Again ... What is it going to COST?????

bass at dmsd.UUCP bass at dmsd.UUCP
Tue Jul 15 17:33:22 AEST 1986


In article <6929 at utzoo.UUCP>, henry at utzoo.UUCP (Henry Spencer) writes:
> 
> If you follow the paths far enough, you find a site which is paying to get
> the stuff in.  Probably paying quite a bit.  As one of those sites, I have
> little sympathy for freeloaders.  Did you really expect that it would stay
> free forever?  Looking at it another way, if your local bill-paying site
> gives it to you free now, why should they cut it off?

Henry, being part of the backbone is a nice status deal, but for every
OFFICAL backbone member there are between 5-20 who also pay toll charges
to get and forward news. I am one, I see the bills every month, and
since I own this small 7 person company, I write the checks every month.
I forward the news to the local university free of charge as a service.
I KNOW what it costs, you don't have a monopoly on that or the right to
judge those who don't pay as free loaders. When I was in the bay area it
was easy to find a local feed ... here in the sticks (200 miles from either
LA or SF) is quite another problem. Many mid west sites share our delima.
The backbone ONLY SEES A FRACTION OF THE TOLL CHARGES. They have little
right to assume authority of the cost of the net ... neither do distant
sites .... Net service should be an open above the table issue.

Heck, I doubt most news administrators on the backbone have seen any
of the phone bills during the last year or know to the even dollar what
the totals are or what each link costs. Many more of us smaller guys who
pay the bills KNOW what the backbone only guesses .... the net is expensive.
Since it often comes out of our pockets, we are much more sensitive to
the spiral of cost caused by the increased flamage and social groups.

> 
> >	2) As a commercial service stargate based news traffic CANNOT be
> >	rebroadcast to ANY OTHER machine or site without paying the fee for
> >	that site as well...
> 
> I would assume that Stargate would permit local rebroadcasting.  If they
> don't, a lot of people are going to be very unhappy.  I'm not an expert on
> the legalities of this sort of thing, but I would assume that Stargate can
> waive fees for subsequent machines in the area if they want to.

THINK -- DON'T ASSUME ANYTHING. If they don't restrict it they will cut
into the user base substantially -- likely to the point Stargate will not
reach critical mass. If they do restrict it, I doubt it will successfully
replace usenet. In either case some form of uucp based usenet will likely
remain in service. I know of a number of sites that could not justify
Stargate service charges to their managment, but will stay on the net
somehow.

> "Multiple machines?  Us?  Heavens no; only one of our machines gets that
> traffic."  I don't see most such sites paying multiple site fees, either,
> and the Stargate people would be silly to expect it.  The Stargate people
> are not dummies, guys, nor are they vicious capitalist swine intent on
> making bundles of money at our expense.  If Stargate costs a lot more than
> Usenet, few people will bother.  They *know* that.

Expect it??? OR demand NOT??? THINK -- what are the business alternatives,
WHAT CHOICES SHOULD WE LOBBY FOR DURING THE NEGOTIATIONS for starting
Stargate???? If they continue to hide behind closed doors with popular support,
then the community will get what they get .... I'd rather have a choice.
The key is in your last statement ... Stargate WILL cost more than the current
near free for most sites, and given any choice they will not likely join
Stargate if the same data remains available over uucp based usenet. Justify
$1,000 to your boss while telling him you can get the same data free.
I agree ... "They *know* that" ... I ask what are they planning to do
about it????? This and most of the other questions are so basic that
they demand discussion prior to settling with a proposed vendor. How
is the vendor going to act if everyone bitches at the last minute???

> Here we come to the heart of the matter.  Usenet *won't* stay intact as a
> cost free competitor, because it is *not* "cost free" and never has been.
> Its costs are rising steadily, to the point where it clearly cannot survive
> in its present form too much longer.  Those sites who pay Long Distance
> bills are subsidizing the freeloaders.  THIS WILL NOT CONTINUE FOREVER!

