[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: OT Re: reason behind fno-exceptions?
From:       Allan Sandfeld Jensen <snowwolf () one2one-networks ! com>
Date:       2001-07-31 7:57:21
[Download RAW message or body]

On Monday 30 July 2001 22:20, Guillaume Laurent wrote:
> On Monday 30 July 2001 19:58, Allan Sandfeld Jensen wrote:
> > C++ exceptions are horrible/horribly implemented,
>
> No, the way they are implemented in g++ sucks. That doesn't mean the
> concept is broken.
What's why I wrote /horribly implemented

>
> > and if you do serious work
> > where you need to know whats going on you deactivate them or use C.
>
> No.
Depends on what you do. Enabling exceptions means you have to write catches 
in order to stay in control. I use pleanty of exceptions in java and ML, but 
I _really_ dont use C++ for anything where I need that level of abstractions. 
i.e. I need to know whats going on, and most c++ programmers do.. 
fno-exceptions is a standard option most makefiles throw on gcc/g++. 
Especially becouse gcc set the previous standard w/o exceptions.

> > Somehow
> > they got accepted in one of the new C++ standards, which is why they are
> > on by default.
>
> There currently is only one C++ standard, and exceptions predate it by more
> than a decade. According to DaEC++, C++ exceptions were designed from 1984
> to 1989. The first implementation as defined in the ARM appeared in sping
> 1992. The Standard was finally voted in november 1997.

They where in fact a part of the original C++ by Bjarne, but for many years 
they remained unimplemented until a _common_ standard was made.

Sorry, I just hate c++ exceptions, and I don't want standard libraries like 
KDE or any other, to throw them unaware users.

-Allan
 
>> Visit http://master.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<

[prev in list] [next in list] [prev in thread] [next in thread]