The mobile phone charger is a power conversion circuit which changes your power line voltage (110 or 220V) into something that is useful for your mobile phone (probably 5V). To do this it needs to have some electronic circuity inside which has to be powered and it has to function even if there is not phone around so it can detect one when you connect it.
The charger could be merely a mechanical device like the power socket itself but it would then require all the charging circuity to be inside your phone. Unfortunately it is quite big and relatively heavy so it would be inconvenient to carry it around all the time.
Regarding the actual 30mW figure: if instead of mW you consider the currents involved you arrive at around 300μA (30mW at 100V). This also means a resistance of \$330\,\mathrm{k\Omega}\$. It is quite difficult to work using resistances higher than and currents lower than this while still having to sense the moment when somebody plugs the actual load.
OTOH 30mW is really, really small. The vampire current draw problems are not as important as many believe. If you want a good review of many aspects of this then I suggest reading "Sustainable Energy – without the hot air", especially the chapter on this topic
Basically what you are asking for is called a digital to analog converter, or D/A or DAC for short. In this case you want the full range to be 0-10 volts.
From your description, it appears the receiving end passively pulls up the line, and is expecting the dimmer to put a variable resistance between it and ground. You want to outright control the voltage, but you only need a low side active pulldown to do it. Here is a circuit that will probably work:
The input to this analog circuit is a 0 to 3.3 volt digital PWM signal from a microcontroller. R5, C4, R2, and C2 form a two-pole low pass filter that makes the average value of the PWM signal. Since your frequency requirements are so low, you can easily create such a PWM signal in a microcontroller with plenty of resolution. For example, a 1 kHz PWM signal will have its PWM frequency reduced by nearly 4000 (over 70 dB) by this filter. Even slow micros can give you 8 bits or more resolution at 1 kHz PWM frequency. The micro would adjust its PWM duty cycle in response to commands received via a UART or some other digital interface.
The opamp is used in the classic positive gain configuration, except that since the transistor inverts the signal the opamp inputs are flipped in response. R1 and R4 form the feedback divider, which in this case causes the circuit to have a gain a little over 3. Ideally you want a gain of 3.03, but the values shown give you a little bit at the top of the range where you know the output will go to maximum. The opamp drives the base current of Q1 to whatever it takes to make the desired dimmer line output voltage. R3 is there so that there will be some voltage change in the opamp output with output change. Otherwise, the opamp output would always be at the B-E junction drop above ground, which could lead to instability. You didn't say what the maximum current is that a dimmer has to sink. This circuit can handle well over 100 mA, which is probably high. If so, you can make R3 higher, but the 1 kΩ shown should work anyway.
C3 is there only for stability. You don't need much bandwidth, so there is no harm in overdamping the opamp. Some capacitance here will be needed since even with R3 there, there will actually be a voltage gain less than 1 from the input to the opamp output.
Edit:
The previous circuit accidentally had the opamp inputs flipped. The transistor inverts the voltage, so the opamp inputs have to be opposite from the usual positive gain configuration. The circuit above is now the fixed version.
I have also updated the circuit for the processor running at 3.3 V instead of 5 V and now show the PWM signal from the micro directly.
Best Answer
This term is a hangover from the original standard which was for fluorescent lamps only where it was used to differentiate between the power in the lamp and the power drawn from the supply by the ballast.
DALI started off as an annex to IEC60929 "A.C. supplied electronic ballasts for tubular fluorescent lamps - Performance Requirements" Annex E.4 - Control by Digital Signals. It is now IEC62386.
Subsequently, the standard has added support for many lamp types which do not use arcs to generate light, including low voltage halogen, incandescent and LED, but the term has remained for historical reasons.
Within IEC standards, terms not given in the definitions section should be found in Electropedia.