Re: [pygtk] pygtk-1.99.5
On Thu, 29 Nov 2001 17:24:16 +0800 James Henstridge [EMAIL PROTECTED] wrote: John Finlay wrote: Hi James, Is there any maintenance effort being done on the old pygtk (0.6.8) source base or is it considered complete? Just wondering because I have encountered some problems that I would report if there is still some development activity. I should probably make another release off the stable platform. There are a few fixes in the tree over what is in 0.6.8, which would be quite useful for many people. I just haven't gotten round to it yet. If you have bug reports, please put them in bugzilla. Where is the bugzilla URL and how do I say this in bugzilla speak: Why does checking to see if a GtkTreeItem is a subtree (with item.subtree != None) result in a segfault? -- Jon Nelson\|/ \|/ Gort, [EMAIL PROTECTED]@'/ ,. \`@ Klaatu C and Python Programmer /_| \__/ |_\ barada Motorcycle Enthusiast\__U_/ nikto. ___ pygtk mailing list [EMAIL PROTECTED] http://www.daa.com.au/mailman/listinfo/pygtk
Re: [pygtk] pygtk-1.99.5
Where is the bugzilla URL and how do I say this in bugzilla speak: Why does checking to see if a GtkTreeItem is a subtree (with item.subtree != None) result in a segfault? Here is a a version of the _wrap_gtk_tree_item_get_subtree source that fixes my problem. Remove foo and use return PyGtk_New( (GtkObject *) GTK_TREE_ITEM... to simulate the old behavior. PyGtk_New was getting passed a NULL pointer. static PyObject *_wrap_gtk_tree_item_get_subtree(PyObject *self, PyObject *args) { PyObject *obj; GtkObject *foo; if (!PyArg_ParseTuple(args, O!:gtk_tree_item_get_subtree, PyGtk_Type, obj)) return NULL; foo = (GtkObject *) GTK_TREE_ITEM(PyGtk_Get(obj))-subtree; if (foo == NULL) { Py_INCREF(Py_None); return Py_None; } return PyGtk_New(foo); } -- Jon Nelson\|/ \|/ Gort, [EMAIL PROTECTED]@'/ ,. \`@ Klaatu C and Python Programmer /_| \__/ |_\ barada Motorcycle Enthusiast\__U_/ nikto. ___ pygtk mailing list [EMAIL PROTECTED] http://www.daa.com.au/mailman/listinfo/pygtk
Re: [pygtk] pygtk-1.99.5
Jon Nelson wrote: On Thu, 29 Nov 2001 17:24:16 +0800 James Henstridge [EMAIL PROTECTED] wrote: John Finlay wrote: Hi James, Is there any maintenance effort being done on the old pygtk (0.6.8) source base or is it considered complete? Just wondering because I have encountered some problems that I would report if there is still some development activity. I should probably make another release off the stable platform. There are a few fixes in the tree over what is in 0.6.8, which would be quite useful for many people. I just haven't gotten round to it yet. If you have bug reports, please put them in bugzilla. Where is the bugzilla URL and how do I say this in bugzilla speak: Why does checking to see if a GtkTreeItem is a subtree (with item.subtree != None) result in a segfault? Go to http://bugzilla.gnome.org/. Create an account if you haven't already done so. You can report bugs or query for existing bugs there. When reporting bugs, use the gnome-python product, and pygtk for pygtk specific bugs. James. -- Email: [EMAIL PROTECTED] WWW: http://www.daa.com.au/~james/ ___ pygtk mailing list [EMAIL PROTECTED] http://www.daa.com.au/mailman/listinfo/pygtk
Re: [pygtk] pygtk-1.99.5
Jon Nelson wrote: Here's a patch for the NULL-pointer dereference problem with GtkTreeItem.subtree *Please* either incorporate this or identical functionality, otherwise using GtkTreeItem.subtree to test if a GtkTreeitem is a subtree is fraught with sigsegs. This bug is fixed in CVS. James. -- Email: [EMAIL PROTECTED] WWW: http://www.daa.com.au/~james/ ___ pygtk mailing list [EMAIL PROTECTED] http://www.daa.com.au/mailman/listinfo/pygtk
Re: [pygtk] pygtk-1.99.5
Hi Jon, This is one of the porblems I encountered as well. I believe the fix needs to be applied in the generate/gktlists.defs file. However, it would also require a patch to the generate/generate.py file as well. I'll file a bug report on this with those changes. FYI the patches I have: --- generate.py.origSat Jan 22 02:36:17 2000 +++ generate.py Sat Oct 27 17:18:56 2001 @@ -208,7 +208,11 @@ impl.write(', ' + self.tp + 'PyGtk_Type, obj))\n') impl.write('return NULL;\n') funcCall = '%s(PyGtk_Get(obj))-%s' % (objects[name], attrname) - if self.decodeRet(impl, funcCall, retType): + retArgs = None + if type(retType) == type(()): + retArgs = retType[1:] + retType = retType[0] + if self.decodeRet(impl, funcCall, retType, retArgs): return impl.write('}\n\n') # actually write the info to the output files and --- gtklists.defs.orig Thu Sep 20 21:46:45 2001 +++ gtklists.defs Mon Oct 15 13:19:36 2001 @@ -1,6 +1,6 @@ ; -*- scheme -*- (define-object GtkTreeItem (GtkItem) - (fields (GtkWidget subtree))) + (fields ((GtkWidget null-ok) subtree))) (define-func gtk_tree_item_new GtkWidget John Jon Nelson wrote: Where is the bugzilla URL and how do I say this in bugzilla speak: Why does checking to see if a GtkTreeItem is a subtree (with item.subtree != None) result in a segfault? Here is a a version of the _wrap_gtk_tree_item_get_subtree source that fixes my problem. Remove foo and use return PyGtk_New( (GtkObject *) GTK_TREE_ITEM... to simulate the old behavior. PyGtk_New was getting passed a NULL pointer. static PyObject *_wrap_gtk_tree_item_get_subtree(PyObject *self, PyObject *args) { PyObject *obj; GtkObject *foo; if (!PyArg_ParseTuple(args, O!:gtk_tree_item_get_subtree, PyGtk_Type, obj)) return NULL; foo = (GtkObject *) GTK_TREE_ITEM(PyGtk_Get(obj))-subtree; if (foo == NULL) { Py_INCREF(Py_None); return Py_None; } return PyGtk_New(foo); } ___ pygtk mailing list [EMAIL PROTECTED] http://www.daa.com.au/mailman/listinfo/pygtk
Re: [pygtk] pygtk-1.99.5
Hi James, Is there any maintenance effort being done on the old pygtk (0.6.8) source base or is it considered complete? Just wondering because I have encountered some problems that I would report if there is still some development activity. Thanks John ___ pygtk mailing list [EMAIL PROTECTED] http://www.daa.com.au/mailman/listinfo/pygtk