Date: Sun, 08 May 2005 18:43:00 +0900 (JST)
Message-Id: <20050508.184300.38707315.garrigue at math.nagoya-u.ac.jp>
To: sven.luther at wanadoo.fr
Cc: julien.puydt at laposte.net, lablgtk at kaba.or.jp, 246846-quiet at bugs.debian.org,
        246846-submitter@bugs.debian.org, debian-ocaml-maint@lists.debian.org
Subject: Re: Bug#246846: Is this bug still present in lablgtk2
 2.4.0+2005.02.18-1 ?
From: Jacques Garrigue <garrigue at math.nagoya-u.ac.jp>
In-Reply-To: <20050508054922.GA24939 at pegasos>
References: <20050507141304.GA11111 at pegasos>
	<427D082C.2050505@laposte.net>
	<20050508054922.GA24939@pegasos>
Mime-Version: 1.0
Content-Type: Text/Plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit

From: Sven Luther <sven.luther at wanadoo.fr>
> On Sat, May 07, 2005 at 08:25:48PM +0200, Julien PUYDT wrote:
> > Sven Luther a écrit :
> > >Hello,
> > >
> > >I would like to know if this bug was still present in the current version 
> > >of
> > >lablgtk2 in sarge ? 
> > 
> > I don't know if it's the release in sarge, but I just rechecked the ones 
> > I pointed as problematic with the current lablgtk2 in unstable, and on 
> > those the result is:
> 
> Thanks for testing and your fast reply.
> 
> > * csview.ml:
> > File "csview.ml", line 51, characters 4-6:
> > Syntax error

csview.ml uses stream parsers, which are provided by camlp4.
There is now a comment in the file:
(* Compile with
     ocamlc -pp camlp4o -I +lablgtk2 lablgtk.cma csview.ml -o csview
   or run with
     lablgtk2 camlp4o.cma csview.ml <file.csv>
*)

> > * events.ml:
> > doesn't crash but gives things like:
> > (ocaml:7938): Gtk-CRITICAL **: gtk_text_view_start_selection_drag: 
> > assertion `text_view->selection_drag_handler == 0' failed

events.ml is old code, which was an attempt to get the cursor position
using the old text widget. This was always buggy. Fortunately, the new
text widget causes no such problem (events2.ml).

> > * tictactoe.ml:
> > File "tictactoe.ml", line 30, characters 26-48:
> > Unbound value GtkNew.make_new_widget

GtkNew is not ported. This was never clear how useful it would be to
implement it on the ML side.
Should remove this test for the time being.

> > * text/test_text.ml:
> > File "text/test_text.ml", line 74, characters 11-25:
> > Unbound value GText.tagtable

Oops. API changes. I've just fixed it in CVS.

Thanks for all these reports.

Jacques