Messages in this thread | | | Date | Sun, 3 Sep 2000 15:02:12 -0700 (PDT) | From | Andre Hedrick <> | Subject | Re: [ANNOUNCE] Withdrawl of Open Source NDS Project/NTFS/M2FS forLinux |
| |
On Sun, 3 Sep 2000, Alan Cox wrote:
> > Linux rejected the code because it does not understand nor does anyone > > have the desire to learn what it does. Since it is not in the kernel > > there is no GPL issue. Upon Microsoft's adpotion of the model they will > > If its your code there isnt anyway. You can license it to multiple people > in general. And why not.
Same point just clearly stated.
> > The things lost and stagnating development is that I can not do new > > features without the code. Some of the newest features in Linux like > > auto_crc_downgrade is being adopted by Intel and a future hardware IO > > access that will allow for sideways disk access will convert the software > > into hardware. > > You don't play with the mainstream IDE code while you are stabilizing a release > you do it for 2.5 - sometimes things just miss deadlines.
I agree, but I can not stabilize the release without portions of the code because of the onion effect. You know this all to well. Once I peel back the parts to fix the bottom-low-level bug you can not put the old layer back in place because of the gaps and the hole punch throughs.
This it will have to wait for 2.5, but everyone needs to get off the issue that it is a filter and understand that it is a command completion pre-handler. I hope that you finally understand the point and we do not have to fight again, next will be to take LT to school about the issue.
I hope by that time I will have a newer model that will force the enduser to do the entire setup with a task-struct that includes a table for setting up table io rules. Then everyone will have to read the 404 page doc to understand how to apply the taskfile. In short if you can not setup the taskfile in user-space and define the needed return, you have no business calling the IOCTL; however, I will allow all BAD-ROOTS to ram shit down the pipes and only certify programs that are known to do things correctly. This will allow OEMs to keep their secrets and remove Linux from needed to know or allow binary modules to load in the driver that are not native.
Cheers,
Andre Hedrick The Linux ATA/IDE guy
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
| |