Fast personal printing _without_ CUPS

[Polytropon]

On Sun, 30 Oct 2011 09:48:08 -0400, Jerry wrote:
> On Sun, 30 Oct 2011 13:59:58 +0100
> C. P. Ghost articulated:
> 
> > On Sun, Oct 30, 2011 at 1:13 PM, Jerry <jerry at seibercom.net> wrote:
> > > The biggest loser in this is FreeBSD itself. Virtually any new PC or
> > > laptop, with the exception of the bargain basement brands, and even
> > > some of them are exempt, now come with "N" protocol wireless
> > > devices.
> > 
> > Instead of devoting so much time and energy whining about the
> > problem here on-list, even though you know full well that we can't
> > do anything about it for known reasons... why won't you lobby the
> > manufacturers of "N" devices, so that they either open their specs,
> > so we can write a driver, or at least release binary blobs compatible
> > with FreeBSD? Wouldn't that be more productive? You're very
> > outspoken on some aspects, so put that rhetorical skill to good use
> > and contact the major wireless chipset vendors; and then follow up
> > with them if you don't get the reply you want, just as you do here
> > on-list.
> 
> Seriously, are you so naive that you believe that his is the only
> venue I use to express my feeling on these matters? I have been
> pestering several corporations for over two years now. I have even
> spoken to several of their representatives, including a developer from
> Brother recently in regards to making drivers easily available to
> operating systems other than Microsoft, and usually a few flavors of
> Linux. The contact I had at Brother was actually a Linux user himself.

Actually, Jerry has a point here. The N networking devices
have similarities with "modern" printers in this regards.
While developing compatible "intelligency" in the devices
itself is a cost factor of O(n), moving this "intelligency"
to software is O(1).

For those not familiar with my abuse of the O notation:

O(n) means linear: The more devices are produced, the more
chips need to be made. In case of printers, those chips
control paper feed and ink pee, as well as scanner,
imaging, local buffer storage, data transfer and so on.

O(1) means constant: Only one set of driver will have to
be developed, one for each "Windows" product line and
architecture that's intended to be supported. The whole
"intelligence" is in there, and data transfered to the
device will control it directly, maybe even unbuffered.

>From a business point of view, investing O(1) in development
vs. getting O(n) revenue sounds very interesting.

What I said regarding printer devices seems to apply to
wireless networking too. The cheaper the better. There
is no intention of continued use in there, as this does
not benefit sales. If hardware could be re-used, what
reason would home consumers (main target area!) have
to buy something new that basically provides the same
functionality?

The more unit sales, the lower the price, and therefore
a wider-spread product spectrum. Of course, the downside
is that the possibilities of use are limited, but again,
that's what customers have been trained to require.



> One company, I believe it was Cisco, told me that FreeBSD
> does not support the system calls it needs to make its devices work
> correctly. I am not a system engineer and since he was talking above my
> head I just let it go.

It _may_ be possible that Cisco depends on "Linuxisms"
here, maybe things like *64() calls, like fstat64() vs. fstat().
I'm not a Cisco engineer, so this is just a very wild
guess. "Doesn't have it" may refer to advanced technology
as well as to legacy one.



> As for releasing technical details, etcetera, I was told point blank
> that such information was confidential and would not be released. Now
> that I can at least agree with.

Of course, it is their right to do so, will all the
implications. The confidentiality could also be a means
to hide the fact that devices come with planned
obsolescence or are intended to spy at users (such
as it is quite easily possible with "Windows" and
a webcam). Other reasons could be secret contracts
with companies or governments for a "data exchange",
you're getting the idea. But as this cannot be proven
properly at the moment, just leave this point mentioned
as is.



> Unlike many socialists, I don't believe
> in working my ass off, spending X amount of dollars and then just giving
> my work away freely to every dirt bag to clone.

If this is not your attitude, well, fine, and fully
okay. However this is not everyones attitude as
some want to improve computers and operating systems
for free, as they see it a chance to do something FOR
the society.

The possibility to "make money" with tools provided
for free is a thing of licensing. You know that FreeBSD
allows its users to create own products with it, even
turn _them_ into something proprietary and then sell
them. This is a good idea from a CAPITALIST point of
view, i. e. take it for 0, sell it for $$$. And why
not? Because the licensing terms don't prohibit it.
This is also a chance for innovation, for individuals
finding their future on a free market.

If this way of working has a moral downside is a
consideration to be taken by every individual on
his own.



> I write several major vendors on a monthly basic. Sometimes even using
> different names so they might falsely believe that there is a larger
> base than actually exists to request support. Now, suppose you were to
> join me. Perhaps a few thousand other users, in other words all the
> FreeBSD base, and wrote on a bi-weekly schedule to a targeted vendor
> base requesting support. I will be happy to supply my own personal list
> and compile other pertinent vendor's names & address's as well.

This woudn't be interesting, as FreeBSD has no market
share. Why "no"? Because it's not a company like the
big software vendors or the OEMs. No market share - no
intention to support. It is that simple. The logical
conclusion: The majority of free projects, no matter
how advanced they are, get no attention as the
predicted unit sales won't justify further activity.



> The only problem I see with this approach is maintaining continued group
> support. The tendency of people to just give up and quite is self
> evident.

If there's a backup for professional work (i. e. payment),
_this_ is what keeps people at work. Only few exceptions
have the time and will to code for free. Their revenue is
the knowledge they gain, the popularity of their software,
and the idea that they actually have improved something.
Of course, this is not what you can buy food for.



> It is the primary difference between an Alpha male and one who just
> bends over and takes it.

As I said: Adoption is the strength of the weak. :-)





-- 
Polytropon
Magdeburg, Germany
Happy FreeBSD user since 4.0
Andra moi ennepe, Mousa, ...