Re: Format String issue w/ patch

[David O'Toole]

> -     if (!num_types) ox_die("%%OCTAL: No machine types loaded. \n");
> +     if (!num_types) ox_die("%OCTAL: No machine types loaded. \n");

There were two "%" characters because I wanted to print out a single
literal "%" (via fprintf to stderr.) Changing it to a single percent
character will actually introduce an error, since fprintf will start
looking for an argument, right...? 

It seems I've got some inconsistency in the error messages though,
thanks for pointing that out. 

>    In my initial grok of the tree I noticed a bit of 
> obfuscation.  I have created a patch to try to minimize 
> core dumps in the octal engine itself witch might be 

What core dumps are you seeing? What system are you running it on? Were
you trying to make a change to the CVS that caused coredumps? In which
sections? 

A good portion of the core is messy because it uses plain C for objects,
which is error-prone and usually looks obfuscated. This is the primary
reason I've started re-writing it in Objective-C (you might already
notice some Objc files in the CVS repository, and more are on the way.)