ATT preparing to impose bandwidth caps
Don Oxford
(06-12) 14:06 PDT NEW YORK, (AP) --
AT&T Inc., the country's largest Internet provider, is considering
charging extra for customers who download large amounts of data.
"A form of usage-based pricing for those customers who have abnormally
high usage patterns is inevitable," spokesman Michael Coe said this
week.
The top 5 percent of AT&T's DSL customers use 46 percent of the total
bandwidth, Coe said. Overall bandwidth use on the network is surging,
doubling every year and a half.
AT&T doesn't have any specific plans or fees to announce yet, Coe
said.
Most cable companies have official or secret caps on the amount of
data they allow subscribers to download every month. Time Warner Cable
started a trial earlier this month in Beaumont, Texas, under which it
will charge subscribers who go over their monthly bandwidth cap $1 per
gigabyte.
Cable companies are at the forefront of usage-based pricing because
neighbors share capacity on the local cable lines, and bandwidth hogs
can slow down traffic for others. Phone companies have been less
concerned about congestion because the phone lines they use to provide
Internet service using DSL, or Digital Subscriber Line technology,
aren't shared between neighbors, but AT&T is evidently concerned about
congestion higher up in the network.
Those who mainly do Web surfing or e-mail use little data and have
scant reason to pay attention to traffic caps. But those who download
movies or TV, particularly in high definition, can hit the caps
imposed by cable companies.
Download caps could put a crimp in the plans of services like Apple
Inc.'s iTunes that use the Internet to deliver video. DVD-by-mail
pioneer Netflix Inc. just launched a TV set-top box that receives an
unlimited stream of Internet video to a TV set for as little as $8.99
per month
Brad Allen
" total bandwidth, Coe said. Overall bandwidth use on the network is
" surging, doubling every year and a half.
That's a normal growth rate for a new service.
I'm pro-usage based charging, however, I always predicate that with
those charges being uninflated, i.e., not paying for customer service
or other welfare-type job titles at the ISP, and within about 5% - 10%
of the actual cost of that extra bandwidth (uninflated, unfaked, most
low cost efficient viewpoint), and in increments that are under $2 per
month (so no "$100,000 per month for every extra googleplex bytes
transfered", regardless of how good a price that seems per byte,
because of the excessively steep increment).
Cubit
"Brad Allen" <q...@sonic.net> wrote in message
news:48519a5a$0$17191$742e...@news.sonic.net...
In many areas these providers have an effective monopoly right now. They
will gouge with their pricing, if you let them.
Steve Pope
>"Brad Allen" <q...@sonic.net> wrote in message
>> I'm pro-usage based charging, however, I always predicate that with
>> those charges being uninflated, i.e., not paying for customer service
>> or other welfare-type job titles at the ISP, and within about 5% - 10%
>> of the actual cost of that extra bandwidth (uninflated, unfaked, most
>> low cost efficient viewpoint), and in increments that are under $2 per
>> month (so no "$100,000 per month for every extra googleplex bytes
>> transfered", regardless of how good a price that seems per byte,
>> because of the excessively steep increment).
>In many areas these providers have an effective monopoly right now. They
>will gouge with their pricing, if you let them.
That, and it's difficult for the user to restrict usage because
many websites just dump as much audio/image streams onto your
connection as they can possibly manage without crashing most
user systems.
It's nefarious if the service providers charge for bandwidth
while the content providers push so much of it at you
unnecessarily that you have no control.
There's not much content anymore that isn't saturated in this way.
Netnews, and Craigslist.
Steve
Roy
Unlimited isn't what it used to be, as providers test monthly caps, with
Sprint's mobile broadband service being the latest to restrict monthly
transfers.
http://www.pcworld.com/article/id,146752/article.html?tk=nl_dnxnws
Jeff Liebermann
You mean I can't watch IP TV on my cell phone? While driving?
Of course, complaining to Sprint Customer Service will probably get
you a disconnect notice:
<http://news.cnet.com/8301-10784_3-9739869-7.html>
I think this is called "cherry picking", where Sprint's long term
retention program targets well paying and docile customers, and let
the other carriers deal with those that use the service to the
advertised limit and complain when the bill is chronically amis.
I wonder what's coming next? Measured rate service?
--
Jeff Liebermann je...@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Steve Pope
>On Sun, 15 Jun 2008 17:07:04 -0700, Roy <aa...@aa4re.ampr.org> wrote:
>>Sprint's mobile broadband service being the latest to restrict monthly
>>transfers.
>>http://www.pcworld.com/article/id,146752/article.html?tk=nl_dnxnws
>You mean I can't watch IP TV on my cell phone? While driving?
Lots of people use Sprint EVDO as their main connectivity provider.
Even if you watch no video, it's quite easy to hit a 5G monthly
limit. If you listen to three hours of internet radio
per evening (I probably listen to at least that much), at
128 kbps, that's a little over 5G/month right there.
I can only imagine youtuber's go way over this, as would
myspacers where nearly every page has music on it.
Insidiously, that music is streaming even when the sound
on your computer is turned off, and the same is true
for lots of other websites. (If you're careful with your
browser settings perhaps you can avoid this.)
Steve
John Higdon
Jeff Liebermann <je...@cruzio.com> wrote:
> I wonder what's coming next? Measured rate service?
Caps and measured service serves two purposes. They put more money in
the provider's pocket, and they keep those pesky competing video and
other services away.
When your connections comes with a pile of usage and bandwidth
restrictions, it isn't really Internet anymore. It becomes an
"information service".
--
John Higdon
+1 408 ANdrews 6-4400
Roy
cheap, fast, reliable, unrestricted
pick three
Jeff Liebermann
Yep. That's about it.
I was expecting differentiated services, where (for example) the ISP
might charge more per megabloat for P2P than for email or news. Also,
different charge rates for outgoing packets to discourage servers.
Maybe charge more for normal business hours than for evenings. Perhaps
traffic averaging over several days to handle bursts and peaks.
Billing might become complicated, but no worse than the average
cellular phone plan, or Centrex pretzel. Well, maybe it can get
worse. Besides the obvious complexity, this also discriminates
against non-ISP provided services, such as VoIP and screaming video. A
lousy idea methinks, but one that I'm sure the major ISP's can't
resist trying.
Personally, I would be elated to see measured rate billing, where big
users pay for what they use, and are not subsidized by light users.
Flat rate made lots of sense when ISP's didn't have the personnel and
software necessary to do complex billing, but that's no longer the
case. As long as the ISP's don't favor their own services, or
differentiate by service, I won't complain.
John Higdon
Jeff Liebermann <je...@cruzio.com> wrote:
> I was expecting differentiated services, where (for example) the ISP
> might charge more per megabloat for P2P than for email or news. Also,
> different charge rates for outgoing packets to discourage servers.
This sounds very "marketplace driven". I use servers at home, so
obviously I would shop for a provider that doesn't "discourage" servers.
