Subject: [coallesced replies] Re: how to create 21,780 new free software jobs (2,530 in R&D)
From: Tom Lord <>
Date: Mon, 4 Nov 2002 21:42:31 -0800 (PST)


	Sounds great! Where do I sign up!? :)

With your neighborhood VC, aol, ibm, sun, hp/c, dell, etc.  One reader
has already suggested he will take it in such directions.


	Does that come with free cheese and beer?

Only if you have a history of hanging out at the alehouse without
being an asshole and with leaving huge tips when you have money.
Think of it as reaping the benefits of both sides of a sliding-scale
payment system.  It's an informal system, though -- you have to be a
decent/interesting human being in the eyes of the staff to pull it
off.  Anyway, the beer can be comped, but the cheese is strictly cost
controlled (as might be expected, if you think about it).

More seriously, as I put it to one editor:


	>> And $80K - $100K/year is big bucks for a network engineer in most 
	>> non-silly valley parts of the world. :)

	> I know.  $190K is a pretty reasonable salary for a small
	> family here in the valley.  I think it's a good level for a
	> "standard professional salary" -- in several places, for
	> example, blue collar workers have had to resort to scamming
	> overtime systems to get up to that range.  Meanwhile, last
	> year I think, I saw job listings for free software hackers
	> in Boston offering something like $55K.  I emphasized the
	> light schedule, 4-week vacation, and high salary to
	> encourage readers who are hackers to not sell themselves
	> short, and to discourage readers who run businesses from
	> creating still more jobs that contribute to a permanent
	> underclass.  (You can quote me on that.)

I would add here, considering FSB RISKS I've mentioned previously,
that the surest insurance against employees turning to the dark side
is treating them with impeccable professional respect.   These are
very high-responsibility positions I've proposed creating.


	[missing elements:]

	1) Anything about how you *get to* 60k people (or at least,
	what are the points on the curve to getting there that are
	viable -- note that the one obvious point is "you can't afford
	your *first* support engineer until you already have 1714

As I said (borrowing from Shivers): "Piece of cake.  Probably only
take a week or two."  An exercise for the reader.  It's the technical
obstacles you've mentioned that led me broadcast this plan rather than
try to turn it into a VC proposal that I hoarded all to myself.  This
would be a lot easier, imo, at this point in history, if there were
technical oversite of Gnome that were focused on producing a light,
tight system with 1984-mac-quality usability and stability.

Broadband helps.  If I can't get out on a thursday night (or if
there's no place to go in my town), at least I can pay a quarter to
catch the show at DNA Lounge.  It's also worth pointing towards
employers like, what was it - GM?, who give all their employees home
computers.  LAN parties are all the rage in East Butfook, these days.


	2) Universities have two classes of users -- well supported faculty,
	and "everybody else" who are usually in the "you're bright, *you*
	figure it out" space.  

Not at CMU, at least when I was there.  The touch-stone of a lot of
technical decisions was how well we were doing with the non-technical
undergrad and faculty population.  "How are the fine-arts geeks going
to cope with this?"  was the cliched way of putting it.  My employers
at that time were way cool in many ways.  CSW ("computing skill
workshop") was (at least back then) an amazing accomplishment.


	On top of that, they're *in one place*, and so can be helped a
	lot more directly by limited resources, or by a neighbor.  You
	also have a near-monopoly position.

There's nothing to prevent the division of users into 60k populations
along geographic and economic-region boundaries.  I think that's a
good way to do it, personally.


	So I'm not sure this business model has any reality under it...

Isn't that what entrepreneurial risk is well-suited to explore?  Are
you sure it doesn't have economic reality?


	I'm not saying there is no market, but I think you
	overestimated willing customers.

	Most people don't pay $132 a year to keep Windows up to date.

Yes, that's one way to characterize the risk.   I don't know you're
wrong.  I don't think you're right.

I think computing can be sold to "the masses" as part of a basic
service (me and MSFT both), and that this is a reasonable plan for
keeping prices steady even while costs fall.


	(I'd even venture to say that most people running some version
	of RedHat are not paying $60 a year for a RHN subscription.)

Almost half-way there.  Can we grab a little off their DSL bill?
cable bill?  cell-phone bill?

Also, the plan, with minor alterations, applies to offices as well as
home users.


	Apparently ISPs can survive getting about $10 a month per subscriber.
	(That supports the organization, facilities, bandwidth, and giving
	bad technical support.)

	Certainly your $11 and services must be on top of that $10, not
	instead of it.

Nope.  Bandwidth is tending strongly towards "free".  You pay too much
for basic cable and basic phone service already, as nearly as I can
tell.  How often do most people make (non-revenue-creating) support
calls to their telcom provider?


	"Good support" seemed to be the selling point for your service.
	I think you underestimated support costs.  

How, specifically?  I think I put forward a plan that, resource-wise,
soundly beats the kinds of support we currently see.   On the other
hand, GNU/Linux distributions are currently so crappy in some critical
ways, that there's some software architecture concerns that need to be
addressed to make the proposal work -- hence (part of) the cry for R&D
bootstrapping investment.


	  Users (and user populations) have very diverse needs, which
	  are satisfied by using many different software packages.  You
	  can never offer "all your binaries."  You must fence off a
	  subset of what you will support, which is another way of
	  saying you divided your market/created a new distro.  That's
	  OK, but it isn't a new business model.

You may or may not have a point here.   You'd have to be more
specific.  Can you give us back-of-an-envelope numbers?