Talk:Google Summer of Code 2006
What points are/aren't relevant
Maybe we should remove/reject/improve some of the ideas. For example: the "explorer enhangements" idea look a bit vague to me. Also, I think ReactOS would be helped a lot more with for example full USB 2.0 functionality, sound or DirectX 9, than with a FOSS clone from already existing windows software.
On the other hand, we're helped with everything, and I think aplying students can judge for themselves what ideas are relevant and which aren't.
A few suggestions:
better video driver support
Make it this good, that you simply can run the installer from (for example) nvidia.com, and implement hardware acceleration for a few basic things (the cursor, moving windows, things like that).
If this isn't enough for a project, we could expand the assignment with the following: full OpenGL support, hardware accelerate every 2D thing also accelerated in Windows XP and/or ensure that driver installers from other types of hardware also work.
Linux has already an NTFS driver which enables safe reading and editing of files. Other things like creating files isn't safe yet. However, especially NTFS read support would be very handy, so at least the users would be allowed to view their files or copy them to their ReactOS partition. For this, the linux driver has to be ported. If this would be the case, it would be very handy if it would be made easy to make a new port at a later time, when more functionality can be done safely.
- Develop wrapper driver that loads linux file system drivers (kernel modules). Perhaps for other drivers as well.
Imo coding such a wapper would be more effort to then porting the filesystem code itself and be a waste of resources. --Dr. Fred 16:12, 21 April 2006 (CEST)
- Create a graphical installer based on our livecd, maybe this should be done together with the ReactOS Package Manager
Would make the installer process slower, etc.