[ros-dev] [ros-diffs] [ion] 52596: [NTDLL]: More attempts at fixing up the loader, this time in the PE side of things.

[Elvis]

I do think Alex is talking about expressions such as (test1 & 55) == 
((test2 & test3) | 110) for example.

Or even (test1 == 44) && (test2 == 110) too.

I do agree the (()||()) style convention can avoid confusion for such 
expressions.

On 7/12/11 22:44, Timo Kreuzer wrote:
> Am 12.07.2011 00:52, schrieb Alex Ionescu:
>> Because that's the convention I keep trying to make you enforce. It
>> avoids confusion.
>>
>> Yes, you will now argue 10 scenarios where this doesn't matter (such
>> as below).
>>
>> And by doing so, you will have proven my point:
>>
>> 1) You can avoid using (() || ()) if you know the 10 exceptions and
>> operator precedence by heart. (Aleksey Method)
> But there is no operator involved in "(BoundEntry)", so you don't need
> to know any operator precedence.
>
>> or
>> 2) You can always use (() || ()) and never have to worry about
>> anything (in this context). (The Sane Method).
> In that case I also suggest using
> SomeVariable = (OtherVariable);
> return (Status);
> if ((x == 2)) return;
> SomePointer = &(Structure);
> (SomePointer)->Member = (0);
>
>>> /* Check if we got at least one */
>>> - if (BoundEntry || ImportEntry)
>>> + if ((BoundEntry) || (ImportEntry))
>>>
>
> Sorry for being sarcastic, but I just had to do this. You're literally
> asking for it :)
>
> I hope it won't lead to the introduction of something like
> #define MY_CONSTANT sizeof(FOO) + 20
>
>
> Regards,
> Timo
>
>
> _______________________________________________
> Ros-dev mailing list
> Ros-dev at reactos.org
> http://www.reactos.org/mailman/listinfo/ros-dev