On Thursday 10 Nov 2005 13:00, Andrew de Quincey wrote: > On Wednesday 09 Nov 2005 01:34, Oliver Endriss wrote: > > Andrew de Quincey wrote: > > > On Tuesday 08 Nov 2005 20:35, Oliver Endriss wrote: > > > > Imho the dish support in stv0299.c should be replaced by a generic > > > > implementation in dvb_frontend.c which can be used by all frontends. > > > > It should be easy to replace the stv0299-specific set_voltage stuff > > > > by fe->ops->set_voltage() calls. > > > > > > I seem to remember asking the original author about that when he added > > > it. He said the problem was that it is really timing specific - he had > > > to specifically optimise the stv0299 code so that it met the > > > requirements. I don't have any dish networks hardware myself so I can't > > > confirm or deny this. > > > > Neither do I. > > I had a quick look over that code. Basically it transmits a 9 bit command > byte using 13v/18v levels to represent 0/1. I didn't see any reason why it > wouldn't be possible to move the signalling algorithm itself into the > dvb_frontend core code, where it would just call the appropriate frontend > set_voltage methods. Of course, timing is very important using such a > simple method - so we'll have to see. > > If I have a go at developing that, would someone with suitable dish > networks equipment be able to test a couple of patches? I'm talking absolute rubbish. I didn't spot the generic implementation of the dish networks protocol which is already present in dvb_frontend.c, sorry.
