Mobile Internet

[skinorth]

@allend:

you state:

"The hub (W35) reports data usage, but by the calendar month and Rogers says explicitly that their billing measurements do not correlate with what the Hub reports.  Any time I have looked at the Hub report, it has seemed chaotic and irrelevant....."

I have posted on this elsewhere, but let me put forth an excerpt from page 31 of the Ericssen W3x Users Guide:

"Note: The data size and packet counters have the upper limits of 4 GiB
and 232 packets (more than 4 billion packets). When these limits
have been reached, the counters wrap around to zero."

The Users Guide is silent on how these counters are synchronized to a monthly period, be it the calendar month, or perhaps a month since the device was started/restarted.

Thus, despite the mention on the Internet Web Page of the W35 User Interface, of Previous Month and Current Month, the implication of the above quote from the Users Guide is that the counters will keep on increasing until they wrap back to zero when they reach 4 Gbyte.  I have not checked what is possible through the CLI user interface which is inaccessible to us in any event.  I also do not know if there is a backdoor to the W35 which Roger could be, or is in fact, using to reset these counters at whatever time suits them.  Somehow I doubt that this is happening.

I have not attempted to track the progress of the W35 counters to determine what patterns, mysterious or otherwise, they are displaying.  It would be interesting to see what the correlation is between Ericssen's version of data volume usage and Rogers version of the same thing.

Of course, the W35 Web Page does include the following caveat as a CYA:

"Note: Approximate figures only. Please refer to your bill for final amounts and charges."

skinorth

skinorth

[allend]

Thanks for that.  I had just figured that using the hubs would be dead-simple and that I would not have to worry about data usage, however when I figured it all out and also that I had no way of knowing where I stood, I was quite flabbergasted.

I find my forthcoming end of billing a really good example of the dilemma, since I have already run one over in spite of my most honest efforts to manage my usage and I dare not use the other for fear of incurring two $10 charges for overage instead of one.  I have two days left this month and could be paying $10 for a few bytes  -- twice.

I have, however solved the problem,  A while ago, I bought a Galaxy Tab and an iPad, intending to compare them and take one back.  As it turned out, they both have their niche and I kept them both.  My wife likes the iPad.  I find the Tab small enough to carry everywhere.  I bought them both with the 3G+ option , and one was locked to Bell, as it turned out when i went to the US and tried a T-mobile SIM and an AT&T SIM in it.  I had no time or inclination to root it and unlock it, so I just used the Wi-Fi for the time being.  I had placed the Rogers SIM into the iPad and was immediately able to purchase data, but didn't since I had no need.

Today, I decided to try to get the Bell Galaxy Tab going on 3G+.  That process is a hassle compared the the comparable Rogers sign-up -- I had to phone to activate the Bell SIM, then it took me about 6 or 7 times entering a ton of info and finding their system was down or I was locked out and then phoning support, going around the Merry-go-Round and listening to music that is not music, then spending an hour with an nice young lady troubleshooting their sign-on process.

Anyhow the upshot is I bought a 5 GB/one month pay-as-you go plan.  Then I got smart.  Not wanting to use my hubs since they are so close to the price escalation point and so close to my month-end, and having already bought the 5 GB, I decided to try the mobile AP feature on the Galaxy Tab.  I set it up in a minute, and connected two computers and the iPad to it.  (My wife's, a Vista machine, is having a problem, though).

At any rate, I ran a speed test and although I see only about 1 Mb/s down,. it is doing 2 up, and my computers seem faster than when on the hubs when the tests show 6mbs.  Go figure.

Anyhow, I could have done the same with the iPad on Rogers, more easily than on Bell, and will activate a plan on it, too when I need to carry it away from home.  I am certain it will be much simpler than the bell experience.

In the meantime, my problem is solved for this month.  I don't care about having accurate data right now on the Hubs because they are unplugged until after billing.

By next month, let's hope that Rogers has had someone put up a decent usage meter for Hub users on the web.

[skinorth]

@allend:

if your wife is using Vista be sure you check out the following URL:

http://support.microsoft.com/kb/928233/en-us

This corrects a Vista DHCP-related problem.  It was what I needed to use my Vista system successfully with DHCP and the W35.  I suspect other devices will also be a problem in using DHCP from them to a Vista system.

skinorth

[allend]

You read my mind.  I was just researching that when the annopuncement email arrived.  Thanks.

BTW, I just checked and the Bell Galaxt Tab I am using on a prepaid plan (with renewal turned off) has an easy to reach web page that reports the data use to the minute and to two decimal places!  

It can be done.

[allend]

Well, so far my Vista machine does not connect to the Galaxy Tab, even with the M$ fix installed.  No matter, that machine is not used much.

