Re: typedef vs. typename confusion

[mrs at wrs dot com (Mike Stump)]

> Date: Thu, 06 Aug 1998 07:39:05 -0400
> From: Noel Yap <nyap@garban.com>

> Why allow people to unknowingly write sloppy code.

Why to you call it unknowingly?  They do it, we tell them they did it,
we tell them to fix their code, all by default.  Is you claim that we
don't tell them to fix their code or that we can't expect the user to
be able to read what we tell them.  Or is your concern that while they
may read it and understand it when we tell them to fix their code,
that they may ignore it?  Do you think that we should also force them,
coerce them to fix their code?

> If they want to write non-compliant code, they should know it's
> non-compliant and, therefore, should be the ones to specify a flag.

You seem not take have read what the others before you have said.
Some of them aren't writing non-conforming code.  The code was written
against the arm, the arm was the standard, their code was perfectly
compliant to the standard at the time.  There is more to the world
than just ANSI/ISO C++, we also try and provide some ARM features.
Some people and projects simply _require_ it in places.  g++ is also a
arm->ANSI/ISO C++ conversion tool.  The tool to be easy to use,
identifies when such a thing is necessary _automatically_ and yes, by
default, and helps the poor user along.  You can claim that this is
bad, and does a disservice to the users.  That users are better off
being forced, kicking and screaming, into the new standard.  You can
claim that C++ is better off by giving everyone that trys a new
compiler, a bad taste in their mouth when none of their code compiles.
Try doing a user survey sometime.

What's reall fun, it when the compiler is so pedantic, that it
complains about your non-compliant system header files, and refuses to
compile any user code, now, that's helping the user!  That way they
can know their system vendor isn't compliant!  Not!