Am Samstag, 2. Juni 2007 14:49 schrieb Mattias Gaertner:
> On Sat, 2 Jun 2007 14:37:29 +0200
>
> Burkhard Carstens <[EMAIL PROTECTED]> wrote:
> > Am Samstag, 2. Juni 2007 14:21 schrieb Mattias Gaertner:
> > > On Sat, 2 Jun 2007 13:45:27 +0200
> > >
> > > Burkhard Carstens <[EMAIL PROTECTED]> wrote:
> > > > Hi all.
> > > >
> > > > The procedure InitializeDebugOutput in lcl/lclproc.pas should
> > > > map DebugText to @stderr instead of @output so stdout stays
> > > > "clean" for application use.
> > > > What do you think?
> > >
> > > DebugLn is debugging output, not errors.
> > > Maybe some warnings should be written to stderr instead, but
> > > often stdout and stderr are not synchronized.
> > > What debugln are disturbing you?
> >
> > e.g. "NOTE: TGtkWidgetSet.InvalidateRect during paint message:
> > TCurvePanel" (*)
> >
> > Currently I am writing a binary data stream analyzer, that must
> > read data from stdin and write it unchanged to stdout. So *any*
> > debugln to stdout is inacceptable in this case. Using
> > --debug-log='/dev/null' seems to be the only way to archive this.
>
> /dev/null, archive, :-)
My fault, I meant achieve (zustande bringen; Ziel erreichen) :-)
> Maybe we should add a function to close the debugtext.
Good idea but don't just close stdout like FinalizeDebugOutput does ;-)
> > (*) TCurvePanel is a selfmade component and maybe I should fix the
> > issue .. :)
>
> Yes. :)
TCurvePanel is a TPanel descendant that contains another panel + some
buttons. I don't call invalidate anywhere (except in
subpanel.onmousemove).
regards
Burkhard
_________________________________________________________________
To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
archives at http://www.lazarus.freepascal.org/mailarchives
