Re: Bug#226417: ITP: kernel-patch-2.4-fasttraks150 -- PROMISE FastTrak S150 TX Series Linux Drivers (for Linux 2.4)

[Tore Anderson <tore@linpro.no>]

* Raphael Bossek

 > http://kt.zork.net/kernel-traffic/kt20020429_164_print.html#3
 >
 > This packge I ITP provides the hardware RAID support, the closed source
 > part of the PDC203xx chipset. All people using a hardware RAID depends
 > on this patch, like me. If someone know a functional source for a OSS
 > driver please reply to this.

  I'm under the impression that most cheap (S)ATA RAID controllers
 really aren't, and that all the RAID functionality is really in the
 driver.  The KT entry you're linking seems to indicate that this is
 indeed the case for the FastTrak S150 TX chipset as well.

  Linux has had a generic one-size-fits-all software RAID layer in the
 kernel for a long time, which is probably much more mature[0] and tested
 than any of these proprietary alternatives.  Isn't this a fully adequate
 DFSG-free replacement for this driver?

  You probably won't get on-disk compability with the proprietary driver
 if you're using MD, though.  I don't see that as a problem -- if one is
 already using the proprietary on-disk format, one has obviously been
 able to install the driver without our (Debian's) help.  New users,
 on the other hand, should be encouraged to use the free alternative.

  If I'm right, I'd object to the inclusion of this package as free
 alternatives to its functionality exist, and I strongly believe we
 shouldn't add packages to non-free when these alternatives exist.
 After all, the very intention behind non-free is that it should
 carry the software that's important to our users but for which no
 free alternative exist.

 [0] My own experience with it certainly indicates its quality,
     reliability, and performance to be higher than that of several
     expensive hardware SCSI RAID controllers.

-- 
Tore Anderson