Re: [Pnet-developers] [Bug #2806] CSCC virtual memory exhausted

[Fergus Henderson]

On 19-Mar-2003, Rhys Weatherley <address@hidden> wrote:
> On Wednesday 19 March 2003 09:22 am, Fergus Henderson wrote:
> 
> > OK, how about "Out of memory, or missing System library"?
> 
> "... or internal compiler error, or bug, or unknown phenomenon, or ..."
> 
> It never ends.  There are a lot of reasons why that message may appear in a 
> non out of memory situation.

OK, then "Out of memory, or some other (unknown) error" would still be
a better message than the misleading "Out of memory" message.

> It is unreasonable to enumerate them all, for 
> an issue that basically boils down to "mdupont deliberately misused the 
> compiler and it acted strangely".

AFAIK mdupont did not deliberately intend to misuse the compiler,
and even if he did, I think it certainly possible that some other
user might do the same thing accidentally.

Saying that the compiler "acted strangely" is IMHO putting it too nicely --
"gave a very misleading error message" would be more accurate.  Issuing
seriously misleading error messages IMHO always constitutes a bug.

Since the fix of converting the misleading error message into a
less-than-fully-informative error message is trivial, why not just do it?
Uninformative error messages are undesirable, but they're not bugs.
Much better to have an uninformative error message than an incorrect one.

-- 
Fergus Henderson <address@hidden>  |  "I have always known that the pursuit
The University of Melbourne         |  of excellence is a lethal habit"
WWW: <http://www.cs.mu.oz.au/~fjh>  |     -- the last words of T. S. Garp.