Quote:
Originally Posted by Wim
So here's the problem.
We had exactly the same back in 2008 for our DL-1x5 chips. So we did development to enable these on Linux and released the driver into the community. This wasn't a complete solution, as there are many variants of Linux and the graphics architecture was not suited to multiple displays in the same way Windows already was. But the information and example drivers released gave all the tools needed to create a driver and a solution on Linux. If there was enough demand for DisplayLink technology on Linux, all the tools were enabled for the community to do it.
Fast forward 5 years, and DisplayLink technology and the graphics architecture issues on Linux still has issues, as this recent post from Plugable shows. It slowly gets better, but it still does not "just work" without bugs and a lot of configuration. It seems the demand on Linux is not there currently, otherwise if it was really desired, the work would have been done to improve this by the community.
As the Linux community hasn't adopted multiple displays or created "easy to use" DisplayLink drivers for the DL-1x5 chips, justifying the work we would have to do to enable DL-3xxx support on Linux is difficult. So if you want to help show there is demand for DL-3xxx Linux drivers, making drivers stable and easy to use for the DL-1x5 family and showing demand for these USB 2.0 products on Linux first would be the catalyst to get DL-3xxx family support on Linux.
Ourselves and Plugable want to see our devices working on Linux. If you want to help develop this support for the DL-1x5 and show that there would also be demand for the DL-3xxx, then Plugable and running a hardware seeding program here:
http://plugable.com/projects/plugabl...amples-program
Wim |
I agree that the driver for the DL-1x5 chips that was released is/was limited to some extent. However, the fact that Fedora 18 has actually made these devices plug and play (Ubuntu's latest release has not yet, I don't think) is actually a good amount of progress, IMO, and I believe others would agree.
As these devices are not a critical system component, it is understandable that development lags behind drivers for other such components (integrated sound and video, for example). We certainly don't need some sort of finished and polished driver that works with every *nix system out there, but it would be ideal to have that base from which others in the open source community can improve upon!
Sadly my field is not in video processing, otherwise I would enroll in Pluggable's program. Even so, it feels like the burden is put on the community to reverse engineer a product when the developers have all the tools needed to get a decent solution out to the community to begin with...