Peter,
I take it this is to prevent the directory not found error from your
previous posting. This appears at first glance to be a legitimate
error status from the call to sys$check_access within cando_by_name.
Do you happen to know what dirspec it was complaining about? Hunches
would include the possibility that do_fileify_dirspec and/or
do_tovmsspec have changed behavior out from under cando_by_name,
perhaps giving it a directory name that is not quite what it really
wants.
At 2:07 PM -0800 3/3/2000, [EMAIL PROTECTED] wrote:
>OK, in addition to changing the use_64bitint symbol name to use64bitint
>in [.vms]subconfigure.com I found that I had to put this logic reversal
>into [.vms]vms.c. Could someone take a close careful look at this?
>
>Thank you.
>
>--- vms/vms.c;1 Fri Mar 3 02:48:20 2000
>+++ vms/vms.c Fri Mar 3 13:55:44 2000
>@@ -4650,9 +4650,12 @@
> if (retsts == SS$_ACCONFLICT) {
> return TRUE;
> }
>- _ckvmssts(retsts);
>+/* hacking for DECC 6.0 - pvhp
>+ _ckvmssts(retsts);
>+ */
>
>- return FALSE; /* Should never get here */
>+ return TRUE; /* hacking - pvhp */
>+/* return FALSE; /* Should never get here */
>
> } /* end of cando_by_name() */
> /*}}}*/
>End of evil hackery.
>
>
>Does vmsperl _need_ this? I'll try to get to the DECC 5.3 box eventually.
>
>Peter Prymmer
____________________________________________
Craig A. Berry
mailto:[EMAIL PROTECTED]
