August 1st, 1995




Bureau of Competition Policy
Industry Canada
50 Victoria Street, 18th Floor
Hull, Quebec K1A 0C9

fax: 819-953-8546

Attn:  Laura Joyce, Cal Gundy

Dear Laura and Cal:

Please excuse my tardiness in getting this information to you.  As 
you're no doubt aware, one of the perils of small business is the 
invocation of the "headless chicken dance", evident around here much 
of the time.

Answering your "Questions for Calgary ISPs" is a nontrivial task.
You've posed questions that, for the most part, have either a large
variation in possible answers, or require broad speculation, as we
are by no means "industry analysts", rather independent businessmen
focussed by necessity on the minutae of our own operations and survival.

That having been said, I'll attempt to tackle your queries, one by one.

1.	The six signatories to the statutory declaration represent the
	majority of commercial Internet dialup access in this city.  As
	previously indicated, you'll need to poll them individually for
	details on their business (eg date of entry into market, etc.).
	Not appearing are the following:

		The Network Center:	

		Run by an engineer of my professional acquaintance named 
		Jack Smith.  Jack is providing full time and ISDN access
		in addition to dialup service.  He has also lined up some
		local computer/software retailers, notably Soft Options,
		to storefront for him and promote his service.  I approached
		him in the course of marshalling support for this action,
		and he declined, confident, based on his many years of 
		experience within the telecommunications industry, that
		left to their own devices, AGT would screw this up on their
		own, and wouldn't require our intervention to do so.

		Cybersurf Internet Access:

		Cybersurf is the true anomaly in the Calgary ISP business.
		They don't provide TCP/IP (SLIP or PPP) access.  Instead,
		users must use their proprietary software, based upon some
		sort of "remote Windows display" scheme, to gain Internet
		access through their system, which also, I believe, limits
		them to serving PC/Windows users.  Coming from a system 
		design background, I personally feel that this is a foolish, 
		misguided way to provide the service, but they promote their
		method as far easier to set up and use than SLIP/PPP.
		They also make some very extravagent claims, which I feel
		are at least deceptive and possibly fraudulent, regarding
		the "speed of access to the Internet" that they provide.
		I approached Gary Gardner, president of the company,
		while gathering signatories, and he too declined.  He was
		apparently under the (I believe) rather silly impression 
		that he was on his way to selling his nonstandard protocol
		to AGT for use in their service, thus lightening their
		user support load.  It probably didn't occur to him to take
		a look at their tariff application, which specifically
		stipulates TCP/IP traffic.

		Calgary Free-Net Association:

		Like other freenets, this group provides extremely limited,
		text-only access to the Internet.  There is no love lost
		between the Calgary Freenet and AGT, reportedly stemming
		from a promise made by AGT to provide Centrex service to
		them, then refusing to do so, sticking Freenet with full-
		price business lines.  The upshot of this is that Freenet,
		with a fixed amount budgeted for phone lines (based on the
		Centrex rate) was then only able to install half the number
		of lines they had counted on, making it extremely hard for
		their thousands of users to gain access to the service.
		Shawn Henry, the Freenet's director, would have been very
		happy to have signed the statutory declaration, but was
		politically prohibited, since AGT is (ironically) a donor.

	There is growing assortment of minor players in the market, some
	running small dialup pools, and some providing only World Wide Web 
	(WWW) services.  But that's the picture with respect to the "major"
	suppliers at present in this market.  To the best of my knowledge,
	none of the companies currently on the scene provided Internet
	service prior to early 1994.  Aside from the obvious (AOL, Microsoft,
	the cablecos, etc.), it's very difficult to speculate on who might
	enter the market in the next two years.

2.	Prodigy, Compuserve, and AOL are at present, not much of a threat.
	Notwithstanding their historical predominance in the field of online
	services, consumers want Internet access, and Internet as provided
	by those companies is still limited and relatively expensive.  We
	(ISPs, that is) seem to be taking from them a steady stream of users.
	Many of the converts are people who signed on with those services simply
	because software was included in a modem purchase, and only after a
	month or two realized how much more expensive it was to get a more
	limited service.

	The Microsoft network is, of course, quite another issue.  As you're
	probably aware, there are movements in the US, spearheaded by the
	Justice Department monopoly-busters, Ralph Nader, and others, to
	prevent Microsoft from bundling software for their online service
	with Windows 95.  I have no doubt that ISPs would be damaged by the
	inevitable lemming-like migration of vast numbers of computer users
	to Windows 95 should it include the native Microsoft network access.
	Obviously, that's a fight already being waged by capable people, and
	one somewhat outside of our realm - at least for now.  See my comments
	below on their apparently pending Calgary service.

3.	This question is somewhat like "Describe selecting, reading, and
	understanding a book".  Setting up an ISP is either a cheap, trivial
	exercise, or an expensive, difficult one.  It all depends on who is
	doing it, and how they chose to proceed.

