This was addressed in a formal, registered letter to Sun sent on August 10. Sun signed for the letter on August 12, and acknowledged their understanding of the claim in phone calls, yet proceeded to do a full press release on August 22.
I have spent considerable time and money defending the GPL. Generally those violating the license immediately recognize their error and correct the problem. The "Right Thing" happens very frequently, and no one notices that there was ever a problem. Each situation usually takes a two or three hours of my time, and perhaps a follow-up letter from a lawyer. That might not sound like a lot, but the cost in time and money adds up quickly.
This was a porting kit, including explicit instructions that listed the eepro100 and tulip drivers by name, and included code from the Linux kernel (e.g. a modified net_init.c) with the copyright notice removed.
The introductory lines from the Sun web page make their intent clear:
This document describes how to set up a Solaris 8 build machine for porting
a Linux network driver to run in the Solaris kernel. Follow these steps to convert a Linux network driver to a Solaris network
driver.
____
The Linux Network Driver Porting Kit provides a wrapper module for a Linux network driver to run the driver under the Solaris kernel.
----
This is both contributory infringment and active inducment to infringe the copyright and license of the code. Describing in text the principles of converting drivers is perfectly acceptable. Sun provided a script to do part of the job, and cookbook style instructions for the fix-ups that need to be done by hand. This is copyright infringement just the same as if they had distributed the already-modified code. (Note: Nor is distributing a patch file a legal end-run around copyright.)
The Sun press release tried to spin this as an aid to vendors that allowed them to convert their network drivers to Solaris, but the examples (and all of the common Linux network drivers) were written by me and released under the GPL.
Another claim is that this is actually permitted by the GPL. The GPL wording explicitly notes that GPL programs may be run on proprietary operating systems. It does not allow GPL code to be extracted and linked into a non-GPL OS. "Run on" and "incorporated as part of" are distinct concepts.
(To forestall the obvious pedantic "but what about..", Solaris is not a microkernel OS. There is a clear distinction between user-level code using e.g. the defined C library interface, and code linked into the Solaris kernel at run-time.)
Every Ethernet chip you will ever encounter does..
on
Linux Failover?
·
· Score: 2
Every Ethernet chip you are likely to encounter allows temporarily overriding the MAC address in software. Almost every Linux Ethernet driver (all the ones written by me) allows changing the MAC address while the interface is down. The driver rewrites the chip's idea of the MAC address when the interface is brought back up.
This feature isn't needed for all fail-over schemes, but it does exist for those schemes which use it.
Veridicom has two fingerprint scanning products, both with (unoffical) Linux support.
Their USB product is just a scanner. Think of it as a minature flat bed scanner that works on direct finger contact with a postage stamp size chip.
They have their own Linux driver and user-level program that writes PGM files. Looking at their protocol I was able to write a program that worked with Scyld's "Univeral bulk USB driver" in just a few hours, so it works with 2.2, and you don't even need special kernel level support beyond the standard add-on USB package.
The product that is comparable to the announced Sony product is the serial port version, with an internal matcher. One mode of operation is to download a few hundred bytes of encoded fingerprint info. The device returns e.g. "28 of 35 points match", which might be good enough for a gas purchase but not good enough to authorize a major funds transfer.
The claim is that these devices can detect living from dead tissue. I don't doubt that is true in controlled cases, but it's probably mostly PR when the device is set up to scan cold, dry finger and still work with warm, wet digits. Even so, it's easier to just kill someone and take their wallet than to cut off their fingers and leave them alive.
Oh, and when is 9mm thick "credit card sized". It might be "credit card outline", but if all of my credit cards and IDs together are only 5.5mm thick.
I have spent considerable time and money defending the GPL. Generally those violating the license immediately recognize their error and correct the problem. The "Right Thing" happens very frequently, and no one notices that there was ever a problem. Each situation usually takes a two or three hours of my time, and perhaps a follow-up letter from a lawyer. That might not sound like a lot, but the cost in time and money adds up quickly.
Have you ever defended the GPL?
Donald Becker
This is not a "cross-compiler" from Sun.
This was a porting kit, including explicit instructions that listed the eepro100 and tulip drivers by name, and included code from the Linux kernel (e.g. a modified net_init.c) with the copyright notice removed.
The introductory lines from the Sun web page make their intent clear:
This document describes how to set up a Solaris 8 build machine for porting a Linux network driver to run in the Solaris kernel. Follow these steps to convert a Linux network driver to a Solaris network driver.
____
The Linux Network Driver Porting Kit provides a wrapper module for a Linux network driver to run the driver under the Solaris kernel.
----
This is both contributory infringment and active inducment to infringe the copyright and license of the code. Describing in text the principles of converting drivers is perfectly acceptable. Sun provided a script to do part of the job, and cookbook style instructions for the fix-ups that need to be done by hand. This is copyright infringement just the same as if they had distributed the already-modified code. (Note: Nor is distributing a patch file a legal end-run around copyright.)
The Sun press release tried to spin this as an aid to vendors that allowed them to convert their network drivers to Solaris, but the examples (and all of the common Linux network drivers) were written by me and released under the GPL.
Another claim is that this is actually permitted by the GPL. The GPL wording explicitly notes that GPL programs may be run on proprietary operating systems. It does not allow GPL code to be extracted and linked into a non-GPL OS. "Run on" and "incorporated as part of" are distinct concepts.
(To forestall the obvious pedantic "but what about..", Solaris is not a microkernel OS. There is a clear distinction between user-level code using e.g. the defined C library interface, and code linked into the Solaris kernel at run-time.)
This feature isn't needed for all fail-over schemes, but it does exist for those schemes which use it.
Their USB product is just a scanner. Think of it as a minature flat bed scanner that works on direct finger contact with a postage stamp size chip.
They have their own Linux driver and user-level program that writes PGM files. Looking at their protocol I was able to write a program that worked with Scyld's "Univeral bulk USB driver" in just a few hours, so it works with 2.2, and you don't even need special kernel level support beyond the standard add-on USB package.
The product that is comparable to the announced Sony product is the serial port version, with an internal matcher. One mode of operation is to download a few hundred bytes of encoded fingerprint info. The device returns e.g. "28 of 35 points match", which might be good enough for a gas purchase but not good enough to authorize a major funds transfer.
The claim is that these devices can detect living from dead tissue. I don't doubt that is true in controlled cases, but it's probably mostly PR when the device is set up to scan cold, dry finger and still work with warm, wet digits. Even so, it's easier to just kill someone and take their wallet than to cut off their fingers and leave them alive.
Oh, and when is 9mm thick "credit card sized". It might be "credit card outline", but if all of my credit cards and IDs together are only 5.5mm thick.