(Neither Sonic nor Speakeasy object to the use of servers, for example.)
> Maybe charge more for normal business hours than for evenings. Perhaps
> traffic averaging over several days to handle bursts and peaks.
> Billing might become complicated, but no worse than the average
> cellular phone plan, or Centrex pretzel. Well, maybe it can get
> worse. Besides the obvious complexity, this also discriminates
> against non-ISP provided services, such as VoIP and screaming video. A
> lousy idea methinks, but one that I'm sure the major ISP's can't
> resist trying.
There are some that believe that eventually, wireless services (phone,
data, whatever) will become nothing more than IP transport. If you think
about it, phone service is nothing more than an application for IP along
with everything else. Billing by the minute or by the byte is a legacy
construct. Rather than come up with ways to constrict demand, transport
companies need to figure out ways to keep up with it. Otherwise, all the
hyperbole about "the information age" is just so much ephemeral twaddle.
> Personally, I would be elated to see measured rate billing, where big
> users pay for what they use, and are not subsidized by light users.
> Flat rate made lots of sense when ISP's didn't have the personnel and
> software necessary to do complex billing, but that's no longer the
> case. As long as the ISP's don't favor their own services, or
> differentiate by service, I won't complain.
So you are saying, in effect, that service should be measured "because
it can be"? I see it as a fallback to legacy paradigms that proportion
and ration the information highway according to the financial and
equipment resources of the masses. You got more money...you get to
participate more. Gone over your limit? Guess you wait until next month
to resume participation.
You might want to note that among what's left of the switched circuit
telephone services, the trend is toward flat rate for local and
long-distance calling, even though billing methods are more
sophisticated than ever.
Roy
> ...
>
> I was expecting differentiated services, where (for example) the ISP
> might charge more per megabloat for P2P than for email or news. Also,
> different charge rates for outgoing packets to discourage servers.
> Maybe charge more for normal business hours than for evenings. Perhaps
> traffic averaging over several days to handle bursts and peaks.
> Billing might become complicated, but no worse than the average
> cellular phone plan, or Centrex pretzel. Well, maybe it can get
> worse. Besides the obvious complexity, this also discriminates
> against non-ISP provided services, such as VoIP and screaming video. A
> lousy idea methinks, but one that I'm sure the major ISP's can't
> resist trying.
>
> Personally, I would be elated to see measured rate billing, where big
> users pay for what they use, and are not subsidized by light users.
> Flat rate made lots of sense when ISP's didn't have the personnel and
> software necessary to do complex billing, but that's no longer the
> case. As long as the ISP's don't favor their own services, or
> differentiate by service, I won't complain.
>
My preference is to have something that is bandwidth controlled.
The solution I like is one where the line slows as you download and
speed up the more you are idle. Example: If you download at full speed
for 10 minutes then your line will run slow for an hour. Under this
scheme you may get a 10 Mbps max rate but you must average 256kbps over
an hour time frame.
The advantage to this scheme is that you a fixed cost (no extra
charges). Need a higher average data rate? No problem, you pay and
they increase your average rate.
John Higdon
<04GdnVyDUqFDW8vV...@posted.southvalleyinternet>,
Roy <aa...@aa4re.ampr.org> wrote:
> The advantage to this scheme is that you a fixed cost (no extra
> charges). Need a higher average data rate? No problem, you pay and
> they increase your average rate.
This is key to many situations. My home traffic is actually, in sum,
quite light. DNS, email, and Usenet do not take much of a toll. But
then...if I happen to be on the phone a lot in a particular month, or
some other server gets in some death grip with mine, or I get some sort
of DDOS attack, I don't want to open a bill for a zillion dollars.
Think about it for a second. The provider would have no incentive to
mitigate the problem because he would see all those counters rolling up
charges. It would be to his financial advantage to let the problem
persist as long as he can.
I've heard of cases where some host or process went berserk and the
hapless customer was handed a bill in a box.
No, thanks.
Steve Pope
>This is key to many situations. My home traffic is actually, in sum,
>quite light. DNS, email, and Usenet do not take much of a toll. But
>then...if I happen to be on the phone a lot in a particular month, or
>some other server gets in some death grip with mine, or I get some sort
>of DDOS attack, I don't want to open a bill for a zillion dollars.
>
>Think about it for a second. The provider would have no incentive to
>mitigate the problem because he would see all those counters rolling up
>charges. It would be to his financial advantage to let the problem
>persist as long as he can.
>
>I've heard of cases where some host or process went berserk and the
>hapless customer was handed a bill in a box.
I don't even thing you need a rogue process or an attack -- just
typical behavior of your computer and of websites.
For example, today "systeminfo" tells me my Windows XP
machine has been up for three hours and 41 minuts, and
"netstat -e" tells me I have transfered 25 megabytes.
That corresponds to 4.8 GB/month, around the size
of the proposed usage caps. But the kicker is I have
been doing no downloading, no listening to music,
no watching videos. Just netnews and web browsing.
(I did go to an IRS site and look at a tax form, but
that is the only download I can recall. I guess Adobe
documents are much bigger than they used to be.
Or maybe I have a zombie, who knows.
Steve
Salmon Moose
>
> My preference is to have something that is bandwidth controlled.
>
> The solution I like is one where the line slows as you download and
> speed up the more you are idle. Example: If you download at full speed
> for 10 minutes then your line will run slow for an hour. Under this
> scheme you may get a 10 Mbps max rate but you must average 256kbps over
> an hour time frame.
>
> The advantage to this scheme is that you a fixed cost (no extra
> charges). Need a higher average data rate? No problem, you pay and
> they increase your average rate.
I'd like a variant of that: you get full bandwidth until you hit your
monthly download cap, then bandwidth drops to dialup levels, but there is
no cell-phone style overage charge. Sorta like maintaining dialtone for
911 on a phone that has been cut off.
If the maximum is implemented as the ISPs claim (yeah I know), 90+% of
users would not be affected at all for the first few years (that is, until
their increasing youtube habits reach that ceiling).
And like your scheme, when you hit the ceiling you can upgrade your
bandwidth plan.
-Moose-
Roy
Your plan works for me except the download cap should be a smaller time
period. Maybe daily or weekly. How about a rolling total over the last
seven days
Salmon Moose
OK, I'll see your rolling cap and raise you a discount for late-night
usage.
After all, late night usage doesn't incur the same cost to the ISP as
daytime usage, because it doesn't cause them to increase peak bandwidth
capacity. Just like electrical time-of-use metering. And this is
supposedly about curbing real ISP costs rather than gouging customers in a
new and exciting way. Customers are already familiar with "free night and
weekend minutes" from cell plans.
Or to keep things simple (for marketing purposes), when the cap is reached
your bandwidth is reduced drastically except during late night, when it is
reduced less drastically.
© The OS/2 Guy ©
the fault of all those posting here. We spent months complaining
about Comcast's illegal throttling and reading of our private email
yet so many said it was their right.