	The best thing to do, then, is present Canada Connect as a case
	history.  Constructing a service such as this can be accomplished
	on the cheap, in a basement with used PCs and lax network standards -
	or it can be done with shiny new workstations, the poshest modems, 
	and high-priced digs downtown.  We've struck a very good compromise 
	between the two, finding what we feel to be a near-optimum balance 
	of high quality gear and low capital and operating costs.

	First, why would anyone want to jump into a market such as this,
	which, given all the present indications, is going to be a bloodbath?
	Quite simply, when we conceived of this company in February/March of
	1994, it was because commercial dialup Internet service was 
	unavailable in Calgary.  I had been operating an old Silicon Graphics
	workstation at home - the machine I learned Unix on - and had a
	'net connection courtesy a friend at the University of Calgary.  It
	was through a scheme called UUCP (Unix to Unix Copy), which allows
	mail and news traffic only.  UUCP is a batch system, which means 
	that one Unix system will call another on the phone when there's
	data destined for the "callee", do the transfer, and hang up.  This
	does not allow real-time use, such as is required for FTP, Web
	browsing, and most other desirable Internet resources.  At the time,
	there were very few computers outside of the University and a few
	local technology companies with 'net connections, and my machine
	was probably among no more than a dozen such systems operated by
	people in their houses.  Aside from the limitations of UUCP, though,
	severe mismanagement of the network, mainly between the University
	of Alberta and the Alberta Research Council in Edmonton, made the
	connection I had virtually unusable.  I had just left a Calgary
	company that I had started up and taken public (on the Alberta
	Stock Exchange), and armed with a modest sum of money from my
	stock I decided to start Canada Connect.  Net-hype hadn't yet hit,
	and I was doing it as much to get myself a good connection as for
	any other reason.

	The next few months were occupied by research, travel (scouting
	appropriate used hardware in Silicon Valley), and money-gathering.
	By early June I'd moved into a small office center, bartering rent
	for access to my network.  The installation, at that point, consisted
	of a four-year-old Silicon Graphics Unix workstation, originally 
	worth about $125,000 , but picked up used and on the cheap through
	connections in California.   The remainder of the summer and early
	fall were spent bootstrapping the system up - getting a low-speed
	dialup network connection from UUnet in Toronto, and establishing
	the basic technical services for operation of the system.  I was
	joined "on spec" by a longtime friend in his early 60s, a casualty
	of free trade.  We contracted a local programmer to do the "tough
	guy" network setup and administration.  In the late fall we installed
	our high-speed network connection (from a company called Insinc,
	now owned by Sprint), and after a couple of months of test and
	debug operation with a few dozen users - friends and people anxious
	for access who had learned of us by word of mouth - we officially
	opened for business (started billing, that is) on December 1st.

	Since that time we've signed on a steady stream of new users,
	performed experiments in marketting the Internet - a "new product"
	that the vast majority of the population still fundamentally fails
	to understand, developed our company's persona through agressive,
	unusual advertising, and witnessed an unforseen flock of competitors
	appear on the local scene, some dumping the service at below cost
	into the market in an effort to achieve predominance and market
	share.  In so doing, we have brought the company into the black in
	under a year.  This has been accomplished through extreme creativity
	in almost every aspect of our operation, and the not-so-creative
	move of not paying our staff much.  At present, of the twelve people
	involved in the operation of the company, eight receive no regular
	compensation.  Two are paid as contractors, and two receive regular 
	monthly pay - the secretary/receptionist ($1100/month), and me 
	($600/month).  The company is being built on goodwill and trust,
	and the belief that concerted, unselfish effort will not only lead
	to paycheques before the end of the year, but to a stock exchange
	early next year.

4.	With lots of money, no time at all.  A company of modest means, 
	such as this, would have a damn tough time just surviving the startup.
	I certainly would not attempt to commence at this time the process
	just described above.  There's just too much competition already
	among the established small players, to say nothing of the result of
	big business (such as AGT) getting into the game.  A year ago this
	was a risky move.  Now it would be unacceptably risky.

5.	As this request has been extended to all the signatories, I am
	only providing information on ours:

	a) Canada Connect Corporation (provincially incorporated).

	b) A copy of our current price list is attached.  Please note that
	   our pricing is still in flux.  For example, here is one of the
	   problems we've been faced with for some time:  What is our annual
	   rate for a regular dialup account?  We can't quote one.  We charge
	   $30/month.  That translates, obviously, to $360/year, and it would
	   seem reasonable to offer some sort of discount for a year's pre-
	   payment.  Our most antagonistic competitor, though, charges 
	   $129/year (they have no monthly rate).  We know they're losing
	   money at that rate, but it means that we can't quote an annual
	   discount significant enough to appear competitive and still make
	   money.  So we have not published an annual rate, instead signing
	   up customers at the monthly rate, and only accepting a year's
	   prepayment at their suggestion, presumably after they've satisfied
	   themselves that our superior service is worth the "premium".  On
	   a few occasions these customers have asked for a modest discount,
	   such as 10%, and we have happily given it.  

	c) I'm an engineer and kinda stunned in the accounting department,
	   but our income statement apparently says that we've grossed about
	   $83,000 since the beginning of the year.  If you wish, we should
	   be able to chart the growth for you - I think we're moving about
	   $15,000/month through the company right now.

