D&C GLug - Home Page

[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]

Re: [LUG] Software Licences

 

Peter Lloyd-Jones wrote:
> Hi Neil
> 
> I passed your quires on to Paul Bender, one of the software developers (and a 
> Linux user).
>>
>> Peter, can you enlighten me on how the JMRI *software* is actually
>> connected to the circuit boards that pass a signal to the devices
>> (trains) themselves?
> 
> The software actually just talks to a serial port which has a network
> interface device attached to it.  In some cases, the hardware attached
> to the serial port is very simple.  The serial port may actually be
> provided through a USB device and it's associated driver.

Presumably the network is how the messages get passed around the track
to the various trains?

>> Might JMRI have some sort of permission to access the interface from
>> whoever creates the hardware? The JMRI site claims to try to be
>> independent of specific hardware systems so presumably there is more
>> than one? If you purchase one of these hardware interfaces, what
>> software are you expected to use to control it? If you use something
>> like JMRI yourself, does your hardware come with some form of licence
>> or permission to use certain software control methods?
> 
> JMRI actually just talks to the hardware using whatever protocol has
> been established by the manufacturer.  The protocols we use are provided
> at no cost by the hardware manufacturers, and are open in the sense that
> the published specifications are available without any license fee for
> use.  In one or two cases, the open portion of the protocol is only a
> subset of the commands that are available.

Open specs are excellent - all free software and open source code
depends on such standards.

> Now, none of this has much to do with the issue revolving around KAM's
> patent.  The patent in question actually has prevented us from
> developing a viable network TCP/IP server.  It talks about buffering
> data from the clients before sending it on to the device attached to the
> serial port.  This is just vague enough that it seems to cover all
> standard techniques for allowing more than one network client concurrent
> access to a network resource.
> 
> I hope this helps to clarify the situation.

Yes, it does. Thanks. As expected, the patent system has broadened the
original claim to the point where the patent becomes sufficiently vague
as to apply to almost anything. As was explained to the attendees of the
DTI conference hosted by Lord Sainsbury and again at the UKPO workshops
on technical contribution, this is the entire purpose of the patent
claim system. In an effort to prevent circumvention, it creates a patent
system that exists solely to perpetuate enormous legal fees. The claims
become so obtuse and are expressed in such perverse contortions of
legalese that only a court can ever say if a patent has been infringed.
Such decisions are then expressed in confusing and controversial
terminology, so guaranteeing appeals and even higher fees.

Evidence, should it be needed, that entrusting a legal instrument to a
group of lawyers and civil servants who earn commission from both the
claiming and contesting of rival claims only ensures that the system
encourages such contests to be as prolonged and expensive as possible,
whilst doing nothing to make the instrument itself either effective or
useful.

Patents themselves are not wrong in principle - it is the system that
implements and regulates them that is rotten.

-- 

Neil Williams
=============
http://www.data-freedom.org/
http://www.nosoftwarepatents.com/
http://www.linux.codehelp.co.uk/


Attachment: signature.asc
Description: OpenPGP digital signature