On Sun, Sep 19, 2010 at 1:33 AM, Antti Palosaari <crope@xxxxxx> wrote: > On 09/18/2010 09:23 PM, Bert Haverkamp wrote: >> >> Every couple of months I scan this mailing list for the keywords usb >> and dvb-c, hoping that some new device has shown up that is supported > > Currently there is Anysee E30C Plus and Technotrend CT-3650. About > Technotrend I am not 100% sure, but I have seen patch for adding DVB-C > support for that device. There is many DRX-K devices, but no drivers yet. > Also there is TDA10024 based "Reddo" available in Finland, but I haven't > looked it. Thus only reliable one is Anysee. > the biggest problem with Linux and DVB-C is getting HDTV work. 1. Nearly no distribution comes with codecs (and those who come are doing it in a not legal way - there are some) 2. Compiling fails for the masses (we've got exactly 3 opensource requests within 1 year). 3. graphiccard drivers are a mess (NVidia is doing the best job in getting their graphiccards work). We've been in contact with Trident for more than a year now they even visited our customers in order to help to improve the DVB-C quality of their solution. The main problem lies with earning revenue. If the chip manufacturer releases opensource drivers, all the product manufacturers will release the same which basically means there's no advantage for anyone anymore and this finally will lead to a pricefighting situation. The price will go down enough that it doesn't make any sense anymore to sell the product - definitely go down enough to not spend any money in R&D for software. What can a product manufacturer do to improve this? nothing. And ultimately it will go back to the chip manufacturer because it's not worth for several companies to sell those products. We can just pick an example, Terratec. They used to offer Linux drivers. Finally they laid off all their software engineers. There's a company starting with P. which is now with a company which starts with H. offering 500 EUR/half year for Linux drivers. Since it's possible to drive those ICs from userspace the failure lies in the structure of this entire linux multimedia project, not being flexible enough for product manufacturers that's why we redeveloped everything and see it works. Those people who are into Linux use to scream loud but use to forget about the average people who are relevant for Manufacturers. I would recommend a situation where mainly the bridge datatransfer is in kernelspace configurable from userspace (we also have a kernel opensource bridge driver especially for embedded systems). The advantage of this is a. tuner configuration can be added in a flexible way from userspace at any time, providing backward compatibility to older kernels b. compiling drivers is no option for normal endusers. c. improved stability, since all the complex code is in userspace, the tuner/demod/etc configuration does not depend on speed either. d. bigger userbase, better testing since everyone at any time should be able to use the latest chipconfiguration package from userspace. e. manufacturers would have an easier possibility to provide support for their devices. sure this way is new but that's the way we went during the last 2 years and we pretty much succeeded in the linux area with that. After all it's only about providing good multimedia support to the enduser. Best Regards, Markus -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