Well, now you can pay them even more money and have your time online
capped. Comcast will literally turn you in to copyright holders if
you download any file that carries a copyright claim. Their goal is
to appease the music and media industry and stamp out P2P
downloading.
It is no one's fault but those who sat back and said nothing. This
always seems to be the case.
Tim...
Brad Allen
" many websites just dump as much audio/image streams onto your
" connection as they can possibly manage without crashing most
" user systems.
"
" It's nefarious if the service providers charge for bandwidth
" while the content providers push so much of it at you
" unnecessarily that you have no control.
"
" There's not much content anymore that isn't saturated in this way.
" Netnews, and Craigslist.
This is simple and easy. The ISP can make an interface in their
network that allows any utility user's computer to send control
statements that allow or disallow various traffic. Each port,
protocol type, etc., can be set to default of ask, drop, bounce, or
accept, and those defaults could be changed at any time, and would
have to be chosen every time for "ask". The ISP would set the usage
meter at the guarded portion of that network link, and only charge for
what the customer accepts. The programs designed to accept this would
be cryptographically impossible for any customer program to use except
the customer designated program, i.e., they would have to sign away
very intentionally the ability of any particular program in their
computer to be able to modify those values, therefore, a web browser
or virus could not decide to suddenly "authenticate" itself through a
button push or EULA agreement or other means and then start
authorizing its own content.
This would have the added side benefit of stopping a large amount of
criminal traffic (SPAM, breakins to your computer, network controlling
of your computer, spyware, etc.).
So, if the ISPs have a good faith attitude toward a proper regiment of
charging for use, then the excuse that they can't control what you do
with your link and don't have control over what others do with your
link is 100% bogus and somewhat irrelevent (they can give you control
over your own link).
Not everyone must understand how to use this system. They either (a)
don't use the Internet as effectively, in the same way that some
people don't pass IQ tests as effectively or pass SAT tests as
effectively, or (b) give up more control than smarter people do, to
purpose-built programs that are meant to do more hand-holding, that
they choose to do that, which will help them with this process.
The protocol for this network control ability would be standarized, so
that freeware and open source programs could be made for any OS and
user level. Perhaps there are already protocols in existance that
could be redesigned or enhanced to do more to fit these purposes, but
nothing should stop anybody from designing a fresh protocol if that's
better.
Brad Allen
" where nearly every page has music on it.
Pretty good imagination, but here's an update:
I go to a school with some young adults and old children in it (as low
as 15?), and they use Myspace (at the school). Replace "music" with
"audiovideo" -- yes, that's right, streaming music AND video.
" Insidiously, that music is streaming even when the sound on your
" computer is turned off, and the same is true for lots of other
" websites.
That's due to the poor design of computers that don't tell the
computer the sound is turned off, therefore there is no need for
sending that music.
" (If you're careful with your browser settings perhaps you can avoid
" this.)
If you don't have to be careful, then a lot wouldn't be. See my other
post for ideas about controlling your use. Having sound not be
transferred when the sound is muted or volume is 0 seems like a good
control point. But there are lots of possible transfers to control,
too.
Brad Allen
Wow, that's awesome. We'll have to test that.
Brad Allen
Comcast: Middle priced. Middle speed. Middle reliability.
Attempting to do restriction. We'll call that the "fast" and
"reliable". That's 2 of 4.
AT&T: Middle priced when you count billing and handling errors,
but otherwise cheap. Very low to low speed. Very bad to
halfway decent reliability. Now attempting to do restriction.
We'll call that the "cheap" & "reliable".
That's 1.5 of 4.
Sonic.net: Middle to high priced. Slightly higher speed than AT&T
in most cases. Slightly higher reliability than AT&T
in most cases. (Both of those are because they use AT&T,
and are otherwise superior.) Can buy out of restrictions,
but not completely, but the restrictions are low to medium.
We'll call them "reliable" without the "fast" or the "cheap".
That's 1.8 of 4.
Fiber connection: Fast and reliable. Not cheap. Unrestricted.
There's 3 of 4.
T-1: Sort of reliable, not cheap, not fast. Unrestricted.
That's 1.5 of 4.
Satellite: Expensive, unreliable, slow, and restricted. 0 of 4.
EVDO & the other cell phone ones: Medium price, medium to slow speed,
questionable reliability, and very restricted. That's about
1.8 of 4.
I'm not pegging it. Sorry. Only fiber adheres.
AES
a private residence), and the problems that might arise from people
"pushing" large amounts of unexpected or unwanted data at you,
especially if you were billed by total volume of data sent down to you.
(This is a follow-on to recent discussions of that last point)
I have a setting in my Eudora email app such that for any message larger
than 500K the mail server at my ISP (which happens to be Stanford
University) just sends the headers, the first few lines of the message,
and a clickable option to either send or discard the remainder of this
large message the next time I check mail. Meanwhile, of course, I
continue to get all smaller messages without trouble.
I understand that this mode of operation can work well for data that
arrives in discrete chunks, namely individual email messages; that it
requires the active cooperation of my ISP; and that the ISP must be
willing and able to store the longer messages on its machine until I
accept or reject them.
What I don't understand are the potential problems I might encounter
related to other types of large-volume data that a wide range of people
or organizations might "push" at me (putting "push" in quotes because I
don't claim to understand its full technical meaning -- if it has one).
I know that Apple somehow silently pushes large upgrades (10s to 100s of
MB) down to me over my miserable 300kB DSL connection, because suddenly
a message pops up, asking me if I want to install some 100 MB upgrade;
and when I click Yes, the upgrade is accomplished and done in a minute
or less -- and although I have no problem with this, I really don't
recall ever having given Apple explicit permission to do these silent
downloads -- or did I, somewhere along the line?
So, now, as a result of the posts on this group, I'm worrying about
things like:
* Can other senders do this to me also?
* If I purchase and install a software app, or click on some PDF
document on a vendor's site, or a video on Huff Post, or register with
amazon, can they later silently start sending large amounts of data at
me, that I've not specifically requested?
* Are Eudora-like mechanisms to protect against this technically
feasile? Could -- or would -- an ISP store the data (potentially very
large files) while they ask me if I want to receive it?
* If someone started sending a flood of data at me, could or would an
ISP (a well-intentioned one) be technically able to continue sending
lower-volume transmissions through to me, while throttling or blocking
the large volume ones that would send my bill skyrocketing?
* Can an ISP even somehow identify at the start "large volume"
transmissions being sent to me, so as to be able to throttle them, while
sending low-volume transmissions unthrottled? What are the technical
limitations on this?
I guess as really broadband residential connections start to appear, we
naive customers/consumers are all going to have learn a lot more about
this . . .
AES
q...@sonic.net (Brad Allen) wrote:
>
> I'm not pegging it. Sorry. Only fiber adheres.
