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

List:       kde-pim
Subject:    Re: [Kde-pim] out of memory
From:       Mark Bucciarelli <mark () easymailings ! com>
Date:       2005-05-26 14:31:15
Message-ID: 4295DDB3.3010000 () easymailings ! com
[Download RAW message or body]

David Faure wrote:

>On Thursday 26 May 2005 15:32, Mark Bucciarelli wrote:
>  
>
>>In C, I'm used to checking that malloc returns a non-null pointer to 
>>make sure you app is not out of memory.
>>
>>How do you do this in C++?
>>    
>>
>
>This was discussed some time ago, and the consenssus was that in KDE code,
>we don't check for out of memory, and we don't need to.
>
Makes sense.  Because of how the OS' work, checking the rval of malloc 
doesn't even seem useful.

Follow up question---is there similar consensus on error handling?

I'm working on dcop iface and I am falling back to my C ways of having 
methods return an integer--0 for success, positive integer for error; e.g.,

#!/bin sh
RVAL=`dcop karm-123 dcopiface booktime libkcal-uid 20050505T1050 360`
if [ "$RVAL" -gt "0" ]
then
  echo `dcop karm-123 dcopiface geterror $RVAL`
  exit $RVAL
fi

It's nice b/c if you use this consistently, then you can bubble up all 
kinds of errors, add more detail later as the library grows and morphs:  
invalid resource, lock failed, async resource currently locked, bad task 
UID, confused resource manager, etc.

But most libkcal fcn's seem to simply return a true or false, and give 
little info on the specific error.  (I did see one use of kdError (sp?)).

So, is this approach (non-boolean rvals) also considered a waste of time?

m

_______________________________________________
kde-pim mailing list
kde-pim@kde.org
https://mail.kde.org/mailman/listinfo/kde-pim
kde-pim home page at http://pim.kde.org/
[prev in list] [next in list] [prev in thread] [next in thread]