On Thu, 24 Jul 2003 07:26:09 -0800, Allen Fear <[EMAIL PROTECTED]> wrote:
>Today Atheros announced that the first open source Linux and FreeBSD >software drivers for 802.11b/g and universal 802.11a/b/g products are now >available as a free download from the Internet. >The 802.11a/b/g Linux driver was created by Open Source developer Sam >Leffler and is intended to be used as a building block for creating fully >featured Linux-based 802.11b/g and 802.11a/b/g products such as network >adapters, access points, and home gateways. The new Linux driver supports >all three generations of Atheros' chipsets for 802.11b/g and universal >802.11a/b/g/ connectivity. >To download a copy of the new Linux device driver please go to: >https://sourceforge.net/projects/madwifi/. >The FreeBSD device driver is a standard part of the FreeBSD operating >system; it is available at: http://www.freebsd.org/. >If you would like to speak with somebody from Atheros about this news, >please feel free to contact me. This morning's release is pasted below for >your convenience. >Regards, >Abigail [snip] I downloaded a copy of madwifi-20030702.tgz from sourceforge.net, and noticed this in madwifi-200030702/README: >The driver depends on two other modules: wlan.o and ath_hal.o. [...] >The HAL is provided in a binary-only form in order to comply with FCC >regulations. I am not a lawyer, so don't take this as legal advice, but I believe that some of this Linux code is illegal. I think the BSD code should be fine, and I think the Linux legal problem should be fixable if the closed source code that actually programs frequencies and power can be moved into a user level program. If there is other code in ath_hal that really needs to be in the kernel, then that may need to be freed if you want a legal Linux driver. Below I've attached a more detailed explanation of my understanding, for those who are interested. If you don't need the details, you can stop reading now. ------------------------------------------------------------------------- Remember: I am not a lawyer, so don't take this as legal advice. I'm just trying to explain my understanding of the situation. I believe that linking proprietary software into the Linux kernel is a violation of the GNU General Public License. Anyone who does does so violates the GNU General Public Licnese and infringes the copyrights of many Linux copyright owners when he or she does anything after that point with Linux that is restrictable by copyright. Linus Torvalds, in a usenet posting, apparently gave some additional permission with respect to his copyrights beyond the GNU General Public License for the purpose of linking in certain types of binary modules, but, to my knowledge, no other Linux copyright holder has, and I have invited other Linux copyright owners to offer such permission in the past. In general, I believe all Linux kernels that contain proprietary software are illegal. I believe this also applies to the case where the proprietary software is a loadable module because there is no exemption in the GNU General Public Licnese when the derivative work is produced RAM that and never written to disk. An imperfect analog that might help in understanding why I think a court would not hold one harmless for just keeping the infringing work in volatile memory would be if one had some files of random bits that, when XOR'ed together in the computer's memory, produced an MPEG of Teminator 3. I also believe that there must be some "glue" code in the provided Atheros Linux drivers whose sole purpose is to support the proprietary ath_hal module working in an infringing kernel, code that has no substantial non-infringing use. Therefore, even if those distributing the code never run Linux, they may be committing contributory copyright infringement. Again, I think this is not a problem for BSD, and I believe the problem in Linux should be fixable by moving the proprietary radio configuration code into a user level helper program. This may require open sourcing some parts of the ath_hal, but presumably not the part that actually programs the radio frequencies and power, which seems to be what the FCC is concerned about (and, by the way, I don't think the requirement that "Manufacturers must take teps that only software that has been approved with a software defined radio can be loaded into such a radio"[1] actually requires closed source, but this message is getting pretty long already). [1] See the URL cited in madwifi-200030702/README: > http://ftp.fcc.gov/Bureaus/Engineering_Technology/Orders/2001/fcc01264.pdf In appendix A, page 2, it defines a new paragraph, 2.932(e). Also see the earlier "discussion" section, paragraphs 30-32. Adam J. Richter __ ______________ 575 Oroville Road [EMAIL PROTECTED] \ / Miplitas, California 95035 +1 408 309-6081 | g g d r a s i l United States of America "Free Software For The Rest Of Us." -- general wireless list, a bawug thing <http://www.bawug.org/> [un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless
