Re: ls is broken, what's next cd?

[Bernhard Voelker]

On 02/13/2018 08:35 AM, Harald Dunkel wrote:

If I got this correctly, then the folks to benefit from this change
are GUI users, using copy-and-paste. Is this an argument to affect
the command line users?


Well, unless one is working on a bare tty without X or via network,
I would guess that there is a terminal application like xterm,
gnome-terminal, a kde terminal, etc., involved in 99% of the cases.
Thus, there /is/ the possibility to use for copy/paste.

Furthermore, the quoting gives you immediate visual feedback that
there's something unusual in the corresponding file name.


BTW, apparently the shell quoting-style in coreutils 8.26 (Stretch)


8.26 is quite old, and there have been further tweaks to the quoting
feature meanwhile (we're at 8.29).


misses to properly quote iso8859 chars, e.g. German Umlaute. Sample:

% ls --quoting-style=shell -al /tmp/Qual*
-rw-r--r-- 1 root root 0 Feb 13 08:27 '/tmp/Qualit?tssicherung final'


I'm not sure what's happening there, but it seems that your terminal
doesn't know how to display the 2-byte representation "c3 a4"; neither
gnome-terminal nor xterm have problems to display it here:

  $ /bin/ls -log /tmp/Q*
  -rw-r--r-- 1 0 Feb 15 07:59 '/tmp/Qualitätssicherung final'

  $ /bin/ls /tmp/Q* | od -tx1z
  000 2f 74 6d 70 2f 51 75 61 6c 69 74 c3 a4 74 73 73  >/tmp/Qualit..tss<
  020 69 63 68 65 72 75 6e 67 20 66 69 6e 61 6c 0a >icherung final.<
  037

Have a nice day,
Berny

bug#30453: Bug in BC calculator version 1.07.1

[Martin Møller Skarbiniks Pedersen]

This is not a bug.
>From the man-page:

$ man bc | fgrep -A5 "expr * expr"
   expr * expr
  The result of the expression is the product of the two
expressions.

   expr / expr
  The  result  of  the  expression  is  the quotient of
the two expressions.  The scale of the result is the value of the
variable
  scale.


So expr * expr is not scaled but expr / expr is scaled.

Regards
Martin



On 14 February 2018 at 07:52, Thomas Bracken  wrote:
> Hello,
>
> Your BC calculator does not do any truncation for the scale value
>
> when it does a multiplication.  See example below.
>
> It works fine on division, but not on multiplication.
>

This is not a bug.
>From the man-page:

$ man bc | fgrep -A5 "expr * expr"
   expr * expr
  The result of the expression is the product of the two
expressions.

   expr / expr
  The  result  of  the  expression  is  the quotient of
the two expressions.  The scale of the result is the value of the
variable
  scale.


Regards
Martin

Re: Cross compiling coreutils to 32 bit

[Assaf Gordon]

Hello,

On Wed, Feb 14, 2018 at 09:59:13AM +0200, Kevin Wilson wrote:
> I am trying to cross compile coreutils-8.27 to 32 bit on
> x86_64 machine under Fedora.
> 

I would recommend to following:

First,
Install "multilib" support. In debian/ubuntu, the command is:
   apt-get install gcc-multilib g++-multilib

With Fedora it should be something equivalent, see here:
   
https://www.cyberciti.biz/tips/compile-32bit-application-using-gcc-64-bit-linux.html


Then,
Assuming you are using gcc, use the following command:
  ./configure CPPFLAGS="-m32" LDFLAGS="-m32" TIME_T_32_BIT_OK=yes
  make

The resulting binaries will then be 32-bit.

The TIME_T_32_BIT_OK is needed due to a recent change that checks
for 64-bit time_t variable.

HTH,
 - assaf

Cross compiling coreutils to 32 bit

[Kevin Wilson]

Hi,

I am trying to cross compile coreutils-8.27 to 32 bit on
x86_64 machine under Fedora.

What I am running is:
untar the tar file.

./configure --target=i686-pc-linux-gn

and then "make".

It builds succesfully.
However, when I am runing:

file src/stty
I am getting:
src/stty: ELF 64-bit LSB executable, x86-64, version 1 (SYSV),
.
...
And likewise for all other generated binaries under src.

I would expect that the generated binary will be 32 bits, as I am
trying to cross compile to 32 bits.
I also tried with to play with:
./configure --host=

 but got the same.

What am I doing wrong ? it could be that I am missing something trivial?

Regards,
KV