Re: [evolution-patches] patch for the bug #64398 [calendar]

[Rodrigo Moya <rodrigo novell com>]

On Tue, 2004-08-31 at 23:01 +0800, Not Zed wrote:
> 
> umm, you should not rely on an assertion to save you from a case which
> could happen in normal code running.
> 
> if it isn't a code error, but a possible run-time case, it should just
> be a normal if() statement.  If it is a code error, the source of the
> error should be fixed ideally.
> 
> i.e. the code shouldn't crash if the assertions aren't present and
> there are no bugs present.
> 
you are right, in fact, get_deltas is called via a g_timeout_add, so I
guess we should be removing that timeout callback when the backend dies,
so that it never gets called when the backend is NULL.

Sorry for not catching it before


Rodrigo Moya <rodrigo novell com>