Re: [Boost-testing] regression.py and Boost.Build

From: Robert Ramey (ramey_at_[hidden])
Date: 2007-09-08 13:51:02

Ahhhh -  The fact that one of our most dedicated testers finds
it inconvenient to test something is itself a red flag.

If our testers can't test it - how can we expect users to
want to use it?

If something isn't testable to demonstrate that it works as advertised
it shouldn't be part of boost.  This should be considered
a defect and be addressed accordingly.  That is it should
be fixed until its testable and passes the tests - just
like any boost library.

Robert Ramey

Martin Wille wrote:
> Robert Ramey wrote:
>> A very good idea - but why stop there?
>>
>> In the root/tools directory one finds
>>
>> boostbook
>> quickbook
>
> How far do you suppose this should go?
>
> FWIW, I'm not going to waste CPU cycles in an XML processing spree or
> brain cycles for any FOP setup nightmare.
>
>
> Regards,
> m