Ahhh .... but the net MUST stay intact to carry items to Stargate via uucp
mail. Also one of the biggest benifits of the current uucp based usenet is
that uucp mail service goes most places. If the current net dries up,
what will happen to mail service??? Is Stargate without mail service still
the same net??? I doubt it. Being able to reply via email is a significant
part of usenet. In any case -- usenet will not die, it will change to cast
off the expensive baggage. Stargate will promote listen only sites that
are issolated from the rest of the net.

> > A lot more can be done to improve the cost of the usenet long haul
> > connections, which WILL LIKELY COST LESS THAN stargate.
> 
> And do less for us, in the long run.  Have you plotted the rise in traffic
> against time?  I recently posted five years of data to net.news.  Try it.
> Then figure out how much time each proposed cost-cutting measure buys us.
> It's usually not a lot.

I have followed the cost of usenet for about 5 years ... I KNOW about
the change in traffic and Costing ... I have voiced my concerns about
the cost of non-technical traffic over the last several years.

> Don't forget compatibility with older modems; few sites have phone lines
> that they can afford to *officially* dedicate to Usenet.  Don't forget
> processing time; vanilla uucp on 750-class machines can't consistently
> run even a 4800-baud line at full speed.  Don't forget that most of those
> modems are very new and still have problems, notably in areas like flow
> control between them and the host.

Most of the fast async modems also support some mix of std 2400/1200/300
access. Stargate will not solve the above problem either, likely it will make
it worse and the same solutions apply. Most of the new modems also handle
flow control with the same type software fixes required as the Stargate
buffer box. News itself is the problem, the data accesses per article
are only a few percent of the disk traffic ... very inefficent. Notes is
much better in some ways. A new dbms type system needs to be written that
doesn't use one file per message and can directly batch/unbatch streams
with a single fork/exec.
> 
> > 	2) Upgrade the uucp server to be full duplex --- IE carry traffic in
> > 	both directions concurrently -- this will likely improve the
> > 	connection costs about 30% for backbones and have little affect
> > 	on leaf sites.
> >
> > 	3) Implement a better I-have/I-want transmission scheme that is
> > 	real-time...
> 
> Are you volunteering to do it?  Usenet has never had any shortage of people
> with great ideas, but we do have a shortage of people implementing them.
> I also suspect that your estimate of cost improvement is high.
> 
> This amounts to redoing the transport mechanism completely, since uucp
> doesn't support this.  That is a lot of work; I sure don't have time to
> tackle it.
> 

I expect the that a good re-write of news to solve the filesystem problem
may also have a local non-uucp queue manager and transport service.
I have considered writing it for a long time, the market for such could
justify the development costs. It is about a 1.5 man year job, more than
my staff and I can afford to do public domain ... although notes is a damn
good start.

> > 	4) Negotiate a reduced flatrate DDS nite service AS A GROUP with
> > 	one of the long haul carriers...
> 
> > 	5) evaluate X.25 major city interconnections...
> 
> 
> Again, are you volunteering to do it?  Don't forget that we really do
> need data-quality transmission, which is a lot scarcer than cruddy voice
> circuits.  Also don't forget that many sites can get their Usenet phone
> bills past the bean-counters only because they are *not* explicitly
> identified as such.
>
> This is already being done, by many people independently.  It helps.
> It doesn't constitute a long-term solution to the problems, though.
>

Don't be so stupid to start discounting suggestions if the suggestor isn't
able to implement it by themselves ... hell stargate wouldn't have
gotten off the ground if Lauren was the only person involved. Good ideals
and goals are much cheaper than false starts.

Its the bean counters you refer to that will stop Stargate ....
you will need the phone service plus Stargate access fees plus
rental fees plus cable fees plus installation charges plus?????

> In case you didn't hear about it, Usenix solicited proposals of precisely
> this kind recently.  I believe they got, essentially, none.

I and many others in usenet are not Usenix members. I don't remember the
USENIX posting requesting alternative proposals ... I certainly would have
responded.
-- 

John Bass (DBA:DMS Design)
DMS Design (System Design, Performance and Arch Consultants)
{dual,fortune,polyslo,hpda}!dmsd!bass     (805) 541-1575



More information about the Comp.org.usenix mailing list