The Galaxy works well, but the AP has to be restarted after a shut down and interestingly, each time the AP is invoked, its name changes.  Another interesting point is that while I have to use a 24 dB yagi on the hubs to get consistent service and no drop-outs, the Galaxy Tab is chugging along, sitting on a table in the middle of the room with no antenna.

---

Just to keep the usage report story up-to-date, this morning the Rogers website shows our usage at 2.26 and 3.06 GB.  Whether this is up to date or not, I have no idea.  The Hubs are currently off and I am using the Galaxy Tab as an access point, since it has quite a bit of cheaper data  available and I don't want to risk running the hubs into another $10 for another 2 GB chunk of data which will expire in a day or two.

---

As for setting up a router, I have an eligible Linksys router lying around and it is just a matter of going through the downloads, flashing, etc.  but I think that going through that process would be useful only for curiosity purposes.  I suspect that monitoring usage locally might well just show that Rogers bills us for *less* data than what we might measure on the link to the tower due to nulling out of handshaking, pings and maintenance traffic.  At least I hope so.  Your results should be interesting.

I doubt that you will find that we are being cheated, because I don't think we are, although in my dealings with *Bell* over many years, I did find errors more often than not -- and only a few in my favour.  That's why I fired Bell as a cell provider and will only deal with them on a prepaid basis and with intense scrutiny of their billing.   I can't recall ever having such issues with Rogers when they were my cell provider, and my only issue with them is the totally unsatisfactory manner in which I am able to see their data measurement in real time, as is my right and the custom in Canada for any good sold by measures.

---

I am a supporter of UBB and think it makes sense that those who use a lot of a commodity should pay more.  I have some concerns that UBB can be abused and is in some cases.  The present "plans" for hubs is an example in that the amount per GB is not consistent and that the chunks are sold in blocks which are too big and which also expire.

To me a UBB scheme should be simple: a flat hook-up monthly fee to cover the fixed costs, then a reasonable charge per GB from zero on up.

On some systems which have limited capacity like cell systems, usage caps or penalty rates beyond some level of consumption may well be necessary to ensure that everyone has access and that the system does not get plugged by a few users who may happen to have a stronger connection.  It seems that even time of day surcharges would make sense, but that adds complexity and impairs predictability and transparency and also can create suspicion of cheating, just as the lack of transparency and apparent randomness in our Hub data reporting does.

If we are able to relate our activities to a data meter from the provider in real time, rather than being surprised hours or days later by a sudden jump in the reading, people would be less likely to suspect monkey business.  It appears that bell has provided such data accounting for the Tab, complete with warnings by email as the data depletes and runs out.   I think Rogers does the same, but have not tried it.  I'd love it is Rogers would send an email report when we cross specific usage points for the hub.

So, UBB is here and here to stay in spite of FUD campaigns against it.  I listened to Spark the other day and although many of the issues were covered, a number of important points were missed, just as they are elsewhere in the media.  One of the points missed was this entire matter of real-time measurement and transparency, and also the need for standardized and regulated definition and supervision of data measurement, if data is to be sold by the measure.

UBB should guarantee that the systems expand and that the cost of data continues to go down, provided that there is meaningful competition, and that is another question.

---

We have to ask this, if Canadians are being overcharged, where are the excess profits going?  Bell and Rogers are publicly owned companies, and after the respective managements have taken their lion's share, any profits go to building more capacity (which drives down prices), dividends to shareholders, and taxes.  We all benefit from increased speed and capacity, and from expenditures of the tax money collected.  Further, most of us are shareholders of Rogers and Bell or both to some extent through the CPP investments, private pension funds, and in many cases direct personal investment in shares.

---

As far as political action is concerned, in many ways we need less regulation, rather than more, since it was political action that created the oligarchy we see and which limited competition and excludes non-Canadian participants who wish to inject large investments and drive down prices.  Political action, unless targeted to one specific issue such as demanding accountability and transparency under weights and Measures, can very well backfire.

[skinorth]

@allend:

You gotta love UBB!  But I also love technical transparency.

As you have so admirably documented, transparency is not part of the picture at this time.  You ever heard of the saying about the foxes watching the hen house?  Yeah, I'm talking about our "friends" at Rogers.

What you and I, and what we all need to do is what xplornetsucks suggested in a previous posting:

"As far as tracking data for multiple computers connected to hub, i think the logical thing to do is add a DD-WRT/Tomato enabled router to you rocket hub via RJ45/wired link.. and let it handle all your wired and wireless connections."

Technically, we need to set up our connection to the Rogers network to include the totally essential feature of a utility connection, the equivalent of the kilowatt-hour meter. This will enable us all to know with certainty that the billings we receive are based on truth and reason.

