Johannes Stezenbach wrote: > Klaus Schmidinger wrote: > >>Johannes Stezenbach wrote: >> >>>Since you Cc'd Klaus: VDR already contains a CI library, and I guess >>>he's not eager to include another one. It would be nice if you >>>could point out to him what needs to be changed in VDR to support >>>Twinhan-style CI, or maybe even send him a patch. But you >>>better sort that out on the vdr mailing list. >> >>I wouldn't like to implement a special method of CI handling just >>for one particular hardware. You should make the interface behave >>just like the existing one - then you can use VDR right out of the box. > > > The problem is that the Twinhan cards hide access to CI behind > a microcontroller, and the protocol for that uC doesn't allow > the implementation of the link layer protocol which is used > by other cards :-( > IIRC there was some discussion with Ralph a while ago about this. > True, down to every word.. > What Manu implements is (I hope) not a Twinhan-specific It is not a Twinhan specific protocol that i implement, (the discussion ended not to have a proprietory protocol, but one that could be reused.. Ralph spent so much time on laying out the specifications..), but a high level protocol, as we discussed (suggestions from Ralph, as Johannes said) on the list a while back, and it stands good in all those discussed aspects. In fact i have it a bit simplified also, such that all hardware specific stuff is hidden in the driver. Currently the driver in it's very experimental state accepts the information in the EN50221 format exactly (ie, the struct ca_msg itself is used to transfer the info, rather than specific ioctl's) without any modifications. This interface could be used for any hardware employing similar high-level communication.. Thereby making tasks a lot simple.. > protocol, but a generic application layer protocol that > could be used for similarly crippled hardware. > Yes, it could be there for similar hardware, and i fell that in the future, there will be more of such cards, and it variants compared to standard cards, the very well reason being the cost factor itself. > It's up to you to decide if you want VDR to support that. > > Meanwhile I suggest Manu explains to interested people I will do that.. but first of all, i think i should understand the VDR code a bit.. I hope it turns out positive. There are many people out there looking for an interface to an existing application. > on the vdr list what would be necessary to support I am not subscribed to the VDR list, but would do that, and discuss there such that people have a better idea.. > Twinhan-CA, if someone is interested they can write > a patch. > I could write out some sample driver usage style also, which would be quite helpful, but not in CPP though. Manu
