Owner: Samuel Ortiz <sameo@linux.intel.com>
-- VPNc
-
- Priority: Low
- Complexity: C2
-
-
-- Moving DNS proxy code to ConnMan core
-
- Priority: Medium
- Complexity: C2
-
- Supporting DNS proxy or resolv.conf direct editing seems more than
- plenty as far as resolving is concerned. So the idea is to move the
- dnsproxy plugin code to ConnMan core and have an additional command
- line option in case one would like to stick with the current
- resolver.c code for editing resolv.conf.
-
-
-- WiFi tethering
- Priority: Medium
- Complexity: C4
-
- WiFi tethering should be done through an extended wpa_supplicant
- D-Bus API, as STA and AP modes are typically mutually exclusive.
-
-
- Session API implementation
Priority: High
Priority: Medium
Complexity: C2
- Owner: Lucio Maciel <lucio.maciel@hp.com>
+ Owner: Henri Bragge <henri.bragge@ixonos.com>
The current service provisioning lacks a D-Bus interface for modifying
existing configurations.
credentials, ConnMan should be able to initiate a WiSPR authentication.
-- IPv6 enhancements
+- DNS caching
+
+ Priority: Low
+ Complexity: C4
+
+ A simple initial implementation would see ConnMan's dnsproxy
+ caching the DNS record based on their TTL.
+
+
+- Power management
+
+ Priority: Medium
+ Complexity: C4
+ Owner: Samuel Ortiz <sameo@linux.intel.com>
+
+ Implement a simple device pm hook that ConnMan's core code would
+ use whenever it decides to put devices in power save mode. Although
+ the kernel runtime power management code should take care of that,
+ not all driver (especially WiFi ones) implement runtime PM hooks.
+
+
+- IPv6 gateway handling
+
+ Priority: Medium
+ Complexity: C4
+
+ We should be able to switch between IPv6 only services and thus
+ change the default IPv6 gateway on the fly. For that we need to
+ improve the connection.c code to properly handle IPv6 gateways.
+
+
+- IP ranges allocation and check
Priority: High
+ Complexity: C2
+
+ For both tethering and private networks, but also to detect invalid
+ static IP configurations, we need to have a core IP range layer
+ that manages all currently used IP blocks.
+
+
+- Personal firewall
+
+ Priority: Low
Complexity: C8
- Owner: Jukka Rissanen <jukka.rissanen@nokia.com>
- Support IPv6 only networks so that system can go online even if
- there is no IPv4 address. Also support more than one IPv6 address
- in one device so that the addresses are reported correctly via
- dbus interface. The autoconf IPv6 addresses need also some tweaking
- so that system will go online properly.
+ Extend the iptables code and provide a D-Bus API for personal firewalling.
+
+
+- PACRunner extensions
+
+ Priority: Low
+ Complexity: C4
+
+ Support more URI schemes, support multiple connections, tighter
+ security integration.
+
+
+- Private networks
+
+ Priority: Medium
+ Complexity: C4
+ Owner: Guillaume Zajac <guillaume.zajac@linux.intel.com>
+
+ The private networks D-Bus API should provide applications with a
+ TUN interface linked to a reserved private IP range.
+ oFono DUN forwarding will use a private network for giving DUN
+ clients access to the default service connectivity.
+
WiFi
Priority: Medium
Complexity: C2
- Dependencies: Core:IPv4LL
Owner: Samuel Ortiz <sameo@linux.intel.com>
Complexity: C2
+
Bluetooth
=========
Complexity: C4
-- DUN server
+
+Cellular
+========
+
+- IPv6 and IPv6v4 cellular data connection
+
+ Priority: Medium
+ Complexity: C2
+ Owner: Samuel Ortiz <sameo@linux.intel.com>
+
+ Support IPv6 and dual stack cellular data connections.
+ oFono already supports it and provide an extensive D-Bus API for it.
+
+
+
+VPN
+===
+
+- l2tp support
+
+ Priority: Low
+ Complexity: C2
+ Owner: Mohamed Abbas <mohamed.abbas@intel.com>
+
+
+- pptp support
+
+ Priority: Low
+ Complexity: C2
+ Owner: Mohamed Abbas <mohamed.abbas@intel.com>
+
+
+- IPsec
Priority: Low
Complexity: C4
+
+
+- Split tunnelling
+
+ Priority: Low
+ Complexity: C8
+ Dependencies: Core:Private networks
+
+ The current VPN support puts the VPN interface at the top of the
+ service list, giving VPNs the default route. When doing split
+ tunneling, the system routes packet to the VPN interface for
+ private IPs, while going through the default interface for the rest
+ of the traffic.