So, let's put together a joint project to instrument our RocketHub installations with an appropriate router and "tomato" firmware.  We all need to document our efforts and the results we get so we can spread the knowledge around for the benefit of all.  We will likely need to start a new thread in the Forum to cover the issues related to this unique technical aspect of the use of the RocketHub.

So let's get started!

skinorth

[skinorth]

I would like to take the data volume/charging issues discussion a bit further and compare it with the power utility volume measurement and charges most of us deal with regularly.

In the first place, common practice is that we are billed for electrical power usage based on what we use as measured by the kilowatt-hour meter installed where the power grid connects to our premises.  Kilowatt-hour meters are designed and tested for accuracy, and can be easily read by both the power utility and the customer.

At one time I took a course on heavy power, and learned that the power grid has its own internal power losses.  In the power grid these can be caused by systemic factors which cannot easily be avoided, as well as by engineering issues, routinginefficiences and equipment failures which can be improved or corrected to reduce or eliminate the power losses.

OK, so get the point, you say.  And well you should!

The point is this: the network infrastructure used by Rogers, wireless, wired, broadband, etc., has many features analogous to the commercial power grids.  It has systemic factors such as protocol overheads which cannot be avoided.  They are a feature of the technology.  They are fixed and relatively easy to quantify, and hence to include in your costing algorithm.

But the network infrastructure also has other losses and inefficiences within its fabric.  The network will have equipment that fails intermittently, it will have routing inefficiencies that affect the user, and it will need regularly to retransmit user data due to the nature of the media used (e.g., wireless) or due to other complex issues.  These are much more variable and hence more of a challenge to include in a costing algorithm.  They are also directly affected by how well the network is designed, implemented and maintained.

Anyone who has ever worked in the management and maintenance of a complex network environment has experienced  that network failures constantly happen.  The MTBF (Mean TIme Between Failure) numbers for equipment are the formal definition of statistically how often failures will happen.  You get lots of equipment out there in a large network, and you get lots of failures.

So, Rogers gets to design, build and maintain the network.  We the customers get to use the result of their efforts.  And we get to pay for it.

The issue is who pays for what.  I want to pay for what I use.  I want those charges to be reasonable and proportional to what I use. I do not want to pay, nor should I have to pay, directly for inefficiencies and failures in the network which are under the control of the network provider.  Rogers needs to have the incentive to implement and maintain the network efficiently.  That is entirely under Roger's control.  The costs need to be paid by the customers/users.  But Rogers needs to have the incentive to make it less costly and more efficient.

So, Rogers (the whole industry, actually) needs to figure out how to put that Gigabyte meter at the point where their network connects to my home or place of business.  At that point we can discuss whether their charges per Gigabyte are appropriate.

In the short-term that router with DD-WRT/tomato may be the only practical way for us the user to implement a "Gigabyte meter" for our Internet connection.  Information is power, and once we have that information we can determine the appropriateness of the volume-based charges we are sustaining.

skinorth

[allend]

I very much appreciate your last two posts and agree with your points, which you make very well. 

One thing we overlook or discount, though, is that we already have a GB meter -- which we paid for  -- installed at  and integral with the entry point.  When we signed up, we expected it would be accurate and reflect what we would be paying, but soon decided it was useless as implemented.  A short period of observation convinced me, that this data print was somewhat random and unpredictable and did not reset when our billing period ends. 

(That is another topic: why in the world does our data allotment expire?  That seems incongruent with the concept of UBB.  Why are we not just paying a flat fee for connection and then a variable amount per unit consumed?  Moreover, do we know the exact moment that our data rests each month?  Midnight?  Midnight where?  Toronto?  Calgary?  Greenwich?)

Additionally, the readings available -- such as they are -- on the hub's built-in router are presented on a page which times out periodically and requires entering a password to re-view and requires revealing the admin password to anyone the owner wants to monitor usage (not a clever feature).   Additionally, web browsers which normally remember usernames and passwords do not automatically fill in credentials on local sites such as 192.168.1.1. making viewing over time a chore.  All in all, the reporting of usage, both local and remotely at Rogers site, is far from ideal, not legal for trade, and not what we expect.  There is no need for that data to be concealed or protected  it should be presented as a widget or some such method, like my desktop calendar or the weather widgets.

BTW, these hubs are programmable.  I have no idea whether they can be updated over the Internet connection or not.  I suspect they can be since I had Rogers Portable Internet previously and the techs installed new firmware on the modem, -- they said -- while it was turned on and I spoke to them by telephone.  So, if Rogers spent a bit of time working on the hub software, maybe they could deal with these issues for us and make the hubs do what they were designed to do and present us with meaningful and accessible data, resetting when the so-called 'plan' does until such time as more sensible billing system is mandated.

