gcl-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gcl-devel] Re: gcl and the random tester


From: Camm Maguire
Subject: [Gcl-devel] Re: gcl and the random tester
Date: 26 Feb 2004 10:27:41 -0500
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2

Greetings!

"Paul F. Dietz" <address@hidden> writes:

> Camm Maguire wrote:
> > Greetings, and thanks as always Paul.  Is it meaningful to get a feel
> > for the compiler reliability modulo this missing function?  I'm pretty
> > sure I won't have time to get it in for at least a couple of weeks,
> > and it would be good to get some feel for how solid the current state
> > is.  Again I can probably contribute computer time if you feel like
> > pooling these tests across machines.
> 
> Yes, it is worthwhile.  Currently these operators are conditionalized
> out of the generator, so we can test without them.  Without them,
> gcl doesn't have any compiler bugs that I know about that the random
> tester can stimulate (although some might exist).
> 

OK.  I want to avoid hype at all costs as we all know there are plenty
of bugs which remain, as is the case with any large and complex
program.  At the same time, I think it is valuable for the project to
be able to make thoroughly qualified, indisputably true statements
about the level of quality which has thus far been achieved, both as
an encouragement to its contributors, as well as an attractor to
future users and developers.  

It would appear then that within the scope of problems encompassed by
your random compiler tester without load-time-value, that we might be
able to demonstrate that GCL is as solid as CLISP, which, again to my
understanding, put both some distance ahead of the other lisps, at
least within this arguably small scope of problems.  

Now as we both know, many other systems show failures because they try
to do more, and of course clisp does the very least complex compile.
It would seem that GCL has achieved a reasonably aggressive compile
with a very low rate of errors, again within this small set of
problems.  

Is such a statement unreasonable to make, presuming of course it is
actually born out in the test results?  If not, perhaps we could put
some such on our website and/or in the release notes at release time.

Take care,

>       Paul
> 
> 
> 
> 
> 

-- 
Camm Maguire                                            address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah




reply via email to

[Prev in Thread] Current Thread [Next in Thread]