oder at eleks.lviv.ua
Sun Oct 7 07:36:08 MST 2007
----- Original Message -----
Subject: Re: [ODE] 0.9-rc1
>> As for measures - I'm also the adherent of hard measures. If validation
>> fails, you should fix your code, not remove validation.
> I'm opposed to the policy to add asserts like in ODE, for the following
> In unlikely situations these pieces of code might be reached (esp. the one
> the normalize methods).
First of all, you are expected to build release with assertions turned off.
So you can be sure user will not see assertion failure.
> It is not very nice if a user reaches such situations
> and the whole applications crashes as a result.
Well, you think it is better to produce incorrect result and keep working
even if user may not notice that error?
I disagree with you completely. It is always better to crash application
once and fix the problem rather than to have it producing incorrect results
or showing unpredictable and/or inexplicable behavior for years.
There can be cases when human lives depend on actions of software. And if a
process dies, you can easily detect it with a watchdog process and take
protective measures. But if process continues functioning but does not do
its job correctly it can be much more dangerous.
> Printing warning, throwing exceptions (if they are used),
> returning/setting NaN
> are ok. Terminating the application is not, as it might happen outside the
> testing phase.
-- ICQ: 36361783
More information about the ODE