Since the unit is programmable and was delivered pre-programmed to Rogers specifications (with some Ericksson features unavailable) it seems to  me that Rogers could very well program them with an upgrade to reset at the appropriate time so that it would track data flow in real time, and possibly program in the same parameters that Rogers uses for billing.  If they roll over at 4GB as I heard but have not verified, perhaps that could be changed in the programming.

As for installing routers with Tomato software for monitoring, that is a good idea in the short term for purposes of confirming how Rogers measures data, but it is not a long term solution for most.  I may do as you suggest, but I travel a lot and one hub travels with me if I am in Canada and the other stays home with my non-techy wife.  I bought the hubs because they have the router built in. 

I am now seeing that my Galaxy Tab may be a better Internet connection solution since it has a more transparent 'plan', is smaller and runs cool; and it is a computer in its own right that comes on instantly.  I should look to see how good a firewall it is, though.  I have Lookout installed, and am assuming that the Tab acts as a NAT router and firewall, but I realise as I write that this is just an assumption and bears investigation.

I'm realising now that taking the hub with me may be a thing of the past since I have discovered how cheap and transparent using the Galaxy Tab as a source of data for my netbook can be.  It is much smaller and cooler running, and I know exactly how much data I am buying and using.  I see the Tab usage data does not update in real time, now that I have watched it a bit,.  I have not figured out how often it updates, but I wonder how Bell cuts off the data when my prepaid allotment runs out since I have turned off auto-renew on the 'plan'. I will see when I have used the whole 5 GB chunk and that will be a while unless I watch some movies of download another Linux distro or two.

Having activated a data pan on the Tab and finding out how easy it is to share that data may well make one of my hubs redundant.   The Tab's data stream is slower than the Hub stream (when the hub is working well and the towers are not plugged), but it is fast enough for my normal purposes.  Anyone want to buy a Rocket Hub?  Or two?  I could do the same as I did with the Tab and put my wife's iPad on a Rogers pay as you go plan.  Then we would not need either hub.  I'm on the hook for another year and a half, though, for both.  Wonder what it costs to cancel early?

It's premature to make such a move at this point, though, since I have not used this setup for more than a day or two and that is not enough time to find the kinks.

[skinorth]

@allend:

you stated:

"Additionally, web browsers which normally remember usernames and passwords do not automatically fill in credentials on local sites such as 192.168.1.1. making viewing over time a chore."

What browser are you using?  I use Firefox which does that no problem at all.

Your comment on UBB and the data volume allotment rolling over on the monthly boundary is apropos.  In a pure data volume usage scenario you would not expect otherwise.

The Hub is programmable, and the Linux CLI is documented by Ericssen (see other posting of mine in this Forum).  However Rogers has elected not to make the CLI accessible to us by not publishing the password.  I can understand the support issues that would ensue if users could go into the CLI.  However, particularly for troubleshooting, CLI access would be invaluable.

Assuming that the W35 firewall allows access through on the telnet port, Rogers (or anyone, for that matter) should be able to access our W35's for whatever suits their fancy.  So keeping password information hidden is also good for the customers' security.

In correspondence with Ericssen it is made clear that they cannot reveal how to access the CLI for contractual reasons without Rogers' permission.  W35's for other vendors have not been so obfuscated.

skinorth

[allend]

My main browser is Maxthon 3, but I also use Chrome, Opera, Firefox, Safari and occasionally Internet Exploder.  I'll have to check into what you say.  I think I had this problem with several browsers, but I'll have to try Firefox next time I have a hub plugged in and see if I can get it to auto-fill the credentials and also auto-update the page.

Just as you replied, I was writing that I have so discounted the built-in data metering in the Hub that I had almost forgotten about it until we got to discussing this question. 

As a result, of our examination of this matter, I think I'll take another look and document the hub measurements and perhaps compare them to the numbers on the Rogers site.  That latter activity will be difficult since the web update seems random and bears no timestamp, but some trend should emerge, especially if a hub is plugged in a while, then unplugged for a day. so that the web reading has to be current.  (That is a big assumption since I have seen the same number remain frozen on the site for days, although the hub was active all the time). 

I can do this measurement fairly easily since I have an extra hub and only use one at a time when home.  I have them both unplugged at present to prevent accidentally going over the 3 GB and getting billed an extra $10 for what might only be a one day or a portion of a day since I am at the last day of my billing cycle.

I observed the local measurement on the hub admin pages for a while when I first got the hubs and decided that the numbers were useless, but did not do a rigorous analysis.  I may do that now.  We'll see.  I'm about to go travelling again. and will not be in place long enough to complete a series of measurements.   (I'm delighted to discover, BTW, that i may not need to take a hub along since hubs are bulky and subject to damage.  I had one exchanged by Rogers early on when the power cord frayed and broke due to the normal flexing that takes place when bundling up the unit for travel. and I have to treat the units with abnormal care)