On 9/16/24 10:13, Cy Schubert wrote:
In message <ZudpAKZS7JgS9ZZ_@int21h>, void writes:
On Sun, Sep 15, 2024 at 03:16:46PM -0500, Dan Mack wrote:
On 14.1 and prior, a CTRL-d will exit a bc session.
Today I noticed that on 3 different 15-CURRENT systems, it appears to
be ignored. Works fine otherwise and I can exit the bc session with
the 'quit' command okay.
I re-tested this on the system console on fresh login just to rule out
any terminal madness.
Here's a paste of what I see:
https://tpaste.us/VYya
I did a fresh install of 14.1 and it works as it did previously.
No biggie, just wondering if anyone else on -CURRENT can confirm/deny
this change on their system.
[void@vm5 ~ ] uname -KU
1400504 1400504
[void@vm5 ~ ] echo 2+2 | bc -l
4
[void@vm3 ~ ] uname -KU
1500023 1500023
[void@vm3 ~ ] echo 2+2 | bc -l
4
Of course the above works because the regression only affects tty users.
Here is some paint on the bikeshed :
enceladus# uname -apKU
FreeBSD enceladus 15.0-CURRENT FreeBSD 15.0-CURRENT #0
main-n271918-d7c87526b1c3-dirty: Mon Sep 2 09:55:54 UTC 2024
root@enceladus:/usr/obj/usr/src/riscv.riscv64/sys/GENERIC riscv riscv64
1500023 1500023
enceladus#
enceladus# bc -lq
scale=36
a(1)*4
3.141592653589793238462643383279502884
^D
^C
interrupt (type "quit" to exit)
ready for more input
quit
enceladus#
Yep. That is borked.
bc(1) now ignores EOF on the terminal while the above still works. You can
circumvent this by putting "export BC_TTY_MODE=0" into your .profile. The
side effect is that line editing will no longer work.
I will have to try that. However why would bc change at all?
--
--
Dennis Clarke
RISC-V/SPARC/PPC/ARM/CISC
UNIX and Linux spoken
