I sent out a reply to the same effect as Junio's longer one in this
side-thread. You can disregard that. Sorry for not first reading the
whole thread.
Tom Grennan <tmgrennan@xxxxxxxxx> writes:
> On Wed, Feb 29, 2012 at 07:26:10PM -0800, Junio C Hamano wrote:
>>Tom Grennan <tmgrennan@xxxxxxxxx> writes:
>>
>>> It's trivial to remove these "quiet" and "silent", but to me that's the
>>> only value added by these patches. More seriously, the remaining
>>> modernization still seems much larger than its value.
>>
>>Don't do that, then ;-).
>>
>>Some older scripts do redirect the output from the commands to /dev/null
>>but that dates back before we made the default reasonably silent, and in
>>"modern" style we tend to keep them sent to their standard output to help
>>debuggability. These quiet/silent takes us to the prehistoric times.
>
> Hey! I am prehistoric:-)
>
> Like I said, I think there is currently a debug distraction with verbose
> mode. However, rather than hiding expected failures and diverting other
> output as I had, perhaps we should dup stderr to stdout in verbose mode
> so error messages show up near the logged invocation when piped through
> a pager (i.e. mimic "|&"). With this, one can quickly scan past the
> noise to focus on the broken cases.
If you have trouble finding the broken case, you can run with -v -i.
I also think you are making an argument for a different feature (which
does not rob us of having all the debug output): test-lib.sh could
perhaps redirect the test output to a file, and dump the file to stdout
only if the test failed. Perhaps --verbose-failing or something like
that.
Otherwise, your proposal is restricting us to having only an "easily
scannable" amount of output per test, perhaps 5-15 lines. Which I
personally think is an insane restriction for something that was
intended for debugging in the first place.
--
Thomas Rast
trast@{inf,student}.ethz.ch
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html