[erlang-questions] Latent bugs in Erlang/OTP

Mats Cronqvist mats.cronqvist@REDACTED
Mon Jan 15 12:51:24 CET 2007


Kostis Sagonas wrote:
> Partly due to extending Dialyzer and partly wanting to satisfying my 
> scientific curiosity, I've been spending a significant portion of my 
> recent time looking for bugs in Erlang code, mainly in OTP.
 > [...]
> There was a relatively recent comment to this mailing list by Mats 
> Cronqvist:
>      <quote>
> 	running dialyzer on well-tested code will turn up tons of
> 	errors. alas, that almost always turns out to be dead code.
>      </quote>
> 
> I guess one point I am implicitly trying to make here is that even dead 
> code might not be so harmless as its "dead" status might suggest, and 
> worth taking a closer look. Typically, competent programmers do not 
> write dead code.

   for the record, i agree with Kostis' comment about competence and lack of 
dead code.
   the problem i was lamenting is one of bootstrapping. if you have a bunch of 
code that generates 1,666(*) dialyzer warnings, it takes some convincing to go 
through all the old cruft and figure out which pieces can be safely removed. and 
  until you've done that the interesting warnings will be lost in the noise.
   perhaps the dialyzer can be enhanced to rewrite the sources dead-code free?

   mats

* actual number



More information about the erlang-questions mailing list