Re: [boost] [thread] age-old, fatal, trivial-to-fix bug

From: Ulrich Eckhardt (doomster_at_[hidden])
Date: 2007-02-10 13:58:47

On Saturday 10 February 2007 12:24, Roland Schwarz wrote:
> Ulrich Eckhardt wrote:
> > (Sorry for the flashy topic, but I reported the bug already ages ago and
> > I'm slightly annoyed that it is still present. But hey, at least I could
> > restrain myself from using multiple exclamation marks.)
>
> If I remember well this topic already has been answered. What you are
> reporting is not a bug, but a cosmetic issue.

The code is simply broken, and dangerously so. I wouldn't call it cosmetic.

> The code in question will never be called.

The broken function is declared in a public header, why shouldn't some user 
call it? Also, I know a way to make sure that it isn't called...

> I intend to fix this in the next release, since some rewrite will be
> necessary to correctly return system messages. It is too short before
> release to introduce new behaviour.

Ah, wait: the patch (probably) applies to older versions, but it was taken 
against the CVS which reports version 1.35!

> Please correct me if I am wrong in my assumption, that the reported
> behavior will not affect a user of the library.

Unnecessary bloat. Maybe, when someone calls the code, crashes.

I'm a bit frustrated, please don't take it as a personal attack.

Uli