>
Fiber is visibly light-years ahead of any other physical medium -- but
does that just push the problem back to the limitations of the hardware
and facilities in the provider central office?
Jeff Liebermann
<HisRoyalRa...@kome.com> wrote:
>In article <tp0c541e1oc6s7q9f...@4ax.com>,
> Jeff Liebermann <je...@cruzio.com> wrote:
>
>> I was expecting differentiated services, where (for example) the ISP
>> might charge more per megabloat for P2P than for email or news. Also,
>> different charge rates for outgoing packets to discourage servers.
>
>This sounds very "marketplace driven". I use servers at home, so
>obviously I would shop for a provider that doesn't "discourage" servers.
>(Neither Sonic nor Speakeasy object to the use of servers, for example.)
As the ISP business grows, that exactly what I expect to see
happening. For example, I need a service that will work for VoIP and
security cameras. 128Kbits/sec line maximum with low latency. No
need for huge download speeds or megabit speeds. 384/384k would work
fine. However, there's no such service that's economical because such
low speed services are expected to subsidize the P2P addicts and IPTV
services. For other customers, I might need a bandwidth plan that is
optimized for servers or for a variety of specialized applications.
One size does NOT fit all.
The problem is how such specialized services are billed and deployed.
At this time, it's perhaps 3 performance tiers and flat rate billing.
With the imposition of rate caps, the bandwidth hogs were throttled.
With the imposition of bandwidth caps, the downloaders were throttled.
My guess(tm) is that time caps are next, where daytime bandwidth is
more expensive than the lighter used nightime or weekend bandwidth.
Next comes performance caps, where some types of packets are delayed
by the ISP, which in effect provide QoS services. For example, ISP's
can offer 15msec lantency for SIP packets, but only 100msec latency
for HTTP at a reduced cost.
How all these caps and traffic management are sold to the GUM (great
unwashed masses) is the real problem. At this time, any change would
be considered a reduction in service and therefore a loss to the GUM.
I have some ideas, but they're not ready for prime time.
>> Maybe charge more for normal business hours than for evenings. Perhaps
>> traffic averaging over several days to handle bursts and peaks.
>> Billing might become complicated, but no worse than the average
>> cellular phone plan, or Centrex pretzel. Well, maybe it can get
>> worse. Besides the obvious complexity, this also discriminates
>> against non-ISP provided services, such as VoIP and screaming video. A
>> lousy idea methinks, but one that I'm sure the major ISP's can't
>> resist trying.
>There are some that believe that eventually, wireless services (phone,
>data, whatever) will become nothing more than IP transport.
Methinks that would only happen if the wireless IP data carriers were
considered a public utility or regulated common carrier. It's
certainly a possibility, but the trend seems to be toward billing
independently by service, not one bill for everything. A common
question I hear is "why is cellular data billing so different from
broadband wire line billing?" For example, I'm billed differently and
seperately for cellular voice, SMS, data, and extensions (added
phones). Visualize being billed seperately for email, web, IP phone,
P2P, IPTV, etc, each with their own rate and bandwidth caps. That
would seem to be a difficult sell unless is were properly sold. For
example, the same ISP could offer an "unlimited" account, but at a sky
high price. The rate, bandwidth, and QoS caps would then be sold as a
discount off the "unlimited" plan.
>If you think
>about it, phone service is nothing more than an application for IP along
>with everything else. Billing by the minute or by the byte is a legacy
>construct. Rather than come up with ways to constrict demand, transport
>companies need to figure out ways to keep up with it. Otherwise, all the
>hyperbole about "the information age" is just so much ephemeral twaddle.
Yep. Sounds good. Now try thinking about it from the ISP's point of
view. The problem is how to supply decent service, make a decent
profit, placate the regulators, pacify the pundits, and do it all
without continuous exponential plant and equipment growth. What
Sprint did to "solve" this problem with wireless is simply dump their
2000 top complainers. I've been told that was a HUGE improvement.
ISP's could do the same thing, but that's only a temporary solution as
the supply of complainers seems infinite. Whether it was legacy or
Neanderthal is irrelevent. If it works, it can be raised from the
dead and sold as the next big thing. I would have not trouble selling
bandwidth limits if it's perceived as a discount off the "regular"
price or as a means of not having light users subsidize the P2P
downloaders.
>> Personally, I would be elated to see measured rate billing, where big
>> users pay for what they use, and are not subsidized by light users.
>> Flat rate made lots of sense when ISP's didn't have the personnel and
>> software necessary to do complex billing, but that's no longer the
>> case. As long as the ISP's don't favor their own services, or
>> differentiate by service, I won't complain.
>
>So you are saying, in effect, that service should be measured "because
>it can be"?
No. Service may be measured because it *CAN* be.
The reason that complex billing was previous impossible was the lack
of software, monnitoring and horsepower necessary to do it
effectively. Simple billing meant low support overhead, which was a
big plus as the ISP biz was growing. That's no longer such a major
restriction where todays computers and staff can easily handle complex
billing. It's an option that I'm guessing(tm) the major ISP's are not
going to resist.
>I see it as a fallback to legacy paradigms that proportion
>and ration the information highway according to the financial and
>equipment resources of the masses.
Yep. That's about it.
For each according to their need.
To each according to what they can pay.
Financial equality works. None of the others seem do to anywhere near
as well.
>You got more money...you get to
>participate more. Gone over your limit? Guess you wait until next month
>to resume participation.
I'm sure something more useful and complicated can be arranged. For
example, an exponentially tiered rate for going over the various
limits. Going over slightly is not penalized much. However, that's
not the way cellular billing currently works. Any overtime is
severely penalized at a horrific rate. Same with late payments on
credit cards, and bank overdraws. Those are intentional and are
designed to trap consumers into going over their limits. It's also
become a huge profit center for providers. There's no incentive to
cut off service after going over some limit. There's a huge incentive
to financially penalize those that do. My guess is the ISP's will
follow the cellular and credit card examples and go for the gold.
>You might want to note that among what's left of the switched circuit
>telephone services, the trend is toward flat rate for local and
>long-distance calling, even though billing methods are more
>sophisticated than ever.
That depends on what they're selling and who's buying. Managed
telecom services basically means billing for everything seperately.
Large companies, with extensive and complex telecom needs will want to
continue the complex model because it gives them more control over the
charges. The GUM isn't so particular and seems to want simplified
billing, even if it is more expensive.
For each according to their need.
To each according to what they can pay.
Billed according to what they can tolerate.
John Higdon
Jeff Liebermann <je...@cruzio.com> wrote:
> I'm sure something more useful and complicated can be arranged. For
> example, an exponentially tiered rate for going over the various
> limits. Going over slightly is not penalized much. However, that's
> not the way cellular billing currently works. Any overtime is
> severely penalized at a horrific rate. Same with late payments on
> credit cards, and bank overdraws. Those are intentional and are
> designed to trap consumers into going over their limits. It's also
> become a huge profit center for providers. There's no incentive to
> cut off service after going over some limit. There's a huge incentive
> to financially penalize those that do. My guess is the ISP's will
> follow the cellular and credit card examples and go for the gold.
