This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: libg++-2.8.1.3 and gcc-20000220 CVS
On Feb 22, 2000, Mark Mitchell <mark@codesourcery.com> wrote:
>>>>>> "Alexandre" == Alexandre Oliva <oliva@lsd.ic.unicamp.br> writes:
Alexandre> On Feb 20, 2000, Mark Mitchell <mark@codesourcery.com>
Alexandre> wrote:
>>> I think we should make the line-wrapping stuff always output at
>>> least k characters after the prefix. That way, even if the
>>> prefix is ridiculously long we don't just wrap the message on
>>> every line. Let's make `k' be 32 or so, for now.
Alexandre> it would be nice if this would bump the line length up
Alexandre> to the next multiple of the user-specified line length,
Alexandre> so that, if I specify that lines should be wrapped on
Alexandre> column 80 and the pathname was 70 characters long or
Alexandre> so, I'd get 160 characters-long lines.
> Why is that the right thing?
There's no right thing when the filename takes almost as much as the
specified line length. It's just that a multiple of the line length
is less arbitrary than 32 (even though you know I like powers of two
:-). The point is that, if you go to the next multiple, you're more
likely to waste less space in the user's terminal than breaking lines
in the middle on every other line.
> But, as we've been discussing, the right solution is to format
> error-messages better. We need to lose the prefix on every line, or
> at least the full path, in order to make error messages more readable.
Yup, that sounds even better.
--
Alexandre Oliva http://www.ic.unicamp.br/~oliva/ Enjoy Guaranį
Cygnus Solutions, a Red Hat company aoliva@{redhat, cygnus}.com
Free Software Developer and Evangelist CS PhD student at IC-Unicamp
oliva@{lsd.ic.unicamp.br, gnu.org} Write to mailing lists, not to me