| 1 |
On Sat, May 13, 2017 at 7:50 PM, Miroslav Rovis |
| 2 |
<miro.rovis@××××××××××××××.hr> wrote: |
| 3 |
> On 170513-12:53-0500, R0b0t1 wrote: |
| 4 |
>> On Sat, May 13, 2017 at 9:25 AM, Miroslav Rovis |
| 5 |
>> <miro.rovis@××××××××××××××.hr> wrote: |
| 6 |
>> > On 170510-20:03-0400, Walter Dnes wrote: |
| 7 |
>> >> On Wed, May 10, 2017 at 01:35:24PM -0500, R0b0t1 wrote |
| 8 |
>> >> |
| 9 |
>> >> > The option is "-fdiagnostics-color=never" or "-fno-diagnostics-color". |
| 10 |
>> >> > You can also set the environment variable GCC_COLORS to the empty |
| 11 |
>> >> > string. The latter is probably more useful in the context of portage. |
| 12 |
>> >> |
| 13 |
>> >> Thank you. I successfully tried... |
| 14 |
>> >> |
| 15 |
>> >> GCC_COLORS="" emerge icewm |
| 16 |
>> > Another tip to remember. |
| 17 |
>> > |
| 18 |
>> >> I suppose the next step is to add GCC_COLORS="" to make.conf. |
| 19 |
>> >> |
| 20 |
>> > I wonder why sticking " --color=n" in the EMERGE_DEFAULT_OPTS in |
| 21 |
>> > make.conf (e.g. mine is: |
| 22 |
>> > |
| 23 |
>> > EMERGE_DEFAULT_OPTS="--keep-going --with-bdeps=y --autounmask-keep-masks --ask --verbose --color=n" |
| 24 |
>> > |
| 25 |
>> > does only partly its job. Erratically, I'd say. You never know if it |
| 26 |
>> > will or not remove color... A bug should be posted for that, but I have |
| 27 |
>> > a partly broken system at this time... |
| 28 |
>> > |
| 29 |
>> |
| 30 |
>> That switch only handles the coloring of portage output. I suggested |
| 31 |
>> using GCC_COLORS precisely because "--color=n" doesn't seem to |
| 32 |
>> propagate to subcommands which do output coloring. |
| 33 |
>> |
| 34 |
>> Another program you might want to disable output coloring for is |
| 35 |
>> CMake, using CMAKE_COLOR_MAKEFILE=OFF. |
| 36 |
>> |
| 37 |
> |
| 38 |
> Thanks for the tip! But let me see... Like the above (repasting): |
| 39 |
>> >> GCC_COLORS="" emerge icewm |
| 40 |
> pr maybe stick in the /etc/portage/bashrc or in some other way? |
| 41 |
> |
| 42 |
|
| 43 |
As Walter suggested make.conf seems the ideal place. Should that fail |
| 44 |
I'm not entirely sure, as I believe portage does some pretty heavy |
| 45 |
environment sandboxing. |
Archives