so long to respond to this John, it's been a little hectic here.
I meant to get back to this sooner, but I've been busy and had
some personal things. Anyways, I have a couple of last minute things
that I hope can get into the beta 2.0.
} > Date: Thu, 5 Jun 1997 06:36:27 -0300
} > From: John Nemeth <jnemeth@cue.bc.ca>
} > To: csuite-tech@chebucto.ns.ca
} > Subject: HIGH: Beta 2 initial testing on Solaris 2.5
} >
} > - seeing lots of 'couldn't set locale correctly' from the echo command
} > - seems Sol 2.5 doesn't like something about the way csuite does locale
} > - ignoring this for now since it doesn't seem to cause any harm
}
} ???
I haven't had time to look at this closely. Also, I learned C
long before locales were invented. So first I have to figure out how
they work, before I can figure out what is wrong.
} > - Sol 2.5 doesn't like something in the install-fr: target of
} > $CS_ROOT/build/makefile
} > - comment out for now
}
} I think the way we do things generates a lot of erroneous errors like:
} make[4]: *** No rule to make target `install-fr'. Stop.
}
} This is simply because there are no french documents to install for that section.
Actually, this breaks it completely. If I don't comment out the
install-fr target in the main makefile, the Sol. 2.5 make program
rejects the makefile, and I can't do anything (i.e. starting at lynx
step 4).
} > - mksub blows up because of '. $CS_ROOT/lib/cs-shlib'
} > - cs-shlib has export NAME=val which isn't a valid Bourne shell construct
} But cs-shlib should use bash, (or at least it does now. :D*) Thanks for alerting us
} to this though. Unfortunately our test machines are somewhat limited with respect to
} problems like this.
cs-shlib is an "include" script. It runs in the context of
whatever script calls it, not in its own shell. I've made the
following change, which seems to have solved the problems.
*** cs-shlib.orig Sat Jun 28 00:13:08 1997
--- cs-shlib Sat Jun 28 00:13:57 1997
***************
*** 10,20 ****
## cs-shlib - a library of shell functions for use from scripts
## Note: libraries must be used with bash-2.0+ as the shell.
! export CS_ROOT=@@@CS_ROOT@@@
! export TEXTDOMAIN=csuite
! export TEXTDOMAINDIR=@@@TEXTDOMAINDIR@@@
lang_init () {
IFS=" ,;"
set -- $HTTP_ACCEPT_LANGUAGE
--- 10,22 ----
## cs-shlib - a library of shell functions for use from scripts
## Note: libraries must be used with bash-2.0+ as the shell.
! CS_ROOT=@@@CS_ROOT@@@
! TEXTDOMAIN=csuite
! TEXTDOMAINDIR=@@@TEXTDOMAINDIR@@@
+ export CS_ROOT TEXTDOMAIN TEXTDOMAINDIR
+
lang_init () {
IFS=" ,;"
set -- $HTTP_ACCEPT_LANGUAGE
} > - mksub should use bash, or cs-shlib needs to be cleaned up
} Unfortunately, in order to run mksub as bash we'd have to know where it is, which would
} mean we'd have to find it, and run 'mksub' on itself. I'm sure you can see the dilema. :)
Yep, but the patch above should fix the problem.
Next note will be the list of patches needed to get it to
compile, which should be out in the next few hours.
}-- End of excerpt from James Fifield
next message in archive
no next message in thread
previous message in archive
previous message in thread
Index of Subjects