On Sunday 16 November 2003 16:38, Omer Zak wrote: > On Sun, 16 Nov 2003, Gilad Ben-Yossef wrote: > > On Sunday 16 November 2003 12:51, Omer Zak wrote: > > [... snipped ...] > > > > 2. If Linus allows use of closed code Kernel modules with Linux and > > > they probably are obfuscation champions of the Linux world, then Ilan > > > should be entitled to ask about the subject. > > > > Linus allows no such thing. He simply noted that the fact that your > > module is linked against a Linux kernel does not automatically make it > > derived work, and therefore whatever license the Linux kernel is under is > > irrevelavnt. Here are his exact words: > > > > "I'm a complete non-believer in binary modules", Torvalds said, > > reiterating what has become the conventional wisdom of the Linux-kernel > > mailing list and experienced Linux support people. "Most houses that use > > Linux a lot say that they won't support binary modules because they > > can't. They may work, but you're not getting the full advantage of > > Linux", he added. On the legal side, there is no specific exception for > > binary-only modules. "They're borderline legal. There's nothing in the > > license that says you're excused from the GPL". > > > > http://www.linuxjournal.com//article.php?sid=6152 > > My interpretation of the above is that Linus had the power to decide > either way and the law would have supported his choice. He chose the > liberal interpretation, and accepted into the official Kernel patches only > from people, who agreed with his (liberal) interpretation. So it is now > the law of the land. > > What if he had decided otherwise? Did he really have the legal power to > decide otherwise? Those are questions worthy of argument by themselves. > Is there anyone in Linux-IL, who has the answers or at least the > imagination to speculate in interesting ways? >
Allowing binary only modules has its advantages as at list some hardware companies have reasons to hide the driver code since it reveals information about their hardware (although as was mantioned, this can probably still be reverese engineered and considering its probably the competition that is most interested in this information they should have the budget to do it). This approach opens up the chance for more hardware support for linux even in cases where the companies are reluctant to reveal the source so in general has its benefits. > > > 3. And, as Ilan said, the Linux-IL mailing list is Linux-IL, not GNU-IL > > > (or GPL-IL), RMS's insistence upon the term GNU/Linux withstanding. > > > > Well, in case anyone cares, I think the question was on topic. I just > > think it was stupid. > > I don't think the question was stupid. If anyone earned the 'stupid' > label in this story it is those managers, who caused the need for the > question in the first place. > > This is because I learned some things from the technical answer to the > question. > --- Omer I second that. Although I don't currently have the use for stopping reverse engineering of anything I write (currently mostly matlab code so its in plain text anyway ;-) the technical answear was interesting. > My opinions, as expressed in this E-mail message, are mine alone. > They do not represent the official policy of any organization with which > I may be affiliated in any way. > WARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html > > > ================================================================= > To unsubscribe, send mail to [EMAIL PROTECTED] with > the word "unsubscribe" in the message body, e.g., run the command > echo unsubscribe | mail [EMAIL PROTECTED] -- Micha Feigin [EMAIL PROTECTED] ================================================================= To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
