Editor's Note: The Bottom Line of Hardware
Feb 16, 2007, 23:30 (19 Talkback[s])
(Other stories by Brian Proffitt)
Re-Imagining Linux Platforms to Meet the Needs of Cloud Service Providers
By Brian Proffitt
A story, to illustrate.
An open source developer schedules a meeting with a hardware
vendor. The purpose of the meeting is to request that the vendor
consider releasing the driver support for their products (we'll
call them Widgets) to open source.
The executive is no stranger to the concept of open source, and
is even sympathetic to it. Still, he has concerns. "Why should we,
Widgetmaker, want to do that?" he asks the developer.
Because open source is good and pure and honest, and it will
help make your company a lot of money, the developer replies. He
further requests that a specific product's drivers be opened: the
drivers for WidgetA, an older version of the company's products
that is a bit obsolete for newer versions of Windows, but would run
lickity-split on Linux, especially with native, open drivers.
For the developer, this makes perfect sense, because WidgetA is
an older, legacy product that isn't generating a lot of new sales,
especially compared to WidgetA+, a top-of-the-line device that's
blazingly fast, even on the new Vista. WidgetA, the developer has
reasoned, is much cheaper than WidgetA+, but opening the device up
to a growing Linux market would be just the thing to help
WidgetMaker's bottom line.
Unfortunately, this suggestion gives the executive pause. Here's
What the developer doesn't know is that while both products are
different in terms of performance and features, fundamentally they
both cost (adjusting for inflation) nearly the same to make.
WidgetA costs about $25 to produce, with WidgetA+ costing $30. This
is despite the fact that right now on the market, the older WidgetA
sells for about $50 and WidgetA+ sells for about $300. Clearly,
because of demand, right now WidgetMaker stands to make a whole lot
more money with the newer WidgetA+.
After he explains this to the developer, the developer thinks
for a moment and replies that selling more WidgetA's will still
generate some profit for the hardware vendor, so opening the
drivers will generate more sales from happy Linux users, who will
clearly demonstrate brand loyalty for all of WidgetMaker's
Ah, the executive replies, but there's still a problem with
that. The $50 price tag is just our manufacturer's retail price
(MSRP). Out on some of the discount Web sites, they're selling it
for $30. And because the retailers are taking a small margin of 4-5
percent, WidgetMaker only sees about $3.50 in revenue per WidgetA
sold these days. But, the executive adds before the developer can
respond, that's not counting the costs in packaging and shipping
So, at the end of the day, the executive concludes, sales of
WidgetA actually lose WidgetMaker money. The company is selling
this devices as loss-leaders, providing replacement parts for
people who just need a new WidgetA or hoping that by selling
WidgetAs to new customers with older machines, those customers will
remain loyal to WidgetMaker and upgrade later to WidgetA+ and later
But brand loyalty is what you will get from the Linux users, the
developer explains. They will be so pleased you opened the WidgetA
drivers, they will keep coming back to your products.
Will they?, the executive asks. Why would they? They will have
WidgetA drivers, which would be open, and will run very nicely on
the older machines that Linux thrives on. WidgetA+ would be closed
and would be vastly overpowered for the kind of machines Linux
users typically run. And, the executive added, how long would it be
until you came back and asked us to open WidgetA+?
The meeting eventually ends, without a resolution, with each
side frustrated that the other can't see their point.
If you think this is a true story, you're right. Only the names
of the companies and products have been changed. Yes, that's
plural: companies and products. To my knowledge, a conversation
similar to this one has been played out multiple times in recent
months across a wide range of hardware vendors.
This story illustrates a huge gap between the needs of hardware
vendors and the open source community: a gap that is only growing
wider and is potentially far more serious than the gap between
Linux and software vendors
I wrote of last week.
It comes, quite bluntly, to this: however hardware vendors feel
about the excitement surrounding Linux, there is currently not a
single compelling business argument for them to shift their
products to open source. None. In fact, as the above story pointed
out, creating a Linux market for their products could actually lose
them money--far worse than just breaking even.
The example I used here showed the problem from a pricing
structure standpoint. There are others, not the least of which the
costs each vendor has in research and development and licensing.
That last is particularly telling: many vendors may not actually
own the intellectual property invested in the devices that they
produce and sell. They may have licensed it from someone else. So,
even if they wanted to open driver code, they may not have the
legal right to do so.
This problem is present in almost every case of hardware
drivers. While it is true that some vendors have generously opened
some drivers, the key word there is generous. Opening drivers has
not been a bottom-line decision, but rather one made out of
altruism. For that, we should be grateful. But we cannot expect
such altruism to exist on a very large scale.
The problem is only getting worse: right now hardware vendors
are in cut-throat competition with each other (see, for instance,
coverage of AMD and Intel price cutting). They cannot afford to
give away anything else. In fact, we should all be looking for
belt-tightening across all the silicon-based industry.
Paradoxically, the advent of Vista has given many hardware
vendors a new source of revenue that will make Linux even less
attractive to them. Vista, as we know, is such an inefficient
resource hog, that only high-end hardware will work well in it.
Thus, many users shifting to Vista will need whole new machines or
new devices to make up the performance difference. One has to
wonder if that's what Microsoft has in mind every time it delivers
a new, bigger version of Windows. Keep loyal to us, they may say to
vendors, and we'll get you a steady revenue cycle. Actually, no one
wonders about this; it's pretty obvious.)
But that revenue lock-in with Windows also keeps vendors from
seriously considering Linux support of any kind, open or closed. I
think that the only reason we saw the support we did was because
Vista took so long to come to market. Linux caught a windfall
because some vendors dabbled in it while waiting for Redmond to
release. Now that Vista is here (note, I didn't say "ready"), that
windfall may be long gone.
The solution is not going to be the status quo. We cannot keep
going the way we are. Asking for vendor support out of the goodness
of their hearts is unrealistic and--to be perfectly
blunt--demeaning. Why should the open source community have to wait
for a handout? Why should we have to compromise with proprietary
drivers while we hope the vendor in question will see the light and
open their code?
We are smarter than that. The open source community has created
thousands of brilliant applications, and one excellent operating
system. Technically, we have no peers.
From a business standpoint, however, a new plan has to be made.
We have to somehow make Linux and open source financially
attractive to hardware vendors. Linux can't be a loss, or a
loss-leader, or even a break-even market for them. It needs to be a
profit center. And then, by God, we will have to beat them all off
with a stick, they'll come running so fast.
What will this new plan be? One suggestion I have heard is to
create our own open source hardware companies and bypass the
current hardware vendors. Difficult, but not impossible. I would
imagine that in such a scenario, if there was money being made, the
hardware vendors would want to jump in the Linux sandbox and
There have to be other ideas, business solutions that will make
Linux and open source technology very attractive to hardware
vendors. Now is the time to plan and implement, before vendors are
too captured by the sway of potential Vista dollars. We need to
change the way we handle hardware issues, and not make Linux
completely dependent on the kindness of strangers.
Let's get the vendors to come to Linux with hats in hand.