Re: Re: [Etherboot-developers] Q: GPL license clarification.

[Timothy L. <tl...@ro...>]

> Such a clarification seems reasonable to me.

I agree.  If world domination is a goal, bios vendors need to be comfortable that they are not going to accidentally GPL their code.
 
> The GPL does place certain other responsibilities on a vendor wishing 
> to distribute software that has been placed under the GPL. The 
> provision about making the source code available comes to mind. It is 
> not difficult to fulfill, but adds some overhead to using GPL software.

Really, this is the most important point.  If a vendor makes use of etherboot with some modifications to suit their needs and does not make those changes available, there will be problems.  As long as one can download the source and reproduce the compiled eb rom that is being distributed the vendor should be OK.
[snip]

> Now, one could argue that Etherboot, because it can stand alone and run 
> on bare metal is separable, and thus could be inserted into a BIOS chip 
> as a storage medium, and because of the mechanism the BIOS uses to scan 
> for extension ROMs, remains a separate program, and thus does not 
> require the BIOS vendor to distribute the entire BIOS under the GPL.
> 
> Because the BIOS extension mechanism was created to allow vendors to 
> use code that may in fact be unknown to the BIOS manufacturer when the 
> BIOS is produced, as long as they follow the PNP BIOS specification, I 
> think one could reasonably argue that Etherboot fits within the spirit 
> of the GPL with regard to aggregation.

If the vendor's implentation follows this standard it would fit nicely.  Essentially this is what happens when people add eb to the bios or to the NIC's rom.  
 
> In fairness, I must admit to seeing the other side of the argument, 
> that hinges on the phrase "combined so that they become effectively two 
> parts of one program". the BIOS extension method does, at run-time, 
> basically make the two pieces of code virtually a single unit. It's 
> just that when you're at bare-metal, _any_ use of other code would 
> probably look like that because it's the most efficient way to do it.

I disagree.  It has the appearance of a single unit but it is not.  However, one of the things that the vendor's should consider is making it possible to upgrade eb separately from the rest of the bios.  It would help ensure that the appearance reflects reality.

> So, I think we are safe (and reasonable) in our interpretation. If 
> someone makes the argument that being a BIOS extension make us 
> "effectively two parts of one program" at run-time there would be a lot 
> of very unhappy vendors.

And at least one unhappy developer;-)  I for one would like to see etherboot become a standard for netbooting.  Unhappy vendor's are not going to make that happen.
 
> I will put in the standard disclaimer about not being an attorney, and 
> though I believe that  my analysis is reasonable, it should not be, and 
> is not intended to be relied upon in a legal context. One should seek 
> "legally" sanctioned advice in this case. :)

Me too...
 
> Now where was I... oh yes, debugging PXE chaining...
> 
I am getting a laptop without a floppy drive so i may have some call for this...

Tim