Re: [Call For Wartectomy] CRLF conversion out of kernel

Albert D. Cahalan (acahalan@cs.uml.edu)
Wed, 14 Jul 1999 01:37:21 -0400 (EDT)


Alexander Viro writes:

> JarJar Binks^W^WSupport for CRLF<->LF translation in the kernel
> must die. Unless somebody has damn good reasons for preserving this
> bogosity I'm going to exterminate the bloody thing. It is guilty in:
> a) breaking mmap() semantics.
> b) breaking lseek() semantics.
> c) breaking write() semantics.
> d) being bloody slow, painful and kludgy.
> e) making FAT support with the new page cache hard.
> f) confusing the hell of unaware victims^Wusers.
> g) bringing the list of magical filename extensions into the kernel.
> h) belonging to userland *and* being already implemented there.
> If somebody has really convincing arguments for preserving the sucker -
> tell. Otherwise it will die.

Obvious reason: compatibility.

Files with CRLF<->LF translation enabled have many of the same problems
found in the /proc filesystem. These files are _not_ regular files.

I suggest a new type code. Then you can kill mmap() and other features.
That leaves a file type that is more like a socket, char device, fifo...
Perhaps "less" will even work right, unlike now.

Type code 0x3000 (030000) has been free since v7, where it was used to
indicate multiplexed char devices.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/