Well, I don't think it's a big deal. I didn't think there was other code
doing ### type stuff, I wonder if that's just a holdover from before we
added the logging priorities. In general I'd prefer to avoid this type of
visibility escalation but if this is consistent with the local code it's
alright.

-- John

On Thu, Mar 11, 2010 at 9:41 AM, Darin Petkov <[email protected]> wrote:

> With the ### I followed what the code was doing on success or failure
> anyway -- just scroll down to the end of the original source file. I'm not
> sure I agree with ### either but consistency is good.
>
> Also, without this logging you don't see any error our build environment.
> It seems that the way this is currently implemented/used,
> client/bin/autotest would simply catch the exception and silently exit with
> an error code of 1. So you don't see anything. Add to that that our build
> system is _very_ noisy anyway.
>
> So, adding more noise in this case seemed to make sense...
>
> Darin
>
>
> On Thu, Mar 11, 2010 at 9:33 AM, John Admanski <[email protected]>wrote:
>
>> I'm not sure I really agree with the ### wrapped around the logging
>> message. I'm sure it makes it easier to spot this particular error, but
>> presumably the exception will be noisy enough to make this visible anyway?
>>
>> -- John
>>
>> On Thu, Mar 11, 2010 at 9:27 AM, Darin Petkov <[email protected]> wrote:
>>
>>> Without this change there was no indication if the test name pattern
>>> matched any tests or not.
>>>
>>>
>>> _______________________________________________
>>> Autotest mailing list
>>> [email protected]
>>> http://test.kernel.org/cgi-bin/mailman/listinfo/autotest
>>>
>>>
>>
>
_______________________________________________
Autotest mailing list
[email protected]
http://test.kernel.org/cgi-bin/mailman/listinfo/autotest

Reply via email to