Which is primarily why I'm against metering. Rather than being used as a
means to equitably distribute costs among customers, it is almost
invariably set up as a trap to catch the unwary.
The period used to measure the data could be a major factor in the
fairness of any metering system. For instance, Apple has a monthly cap
on its .mac subscribers, but measures the usage semi-monthly. If, in the
two week period, the customer uses more than *half* his monthly transfer
allotment, the punitive measures kick in.
I would suggest going the other way. Make the measurement annually. That
way, a customer's own "light" months could subsidize his "heavy" months.
I don't see why that wouldn't be fair.
I'm hardly an entry-level user, but I don't do video downloads, and in
fact do minimal surfing on the web. I have no idea what my monthly
throughput happens to be. I handle DNS services for about a hundred
domains (some of them well-known and popular, but with long refresh and
TTL times), and email for a dozen or so users (with POP and a web
interface). In addition, I have four kitty-cams in the house that update
to a server (viewable on the web) every ten seconds, as well as a myriad
of remote control devices.
All of my telephone traffic is via VOIP. During a busy day, there might
occasionally be two or three calls active at once.
All of this is handled by two 6 Mbps DSL connections. They aren't
particularly cheap, either. But I'll be G-D if I'm going to put up with
something like an extra two seconds on a phone call dropping me over to
the next measurement tier for an extra fifty bucks or whatever. I've
heard the stories of the Verizon users that got billed thousands of
dollars extra for violating the sacred 5GB "secret cap" on EVDO. That's
nuts.
AES
Jeff Liebermann <je...@cruzio.com> wrote:
Jeff,
As a residential customer/at-home consultant (not in IT field) I could
live with a variety of billing schemes. One thing I couldn't live with
would be any situation where any person or organization pushing large,
unrequested amounts of data at me would, by pushing me up against either
bandwidth or volume caps, block me from receiving simultaneous small
(but still possibly unrequested) transmissions addressed to me.
In other words, "large" (ergo expensive) unrequested transmissions,
especially any that would run me way over a cap (of any kind), or that
would run my bill way up above normal, get blocked, or severely
throttled, or held up by the ISP until I agree to receive them --- but
"small" transmissions to me always get through (spam filtering aside),
even if unrequested, and despite throttling or blocking of large
transmissions.
Is this technically feasible, within reason? Is there a feasible way
for an ISP to distinguish between "large" and "small" traffic for me (in
the broad definitions given above), as this traffic begins to arrive in
their facilities?
Or to distinguish between "requested" and "unrequested" traffic coming
in addressed to me? -- especially when some of the requested traffic may
be standing subscriptions I've put in place, and some may be one-time
requests I've sent to someone.
Steve Pope
>" Insidiously, that music is streaming even when the sound on your
>" computer is turned off, and the same is true for lots of other
>" websites.
>That's due to the poor design of computers that don't tell the
>computer the sound is turned off, therefore there is no need for
>sending that music.
Well website design figures into it as well -- even if you
turn off "play sounds automatically" in your browser, websites
will stream sound at you because the sound players on many websites
do not conform to any particular standard.
>If you don't have to be careful, then a lot wouldn't be. See my other
>post for ideas about controlling your use. Having sound not be
>transferred when the sound is muted or volume is 0 seems like a good
>control point.
If it can be done.
Steve
Steve Pope
>I know that Apple somehow silently pushes large upgrades (10s to 100s of
>MB) down to me over my miserable 300kB DSL connection, because suddenly
>a message pops up, asking me if I want to install some 100 MB upgrade;
>and when I click Yes, the upgrade is accomplished and done in a minute
>or less -- and although I have no problem with this, I really don't
>recall ever having given Apple explicit permission to do these silent
>downloads -- or did I, somewhere along the line?
Apple is one of the worst offenders. Even if you're completely
uninstalled Apple Qucktime, Apple continues to auto-download updates for
it. There's no way to stop this as far as I can tell.
Never install Quicktime, unless you have a way of restoring
your computer to its previous state.
Steve
Roy
> ...
>
> I'm not pegging it. Sorry. Only fiber adheres.
Actually they all adhere. No one scores more than 3.0 on your scale.
I think you missed the point. You cannot have it all. Something must
give. You want to pick a provider as close to 3.0 with the emphasis on
the attributes you want.
A+B+C+D <= 3
You can certainly pick a lower rated provider. Remember the Indiana
Jones movie: "He chose poorly."
The bottom line is there is no perfect solution, everything is a tradeoff.
John Slade
"Don Oxford" <m...@witty.com> wrote in message
news:1ec81f5f-5e47-4425...@d1g2000hsg.googlegroups.com...
> http://www.sfgate.com/cgi-bin/article.cgi?f=/n/a/2008/06/12/financial/f140613D00.DTL&tsp=1
>
> (06-12) 14:06 PDT NEW YORK, (AP) --
>
> AT&T Inc., the country's largest Internet provider, is considering
> charging extra for customers who download large amounts of data.
>
> "A form of usage-based pricing for those customers who have abnormally
> high usage patterns is inevitable," spokesman Michael Coe said this
> week.
>
> The top 5 percent of AT&T's DSL customers use 46 percent of the total
> bandwidth, Coe said. Overall bandwidth use on the network is surging,
> doubling every year and a half.
>
> AT&T doesn't have any specific plans or fees to announce yet, Coe
> said.
If and when they start charging because I may use more, I will dump
them quickly.
John
** Posted from http://www.teranews.com **
Steve Fenwick
spo...@speedymail.org (Steve Pope) wrote:
Huh? On a Mac, I go into "System Preferences", pane "Software Updates",
and uncheck the "Check for Updates" box (or set it to "Manually" on some
versions). No forced updates that I can tell.
Steve
--
steve <at> w0x0f <dot> com
"Life should not be a journey to the grave with the intention of
arriving safely in an attractive and well preserved body, but rather to
skid in sideways, chocolate in one hand, chip shot in the other, body thoroughly
used up, totally worn out and screaming "WOO HOO what a ride!"
Steve Fenwick
Meer.net over Covad backhaul: fast (ADSL2, 15Mbits), reliable,
unrestricted. Not cheap. I think they can be said to adhere.
John Higdon
Steve Fenwick <nos...@nospam.invalid> wrote:
> Huh? On a Mac, I go into "System Preferences", pane "Software Updates",
> and uncheck the "Check for Updates" box (or set it to "Manually" on some
> versions). No forced updates that I can tell.
Not only that, on the automatic "check for updates", you are always
given a list of the updates and the ability to pick and choose. If you
don't ever want to update QuickTime, you don't have to. I think that's a
bad choice, but you do have it.
