Re: [Re: [gtkmm] IOCondition holds illegal values]

[Murray Cumming <murrayc usa net>]

Eric Buddington <eric ma-northadams1b-39 bur adelphia net> wrote:
> On Tue, Sep 24, 2002 at 09:12:44PM +0300, Tassos Bassoukos wrote:
> > On Tue, 24 Sep 2002, Eric Buddington wrote:
> > 
> > > In the case of a file descriptor that is a pipe to a dead child, I am
> > > getting values '20' and '12', which match none of IO_IN, IO_OUT,
> > > IO_PRI, IO_ERR, IO_HUP, or IO_NVAL.
> > 
> > Glib::IOCondition is a set of ORed flags, you have to check for
individual
> > flags by ANDing with the value. The values you are getting are equivalent
> > to (IO_OUT | IO_HUP) and (IO_OUT | IO_ERR) respectively.
> 
> Great - thanks for that fast answer.
> 
> Now - is this sort of thing documented in the gtkmm code, or the
> underlying gtk code? If I can figure it out, I'm inclined to
> contribute a doc patch or two.

We use Doxygen. It should be quite easy to figure out just by looking at the
existing documentation in the .hg/.ccg or .h/.cc files. glibmm probably uses
just .h/.cc files.


Murray Cumming
murrayc usa net
www.murrayc.com