Kustaa,
thanks for the detailed explanation of copyrights!
Probably such an approach would be good for the court, but there is an
other point of view: moral rights.
We all know that there is a lot more then hex addresses and register
names we are talking about. I think Microchip is not concerned about
include files themselves, but about copying / stealing of the knowhow
(resources, time, money...) they are investing into the development of
pic MCUs. The statement that the include files can be used only with
their own chips is just an additional measure of protection. This
statement is even not written in the .inc files: they wrote it when I
asked them about the licensing status, but I believe we should respect
it anyway. I already explained them that we have scripts to convert
MPASM include files to .c and .h files. This is not a problem for them,
they just want that the statement is valid for the generated files too.
Maybe you are right and legally we don't have to obey them, but this is
their moral right which we have to respect.
This is my personal opinion which can also be changed if somebody shows
(and convinces) me that I'm wrong.
P.S.: and this is also why I wrote that I'm not 100% sure: there is no
big difference if the files are generated from XML database, .inc files
or from documentation...
Borut
On 03/30/2011 08:14 PM, Kustaa Nyholm wrote:
> On 3/30/11 19:05, "Borut Razem"<borut.ra...@gmail.com> wrote:
>> But anyway, the pic .h and .c files, produced from the their
>> XML database wold have the same legal limitations as the current ones,
>> generated from MPASM include files: they can be used only for chips,
>> produced by Microchip. So no GPL.
> So many things going on this paragraph, hard to know where to start.
>
> Copyright is about the end result, not about the process, ie no matter
> what you
> do or how you do it, if the end result is close enough to the original then
> it is a copy.
>
> Another aspect of copyright is originality, which in the context of
> copyright law means 'something that required some creativity to produce',
> ie not every variation of a something is original art in the meaning of
> the act,
> thus not every man made thing is copyrighted. The threshold is low, much
> lower
> than the inventive step requirement in patents for example, but it is
> there.
>
> Thirdly copyright protects the non-utilitarian aspects of the
> art work, it protects the form not function. There is no copyright
> on the shape of a piston or engine block (unless they are exceptionally
> artistic!), there maybe patents or that BMW logo might be trademarked,
> but copyright, I don't think so.
>
> Now lets look at the existing header files (just had a look at
> pic18f2455.h)
> in the light of the above.
>
> Take for example this fragment:
>
> extern __sfr __at (0xF62) SPPDATA;
> typedef union {
> struct {
> unsigned DATA : 8;
> };
> } __SPPDATA_t;
> extern volatile __SPPDATA_t __at (0xF62) SPPDATAbits;
>
>
> The form of this fragment is determined by the C-language standard, so
> there is no originality there. The hex addresses are purely functional,
> so no copyright there. The names SPPDATA might have some originality
> in the choice of the acronym/mnemonic, but since that is really the
> only way to refer to that register, its the name of the thing,
> you cannot separate the function from the form and thus this is not
> copyrighted. I believe no single word or name has ever been considered
> copyrighted.
>
> So I see no way the above fragment can be copyrighted.
>
> How about the comments?
>
> Well there might be something in there if the comments would somehow
> be original but they seem to be very much like any other similar comment
> in hundreds of header files.
>
> No copyright there.
>
> How about copyright as a collection of items that in themselves
> are not copyrightable? Or the order of definitions on those files?
>
> I do not think that would pass the originality threshold.
>
> So in my view there is no copyright on those header files and the
> question of license is moot.
>
> No copyright => no one can enforce any license.
>
>> I think the correct way to produce the GPL .h and .c files is directly
> >from the documentation, as Wes suggested.
>
> The process of how you make a 'copy' makes no difference.
>
>> But even for this approach I'm not sure if it is 100% legal.
> Again, no matter how you produce a copy of a copyrighted work, distribution
> of that copy is the prerogative of the copyright holder and they can
> allow it or forbid it. They can allow it for free or under almost any
> term they want. The permission can be explicit or implicit, as if
> a thing has been in circulation for a long time and they have know about
> it and not acted upon the infringement they may loose the copyright.
>
>
>> Has anybody take a detailed look into the
>> documentation if there is a statement which prohibits such actions?
> At least on the 2455/2550/4455/4550 datasheet there is no such statement,
> it only says "(c)2006 All rights reserved". So it can be argued that
> Microchip
> has created an implied license to use those names and addresses in anyway
> we see fit when they published them, what else purpose there could be
> for publishing them. But I do not believe the names and address are
> copyrighted
> in the first place, so weather they approved or forbid it is moot as they
> have no control over it.
>
> br Kusti
>
------------------------------------------------------------------------------
Create and publish websites with WebMatrix
Use the most popular FREE web apps or write code yourself;
WebMatrix provides all the features you need to develop and
publish your website. http://p.sf.net/sfu/ms-webmatrix-sf
_______________________________________________
Sdcc-user mailing list
Sdcc-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sdcc-user