Brad Allen
Roy <aa...@aa4re.ampr.org> wrote:
" Brad Allen wrote:
" > ...
" >
" > I'm not pegging it. Sorry. Only fiber adheres.
"
" Actually they all adhere. No one scores more than 3.0 on your scale.
Ahh, of course you're right.
" I think you missed the point. You cannot have it all. Something
" must give. You want to pick a provider as close to 3.0 with the
" emphasis on the attributes you want.
"
" A+B+C+D <= 3
"
" You can certainly pick a lower rated provider. Remember the Indiana
" Jones movie: "He chose poorly."
"
" The bottom line is there is no perfect solution, everything is a tradeoff.
Yes, right. Ok. We can continue to test it. The question is perhaps
something like whether the measurement tripod of "fast, cheap,
reliable/good quality" will be unbalanced with another leg to the
point that it will mismeasure (which is a bad metaphore because that
measurement says that one leg will always be bad, and in any case more
than one leg could be bad). I don't have the time I used to to
actually think about that. (Note: added measurement "leg" is
restriction, which was always there, but just not separately measured.)
Brad Allen
" To each according to what they can pay.
" Billed according to what they can tolerate.
Seems pretty accurate. (I get slightly miffed because I'm a slightly
washed member of the mass with frequently unbalanced pay.)
Brad Allen
" as a means to equitably distribute costs among customers, it is
" almost invariably set up as a trap to catch the unwary.
That's the same sledge hammer I've been attempting to get ready
myself. I explain how measured use is a good thing, and how it is
not, and it is almost always differentiated in how it is done.
There's a difference between measured use and bad implementation, or
exorbitant billing, or criminal billing, or stealing, or anything
else.
Brad Allen
For today's MBA's, that don't know anything about building Internet
standards? (I'm not sure about that: does the bribe and steal method
actually work?)
" Is there a feasible way for an ISP to distinguish between "large"
" and "small" traffic for me (in the broad definitions given above),
" as this traffic begins to arrive in their facilities?
I know it's theoretically possible: the customer can (even
cryptographically) authenticate certain traffic types as acceptable,
only giving those people the key that deserve to communicate with
them.
" Or to distinguish between "requested" and "unrequested" traffic
" coming in addressed to me? -- especially when some of the requested
" traffic may be standing subscriptions I've put in place, and some
" may be one-time requests I've sent to someone.
Theoretically, not entirely, but a large amount of that can be. You
can put the rest into maybe authentifications, with strict pause
controls -- you get your version of a reasonable "pop-up" to see if
you want to unpause questionable transfers. (Yes, I'm referring to
that which includes SPAM, which means that SMTP/IMAP/POP/BSMTP would
need to be updated to handle partial transfers, authentificition,
etc.)
Brad Allen
"
" The solution I like is one where the line slows as you download and
" speed up the more you are idle. Example: If you download at full
" speed for 10 minutes then your line will run slow for an hour.
" Under this scheme you may get a 10 Mbps max rate but you must
" average 256kbps over an hour time frame.
"
" The advantage to this scheme is that you a fixed cost (no extra
" charges). Need a higher average data rate? No problem, you pay and
" they increase your average rate.
Comcast does a sort of combination of this and other junk. The "this"
part (what you describe) works really well. I can dump data any
direction (it's bidirectional control of this sort), sometimes
uploading 5MiB files at about 300KiBy/s to 400KiBy/s (yesterday) or
downloading 100MiBy at 3MiBy/s (yesterday), and the rest of the time I
just meter my traffic to something under 70KiBy/s outward, and Comcast
does the rest (it usually doesn't exceed 1MiBy/s, and anything slower
is caused by the sender sending slower). If I didn't meter traffic in
this manner or I had a bunch to dump in some direction, I'd see
Comcast ratchet down my speed slowly but much more so than I want, so
I do it from my side preemptively to avoid trouble. I also pay
attention to peak use times, and avoid higher sustained use at those
times (but don't care about my immediate needs: if it can't handle
those, then it isn't worth it).
I also pay an arm and a leg for service (about $60/month part of a
triple play at $160/month so it gets discounted about $10/month, so
otherwise it'd be $70/month), but $70/month isn't that much for richer
people, so it might still be considered "cheap" even though it is so
expensive.
I don't need it to be done this way (there are lots of other ways that
ought to work great, even better), but I'm just reporting that it
seems to currently work OK, even well, from a user's perspective. Of
course, they're trying to change it, so that is part of the question
still.
Brad Allen
I once had to fall back to modem this year. I turned off all big
transfers, and the line was still unusable. Unauthorized transfers
filled the pipe for hours. Unfortunately, there aren't tools in
Microsoft XP to find out what processes are using bandwidth and/or
disk (another of my problems), so I had to kill everything. It
eventually died down, but "kill everything" didn't work since some
stuff stayed anyway or turned off the computer, and it would still
transfer unauthorized transfers.
Brad Allen
" > your monthly download cap, then bandwidth drops to dialup levels,
" > but there is no cell-phone style overage charge. Sorta like
" > maintaining dialtone for 911 on a phone that has been cut off.
[...]
" > And like your scheme, when you hit the ceiling you can upgrade your
" > bandwidth plan.
" >
" > -Moose-
"
" Your plan works for me except the download cap should be a smaller
" time period. Maybe daily or weekly. How about a rolling total over
" the last seven days
That would seem reasonable if you add to it the ability to do one-time
upgrades in bandwidth in heavy weeks -- log in to their bandwidth
website and purchase another 5GiBy that week for $0.50 or 100GiBy for
$0.75 or something, and the gates are opened up more until that's
used. No big surprises, and full utility, the way a utility is
supposed to work.
Brad Allen
" usage.
"
" After all, late night usage doesn't incur the same cost to the ISP
" as daytime usage, because it doesn't cause them to increase peak
" bandwidth capacity. Just like electrical time-of-use metering. And
" this is supposedly about curbing real ISP costs rather than gouging
" customers in a new and exciting way. Customers are already familiar
" with "free night and weekend minutes" from cell plans.
"
" Or to keep things simple (for marketing purposes), when the cap is
" reached your bandwidth is reduced drastically except during late
" night, when it is reduced less drastically.
Quite good.
Jeff Liebermann
(Steve Pope) wrote:
>For example, today "systeminfo" tells me my Windows XP
>machine has been up for three hours and 41 minuts, and
>"netstat -e" tells me I have transfered 25 megabytes.
>That corresponds to 4.8 GB/month, around the size
>of the proposed usage caps.
My daily "idle" usage is about the same. I fired up Wall Watcher to
see what was happening. About half of my traffic was compliments of
Skype and other distributed directory servers. I'll do some more
analysis later.
Also, the final version of DD-WRT v24 includes graphs of daily WAN
traffic for quota management. So far for this month mine looks like
this:
<http://802.11junk.com/jeffl/crud/dd-wrt-wan-graph.jpg>
Looks like about 5GB/month for me at home. The office graph is about
twice that.
