I apologize if this seems like a silly suggestion, but given a project like ReactOS, would it be feasible, or even possible, to take a portion or subset of this project to create a compatibility layer for windows drivers? I cannot even imagine how successful an open source operating system would be that had such a feature.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Sorry for the late reply. Most of Genode-related discussion tends to happen on the mailing list. Maybe, you'd like to join it? See http://genode.org/community/mailing-lists
Your suggestion is not silly at all. But to be honest, I think that the situation of device drivers looks pretty good on Linux, too. So to me, porting drivers from Linux looks more attractive than reusing binary drivers. There are two reasons for this stance.
First, binary drivers come with different licensing terms so that it is not clear under which conditions they may be distributed. Those conditions may differ from vendor to vendor. When porting drivers from an Open-Source kernel, in particular Linux, the terms are uniform and clear.
Second, reimplementing the Windows device driver API looks like a heroic task when considering complex devices, in particular GPUs. But those are the only devices that would really benefit from such a possibility. Drivers for simpler devices (i.e., network cards, disk drivers etc.) are not so interesting because there is good driver support in Open-Source OSes already.
Which particular device drivers do you have in mind?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I am mostly concerned with graphics drivers. I have been unable to use any type of open source operating systems despite doing a lot of coding work because of the need for a reliable GUI and GPU drivers (for games, movies, and webbrowsers).
I don't have experience in this though, so I really don't understand the amount of work involved, but if there was an operating system that was able to launch valve's new linux version of steam with windows grade opengl/directx hardware support would be pretty significant.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thanks for clearing up your incentive behind the question. From the insights I got so far, GPU drivers are insanely complex. So I think that the undertaking of creating an environment for executing such binary drivers in a reliable way would be a tremendous effort. I would certainly not take it on. Out of curiosity, has ReactOS succeeded in using binary GPU drivers?
From following the announcements at Phoronix, I gathered that Value has become very active in Linux driver development. They seem to collaborate directly with GPU vendors to bring Linux driver support on par with Windows. So the situation on Linux seems to get better. For Genode though, I think that your goal is out of reach for now.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I apologize if this seems like a silly suggestion, but given a project like ReactOS, would it be feasible, or even possible, to take a portion or subset of this project to create a compatibility layer for windows drivers? I cannot even imagine how successful an open source operating system would be that had such a feature.
Sorry for the late reply. Most of Genode-related discussion tends to happen on the mailing list. Maybe, you'd like to join it? See http://genode.org/community/mailing-lists
Your suggestion is not silly at all. But to be honest, I think that the situation of device drivers looks pretty good on Linux, too. So to me, porting drivers from Linux looks more attractive than reusing binary drivers. There are two reasons for this stance.
First, binary drivers come with different licensing terms so that it is not clear under which conditions they may be distributed. Those conditions may differ from vendor to vendor. When porting drivers from an Open-Source kernel, in particular Linux, the terms are uniform and clear.
Second, reimplementing the Windows device driver API looks like a heroic task when considering complex devices, in particular GPUs. But those are the only devices that would really benefit from such a possibility. Drivers for simpler devices (i.e., network cards, disk drivers etc.) are not so interesting because there is good driver support in Open-Source OSes already.
Which particular device drivers do you have in mind?
I am mostly concerned with graphics drivers. I have been unable to use any type of open source operating systems despite doing a lot of coding work because of the need for a reliable GUI and GPU drivers (for games, movies, and webbrowsers).
I don't have experience in this though, so I really don't understand the amount of work involved, but if there was an operating system that was able to launch valve's new linux version of steam with windows grade opengl/directx hardware support would be pretty significant.
Thanks for clearing up your incentive behind the question. From the insights I got so far, GPU drivers are insanely complex. So I think that the undertaking of creating an environment for executing such binary drivers in a reliable way would be a tremendous effort. I would certainly not take it on. Out of curiosity, has ReactOS succeeded in using binary GPU drivers?
From following the announcements at Phoronix, I gathered that Value has become very active in Linux driver development. They seem to collaborate directly with GPU vendors to bring Linux driver support on par with Windows. So the situation on Linux seems to get better. For Genode though, I think that your goal is out of reach for now.