Re: A reply on the RTLinux discussion.

[Karim Yaghmour <karim@opersys.com>]

Mark Mielke wrote:
> > yodaiken@fsmlabs.com wrote:
> > > 1. MYTH: "The acceptance of Linux in embedded is being harmed by
> > >     uncertainty over intellectual property"
> > That is your own assesment, which is of course somewhat biased since I
> > don't see you coming out and telling us: "True, the market has a problem
> > with my patent." This standpoint is the only one expected. But given
> > that you are not the only vendor out there, others have testified on this
> > list that indeed Linux has a hard time penetrating in the embedded/rt
> > market precisely because of your patent.
> 
> Sounds like people who can't get customers calling foul to the one
> that can.

Well, again this runs the same way as Victor's other personnal attacks.
Dismissing arguments because of who points them out doesn't really
work.

> > I will leave it to the average LKML reader to decide whether the problems
> > I discribe are "myth" of "fact" in light of all the testimony presented on
> > this list.
> 
> I would label it "seriously exaggerated fact to obtain market share".

As you wish.

> > Also, you omit to explain why the 11,000 developers sampled by the VDC
> > point "real-time limitations" as their #1 show-stopper for using Linux.
> 
> "Real-time limitations" *are* the #1 show-stopper for using Linux. This
> isn't an argument for or against your point.

In light of every person I met in the last 2 years I started working against
this patent and in light of all the testimony seen here, it certainly is.
Read the start of this thread to get the full argument behind this.

> > > And most of embedded Linux use does not require hard real-time.
> > You're trying to play the "niche" trick here, pointing out that your
> > patent is OK because it's a niche market. As I pointed out very
> > early in this thread, embedded/rt is far from being a niche.
> 
> Actually, it looked like he was just presenting a fact. Most people that
> say they need RT, don't, and the few that do, do not tend to mind patents.

These claims have been contridicted a couple of times now by many people
outside myself.

> > > 2. MYTH "The patent license is a terrible burden and terribly vague".
> > FSMLabs cannot be compared in any way to any other open source company:
> > I can rewrite Qt and distribute it under the license I like
> > I can rewrite MySQL and distribute it under the license I like
> > I can't rewrite RTLinux and distribute it under the license I like
> > As for your dismissal of the "motivation behind the act", it does not
> > dismiss any of our arguments, but only reinforces them.
> 
> Most companies with serious RT requirements do not wish to re-write
> and distribute 'as they like'. In fact, the only reason they have to
> do this now, is because the tools are not mature enough.

That's not the point is it? The point is that Victor's patent and its
use to make money from GPL software is unique in the open source
software world. Hence, any time you talk about this issue, it has
to be taken outside the standard open source frame of discussion
because there is no other equivalent case.

> The problem here is obviously open-source companies that prefer to
> avoid paying for software, because they do not sell software, and
> companies that sell software, and do not mind paying for software to
> sell software.

Did you read all the emails stating that clients simply don't use
Linux because of this patent and because of Victor's lack of clarity
on the actual interpretation of how far reaching his patent is?

> The fact is, patents are legal. Live with it, or convince a few
> senators that they should remove 'patent infringement' from the set of
> actions that can be tried in a court of law.

I said it before that I am aware of the issues involved.

> You are not getting anywhere here.

Actually, it is this thread that is not getting anywhere and that
is getting us back to the exact point I summarized earlier in the
following post:
http://marc.theaimsgroup.com/?l=linux-kernel&m=102252033827888&w=2

We should leave it at that.

Karim

===================================================
                 Karim Yaghmour
               karim@opersys.com
      Embedded and Real-Time Linux Expert
===================================================