Roy
> On Mon, 16 Jun 2008 21:26:27 +0000 (UTC), spo...@speedymail.org
> (Steve Pope) wrote:
>
>> For example, today "systeminfo" tells me my Windows XP
>> machine has been up for three hours and 41 minuts, and
>> "netstat -e" tells me I have transfered 25 megabytes.
>> That corresponds to 4.8 GB/month, around the size
>> of the proposed usage caps.
>
> My daily "idle" usage is about the same. I fired up Wall Watcher to
> see what was happening. About half of my traffic was compliments of
> Skype and other distributed directory servers. I'll do some more
> analysis later.
>
> Also, the final version of DD-WRT v24 includes graphs of daily WAN
> traffic for quota management. So far for this month mine looks like
> this:
> <http://802.11junk.com/jeffl/crud/dd-wrt-wan-graph.jpg>
> Looks like about 5GB/month for me at home. The office graph is about
> twice that.
>
I guess I am a bit busier. I have averaged about 30kbps download for
the last year which comes out to 10GBytes/month I think. I think
watching Bloomberg TV while the stock market is open probably
contributes significantly to that number
Jeff Liebermann
<HisRoyalRa...@kome.com> wrote:
>In article <o7ei54tkjarfrhuds...@4ax.com>,
> Jeff Liebermann <je...@cruzio.com> wrote:
>> There's a huge incentive
>> to financially penalize those that do. My guess is the ISP's will
>> follow the cellular and credit card examples and go for the gold.
>Which is primarily why I'm against metering. Rather than being used as a
>means to equitably distribute costs among customers, it is almost
>invariably set up as a trap to catch the unwary.
Yep. I'm not sure that such predatory pricing was initially
intentional. As various companies that charge by the month soon
discovered, a rather substantial percentage of their customers will
tend to pay late or forget to pay their bill. Short due dates are not
helping. For example, my Master Charge bill usually arrives with only
a 2 to 4 day window to pay (including time for my check to arrive).
Visa is a bit better with a 5 to 8 day window. When I complained,
they moved the billing date back, but also moved the due date the same
amount, resulting in no net improvement. I suspect Ma Bell will
eventually figure out the technique and do the same. I hope they do,
because I expect this nonsense to precipitate some consumer protection
legislation.
>I would suggest going the other way. Make the measurement annually. That
>way, a customer's own "light" months could subsidize his "heavy" months.
>I don't see why that wouldn't be fair.
Some cell phone plans already do that. For example, Cingular has roll
over minutes. It's certainly possible.
<http://www.wireless.att.com/learn/why/rollover.jsp>
>I've
>heard the stories of the Verizon users that got billed thousands of
>dollars extra for violating the sacred 5GB "secret cap" on EVDO. That's
>nuts.
<http://finance.yahoo.com/expert/article/millionaire/30108>
<http://www.cbc.ca/canada/edmonton/story/2007/12/12/cell-phone.html>
The real danger is not sloppy usage or secret bandwidth caps. It's
getting the phone stolen and having some bozo rack up a huge charge or
tries to use it for identity theft. If your phone walks away, login
online and lock the phone.
Back to the original topic. What exactly happens when you hit the
proposed bandwidth cap? Does internet access die? Does it slow down?
Does the billing rate increase linearly? Exponentially?
Astronomically? There's been considerable discussion on the setting
of a bandwidth cap limit, but not much on what happens when you go
over. I think that's the key issue.
--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558 je...@comix.santa-cruz.ca.us
# http://802.11junk.com je...@cruzio.com
# http://www.LearnByDestroying.com AE6KS
Jeff Liebermann
>Is there a feasible way
>for an ISP to distinguish between "large" and "small" traffic for me (in
>the broad definitions given above), as this traffic begins to arrive in
>their facilities?
Feasible, practical, or economical?
Feasible is certainly possible. Some form of traffic analysis or
packet content sniffer, with a monsterous buffer able to store
sufficient data to make a decision. It certainly would work, but your
latency will suffer.
Practical is dubious but possible. It's like the current spam
filters. They're functional as long as the incoming traffic is easily
classifiable. However, SSL and VPN tunnels will fail. Same with a
bunch of little known protocols.
Economical is very dubious. Would you be willing to pay the ISP for
this service? Or would buying a bit more bandwidth be cheaper?
>Or to distinguish between "requested" and "unrequested" traffic coming
>in addressed to me?
I don't think you want to design the rule set and filters necessary to
do that effectively. For example, I may want to watch flash videos on
YouTube, but not in the intro to bloated web pages. How are you going
to make the distinction?
> -- especially when some of the requested traffic may
>be standing subscriptions I've put in place, and some may be one-time
>requests I've sent to someone.
Again, the problem is the complexity of the rule set. Methinks it's
too complex to be useful or practical.
Brad Allen
" proposed bandwidth cap? Does internet access die? Does it slow
" down? Does the billing rate increase linearly? Exponentially?
" Astronomically? There's been considerable discussion on the setting
" of a bandwidth cap limit, but not much on what happens when you go
" over. I think that's the key issue.
I agree that that's super important. I think that's the #2 issue,
with price being #1. Without both being correct, I don't think it can
work right.
Brad Allen
" too complex to be useful or practical.
I think it could go through an overly complex phase where things just
aren't done right, but eventually it could be properly categorized in
a very simple, intuitive, complete and correct way, as long as it is
not controlled by those who want to shape our use for us. Those types
of people can never get anything right, and everything fails around
them.
Brad Allen
" things like:
"
" * Can other senders do this to me also?
If both computers agree to it, then TCP will flow a lot. If only one
computer agrees to it, then any protocol packet type could flood your
computer. This is what I was talking about in another post: you'd
only allow traffic that had proper authentification.
Does IPv6 have this, btw? Doesn't matter: it can be added.
" * If I purchase and install a software app, or click on some PDF
" document on a vendor's site, or a video on Huff Post, or register with
" amazon, can they later silently start sending large amounts of data at
" me, that I've not specifically requested?
Did you agree to their contract? If so, then the contract may say so.
TVs don't have a contract when you flip channels, and street signs
don't either, but a lot of WWW sites do.
" * Are Eudora-like mechanisms to protect against this technically
" feasile? Could -- or would -- an ISP store the data (potentially very
" large files) while they ask me if I want to receive it?
They could store it, or more importantly they could not transfer it at all.
" * If someone started sending a flood of data at me, could or would
" an ISP (a well-intentioned one) be technically able to continue
" sending lower-volume transmissions through to me, while throttling
" or blocking the large volume ones that would send my bill
" skyrocketing?
That depends if they can differentiate between them. That's why you
need a way to authenticate each item, type, source, destination,
etc. of every communication very specifically for specified
authentification to work, otherwise someone could bypass that system
somehow.