	   At present we have about 600 users on the system, including 30
	   or 40 freebie accounts (friends and promos).  We'll be doing a
	   culling of the freebies soon and paring it down to "good friends"
	   and the truly deserving.  These users are being serviced by 48
	   dialup modems, which is significantly fewer than I'm comfortable
	   with.  The experience to date, as we've increased our modem pool
	   from an initial 16 to 32 and finally 48, is that we shouldn't be
	   running more than a (roughly) 10:1 user-to-modem ratio.  That's
	   the point at which people start to get busy signals during peak
	   (evening) hours.  We've been stalling on the installation of
	   another 16 for a month or so, because we've outgrown the building
	   we're in (and there's little point in adding more lines here just
	   before we move), and because our connection charges (AGT lines for
	   the dialup pool plus our network connection through Insinc) are
	   beating the hell out of us.  This year to date, those charges have
	   come to about $20,000.  And that's with the existing Centrex voice
	   lines - the next 16 will be POTS (plain old telephone service), at
	   about twice the cost per line, because AGT is refusing to give us
	   any more Centrex lines.

	d) We're trying to keep pretty squeaky-clean books, in anticipation
	   of going public next year, but nothing's been audited yet.  
	   Attached are the best we have, which are apparently pretty good.  
	   Please note, however, that they do not yet reflect the 3-4 man-years 
	   worth of so-far-unremunerated effort invested by a number of people,
	   nor the approximately $30,000 worth of equipment supplied by myself,
	   and yet to be rolled into the company.

	e) We'll have to hit the books on this one to give you more detail
	   than the rather rough "$20,000 of $70,000 total expenses" that the
	   income statement shows, but that gives you somewhat of an idea.

	f) A copy of our Centrex contract is attached.

	g) There has been no correspondence between AGT and Canada Connect
	   regarding the provision of Internet services.  AGT, it seems, only 
	   generates correspondence when dealing with the CRTC or CRTC-related
	   issues.  I do not perceive the CRTC as being the governing authority
	   in this matter, so I have not engaged in the "regulatory debate"
	   between AGT and the CRTC, which AGT appears to control.  I chose
	   to bring this matter to the Competition Bureau because the Bureau
	   is the body charged with enforcing the relevant law, and is not
	   in AGT's pocket.  In fact, AGT has explicitly refused to document
	   any of the matters in dispute, most recently their refusal to supply
	   Canada Connect with additional Centrex lines.

6.	I don't think that there's any doubt that the telcos are wetting 
	themselves with fear over what the cablecos can do to them in the
	bandwidth department.  Here's my take on the situation:  AGT, and
	the other telcos in this country, should be totally and absolutely
	prohibited from competing with us until the cablecos can provide
	competing local loop, assuming that we (the independent ISPs) are
	given equal access to the coax.  At that point, and only at that
	point, will the monopoly that we're combatting with this action
	cease to exist, permitting entry by the telcos and cablecos alike
	into fair competition with us.

7.	Yes.  If you tell them they have to.

It is interesting to note that some of the more "provocative" and "radical"
ideas offered above and in our Statutory Declaration may have been borne out
by AGT President Don Lowry himself in his letter of June 26 to the CRTC.
The bullying tone of this letter, and the "interim approval" granted to AGT 
shortly thereafter, hint at the political pressure rumored to have been brought 
to bear on the CRTC.  But especially interesting is his comment on pending 
competition:

	"AGT is aware that large U.S. based Internet access service 
	providers, specifically Microsoft and America On-Line, have
	installed systems in Edmonton and Calgary and are expected to
	commence service by the end of September."

AGT may have been aware of this, but to the best of our knowledge nobody 
else (except, of course, Microsoft and AOL) was.  This would suggest that
the "paranoid" claim put forth in our Statutory Declaration, that

	"...AGT/Telus is uniquely privy to confidential and proprietary
	information...and the potential exists for abuse of this confidential
	and proprietary information."

is more than justified.  This would appear to be compelling evidence that
AGT is quite prepared to exploit this insider's knowledge in order to further
its own agenda.

I hope that this answers your questions, and please accept my apologies if
I've strayed too far into rambling tangents or diatribe.  I shall be more
than happy to, in a more timely manner, answer any further queries you might
put forth.

Thank you.

					Yours,




					Jonathan Levine