[glib/gdbus-merge] GDBus: add TODO item about maybe having to rework ::g-properties-changed
- From: David Zeuthen <davidz src gnome org>
- To: commits-list gnome org
- Cc:
- Subject: [glib/gdbus-merge] GDBus: add TODO item about maybe having to rework ::g-properties-changed
- Date: Fri, 7 May 2010 19:03:35 +0000 (UTC)
commit f1855c2f77fb76347c332b21865e5513f89b15b9
Author: David Zeuthen <davidz redhat com>
Date: Fri May 7 15:02:37 2010 -0400
GDBus: add TODO item about maybe having to rework ::g-properties-changed
gio/gdbusconnection.c | 13 +++++++++++++
1 files changed, 13 insertions(+), 0 deletions(-)
---
diff --git a/gio/gdbusconnection.c b/gio/gdbusconnection.c
index f4b18e3..f09328d 100644
--- a/gio/gdbusconnection.c
+++ b/gio/gdbusconnection.c
@@ -28,6 +28,19 @@
* - Need to rewrite GDBusAuth and rework GDBusAuthMechanism. In particular
* the mechanism VFuncs need to be able to set an error.
*
+ * - The GDBusProxy::g-properties-changed signal currently looks like this
+ *
+ * void user_function (GDBusProxy *proxy,
+ * GHashTable *changed_properties,
+ * gpointer user_data);
+ *
+ * which is problematic because some people frown upon GHashTable
+ * usage in public API (in particular some of the JS people). Maybe we
+ * need to rework it, maybe it doesn't matter since GDBusProxy is
+ * a low-level API and, for C code, we expect code generators to
+ * spit out subclasses that automatically hook up to this signal
+ * and does g_object_notify() anyway? Hmm...
+ *
* - probably want a G_DBUS_NONCE_TCP_TMPDIR environment variable
* to specify where the nonce is stored. This will allow people to use
* G_DBUS_NONCE_TCP_TMPDIR=/mnt/secure.company.server/dbus-nonce-dir
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]