" * Can an ISP even somehow identify at the start "large volume"
" transmissions being sent to me, so as to be able to throttle them,
" while sending low-volume transmissions unthrottled? What are the
" technical limitations on this?
Same as previous point.
" I guess as really broadband residential connections start to appear,
" we naive customers/consumers are all going to have learn a lot more
" about this . . .
True enough, but the concepts aren't that hard for average IQ people.
You'll have no trouble figuring it out. Eventually we will have
limitations on bandwidth that will affect everybody, and lots of
applications available that can usefully use more than is available,
and everyone will learn how to deal with this naturally.
Jeff Liebermann
Methinks you have it backwards. Things always start simple, and
rapidly become more complex.
For example, we started with exactly 10 commandments to help
distinguish between good and bad. Simple rules for simple living. The
original leaders managed to inflate that into about 600 laws before
they got tired and decided that everything else would be an
interpretation, instead of a new law. Nice try.
Fast forward to today, and we have buildings full of laws, rules,
codes, regulations, dictates, pontifications, edicts, torts, case law,
and interpretations. All this just to distinguish right from wrong.
If you happen to have a "simple, intuitive, complete, and correct way"
to distinguish between right and wrong, I would welcome it. Then we
can close the law libraries, dismiss the attorneys, and live happily
ever after.
It's like that with any filter. Spam filtering started out simple
(keyword), became complex (Baysian), and rapidly approached cryptic
(AI). Same with OS identification. NMAP has turned into a really
complex algorithm for guessing the server OS. I suspect that sniffing
traffic for desired (good) and undesired (bad) traffic, will have
similar problems.
Jeff Liebermann
Ok, let me try to simply the way it works.
You have a choice of two services. The first is slightly more
expensive, but offers some slack if you over-use your quota or limits.
The terms of service are restrictive, but not unreasonable. However,
the provider goes out of its way to dump complainers, abusers, and bad
payers. Basically, this one makes its money on long term and reliable
users with modest and fairly consistent requirements.
The 2nd service provider is somewhat substantially cheaper, but
financially penalizes you if you go over the limits, pay late, or
otherwise walk onto a prepared land mine. The rate schedule is
intentionally incomprehensible. The terms of service are highly
restrictive, overly broad, and vague. However, they are not commonly
enforced and are waiting to be used in court to backup the service
providers position. Basically, this one makes its money on the users
that can't plan their usage and payments.
The 2nd service provider is obviously cheaper. Would you still prefer
the cheaper provider?
If that's not good enough, perhaps a different pricing schedule might
be appealing. I'll give you a 33% discount off the retail service
rate if you agree to pay 20 times the retail rate if you exceed your
*DAILY* quota. Why daily? Because the cost is in the PEAK bandwidth,
not the average. If everyone on the system decided to watch the
Victoria's Secret Fashion show at the same time, the ISP would need to
buy more expensive capacity to meet the peak demand. Would you still
go with the cheaper service provider?
Brad Allen
" prefer the cheaper provider?
Your thoughts are well considered and explained.
For the record, my thought is that price doesn't just include when you
are in a straight jacket and enjoying it; it also includes when you
can't do things because of that straight jacket, or when you die when
you attempt to circumvent that straight jacket. Therefore, the
straight jacket itself has a heavy cost, and the service that doesn't
come with a straight jacket could be lower cost even though if you
stay compliantly in the straight jacket version, the other service
would be "more cost".
Brad Allen
" way" to distinguish between right and wrong, I would welcome it.
" Then we can close the law libraries, dismiss the attorneys, and live
" happily ever after.
Sure (although this is an incomplete description): theft is bad and
accountability is good, and the goal is life and happiness, and
optionally for those who have a high quality enough religion (such as
reformed ones), religion.. Basically about three commandments.
I have a very, very complex view of theft, however. It consideres
everything, and is all-applicable. Theft of life is by far one of the
worst types of theft. Therefore, no need for bodily harm and murder
type laws, but essentially that's mostly just a nomenclatural
difference.
Why is theft bad? It destabilizes and takes away from purpose (see
above); note its inefficiency.
Accountability means if you steal from me, then you must be
accountable for that action, therefore accountability causes some
legitimate events which if they weren't part of an accountability
would themselves be theft. For instance, if I kill you because you
stole $2,000 from me in today's economy, that is a legitimate
accountability. With some more time, I could quantify various
relative levels of everything, for instance, accountability.
I'll throw that out there as an example.
" It's like that with any filter. Spam filtering started out simple
" (keyword), became complex (Baysian), and rapidly approached cryptic
" (AI). Same with OS identification. NMAP has turned into a really
" complex algorithm for guessing the server OS. I suspect that
" sniffing traffic for desired (good) and undesired (bad) traffic,
" will have similar problems.
Spying on traffic to see what it is once it was already put into place
seems like a bad approach. Categorizing (author, etc.) the traffic
sooner and more often would cause the ability to figure it out much
better. E.g., with SMTP, only allow from sources you authorize. If
you are getting a new source that you haven't authorized, it can go
into your list of "unauthorized sources", and you can study each one
to see if it is authorizable. Protections would be put into place
during that study as well. That way, you don't hunt spam; you hunt
non-spam, and deny the rest.
Rules like "no pictures unless the user clicks, and then only if the
pic is less than 100K" seem quite fine, too, or generated thumbnails
of each, etc.
John Slade
"© The OS/2 Guy ©" <TheO...@gmail.com> wrote in message
news:91ee04ab-c0c5-4350...@s21g2000prm.googlegroups.com...
> There should be no download caps whatsoever but it is coming about at
> the fault of all those posting here. We spent months complaining
> about Comcast's illegal throttling and reading of our private email
> yet so many said it was their right.
>
> Well, now you can pay them even more money and have your time online
> capped. Comcast will literally turn you in to copyright holders if
> you download any file that carries a copyright claim. Their goal is
> to appease the music and media industry and stamp out P2P
> downloading.
>
> It is no one's fault but those who sat back and said nothing. This
> always seems to be the case.
What can somone say? Nothing with letters or their mouth. What they
can do is vote for someone who will prevent ISPs from putting in caps.
That's the only power we have over utilities, the law and we get laws
through elected officials. I know people hate politicians, it's fashionable
to do so, but when you want something done, you have no better friend or no
worse enemy. Rememebr it was congress who voted to take DARPAnet and create
the Internet out of it in the first place. So before people start whining
they should start writting letters, sending e-mails and calling their
congressional representation. They figure that one person who takes the time
to do the above is worth thousands of like-minded people who don't.
If they go to a measured rate and restrict usage. It will kill a lot of
upcoming industries and also it will kill video and audio streaming over the
Internet. It's quite easy to stream 5GB worth of data in a day. I can see
other providers cashing in on the fact that they give you unlimited access.
This will eventually lead to much competition and eventually the big guys
will follow suit. If they don't then you will have content providers
investing in wireless and/or wired networks of their own rather than depend
on the current crop of big ISPs.
John