Wednesday, January 28, 2015

SIP Wifi Phone Idea

SIP Wifi Phone Idea

So we got cell phones that use cell towers, then we got SIP phones that use the internet, then we got cell phones with wifi that can use software SIP. Let's put it all together to get a cell phone with wifi that uses soft SIP, the internet and cell towers for phone calls from anywhere to anywhere.

Voice communication takes very little bandwidth (4kB/s) and the slowest wifi protocol can handle several of those simultaneously (1mb/s or 100kB/s or 25 x 4kB/s), and wifi routers typically can handle much higher bandwidth (up to 1gb/s or about 100mB/s or 1,000 x the slowest wifi protocol) than the lowest protocol allows. Some quick dirty math tells us a single home wifi router can handle 4,000 phone calls simultaneously by bandwidth alone. Of course there's data overhead for the various protocols and there there's data compression algorithms but that's the order of magnitude for voice communications and wifi capacity.

Furthermore, there's no reason a device cannot use multiple hotspots simultaneously (or discriminate between hotspots) using the existing system which allows it to use a single hotspot. Ever wondered how your cell phone knows which cell tower to connect to, when several towers are in range therefore your phone receives a signal from all of them? It's a method called generically signal discrimination, or in this case data discrimination. Two towers could broadcast on the same frequency, but they each have a specific data identification code in the datastream, and your phone has a way to discriminate between towers using this ID in the datastream. I'm just pulling this off the top of my head but that's the idea basically. Well, the same thing can be done with SIP/wifi. And in fact it's already done with wifi specifically where there's discrete channels within a range of frequencies.

Distributed network communications.

It all started with BitTorrent, but now every media, data, storage and streaming provider is gradually switching to this model. The idea is to distribute the load across several connections and machines so that no one connection or machine handles the total load and to increase the overall capacity at the same time. The same can be done with phone service.

Here's the first problem. How do you connect to your neighbor's wifi without first setting up a password on his router? Well now you can't unless you hack his router but that's not the idea here. When you're at home or at work, it's easy to do yourself and make a call this way. But when you're on the road or elsewhere, it's got to be set up independently beforehand somehow. For the router, it's pretty simple. Write a new firmware that allows this kind of communication without a password, or build routers with this firmware at the factory. Combined with this router firmware, set up servers that monitor those routers specifically for this kind of communication, then make this info available to phone providers for billing purposes. Additionally, subtract all of this data usage from the guy who's router you just used to make your call so he doesn't get billed for it by his ISP. All of this can be done already using the MAC address in your device.

Here's another problem. Do you want your internet to bog down cuz there's too many people making calls through it? Of course not, so the router must take this into account and limit the amount of bandwidth dedicated to this, or give you the option to just turn off this kind of communication so you can enjoy your internet without any problem.

What about video calls? That takes a whole lotta bandwidth. You could do that at home or at work, but maybe not everywhere else. I'll leave it at the for now.

Money. How much would all that cost? No more than it already costs. It would probably cost less because network resources would be distributed, therefore costs would also be distributed exactly like it happened with other existing distributed systems. Some wifi hotspots in Montreal for example are free and are set up to use the MAC address on your device (Ile Sans Fil for example). Cell phones like mine can use apps to make SIP/wifi calls (it's an Android phone). VoIP providers are many and offer super cheap prices for basically everything you want in a phone service ( for example).

The biggest problem is cell phone companies. They don't like this idea one bit. It takes away their power. Their power to bill you a hundred bucks each month just for a phone. Their power to bill you more for data, surcharge, usage, cancellation, opening a new account, buying a new phone, switching to a new 3-4-5G+ network, locking you into a multi-year contract, paying your bill in person (da hell kind of practice is this?!?), etc, etc, etc.

OK, so this idea isn't ridiculous. Every individual element is already done on some level, it's the whole package that just needs to be put together. And if I can come up with this idea and I'm nobody, then somebody else will come up with it too but this guy will be in a position to actually do it.

Martin Levac Copyright 2015

Tuesday, January 27, 2015

Public Mobile Contract Modification

Public Mobile Contract Modification

Here's the situation. Public Mobile modified the service contract, but did not notify me of any of it until several months later. In Quebec, any modification of "service contracts" like phone service for example must be notified directly to the customer 30 days before the modifications take effect. In addition to this obligation, the notification itself must include the clause before and after the modifications, the date the modifications take effect, and the option in case the customer refuses the modifications.

The specific modification is about the billing period, which used to be "monthly, or month-to-month, same monthly date", but now is "30 days", which means it will change every few months until 6 years have passed to come back to the same original date of signing. In fact, this modification reduces the value of the service by exactly 5 days per year, plus 1 day for leap years. To a single customer, it ain't much. But to a company with 150,000 customers, it's anywhere between $625,000 and $1,500,000 per year if not more. In financial terms, we call this a Big Deal. In Quebec legal terms, we call this a Big Fraud. Of course, not all of those 150k customers are in Quebec, but the result is the same because the price is the same everywhere and the reduction in value is the same too and it's the same company which profits from this reduction of value. "Reduction of value" is the same as "increase in price".

In the service contract, there's a clause that says "without notification". This clause in Quebec is null and void.

Here's the math.

Price before modification for 1 year service: 12 x $25 = $300/year
Price after modification for 1 year service: 12 x $25 + (1/6 x $25, or $4.17) = $304.17/year

Multiply this $4.17 number by 150,000 for company profits without any increase in service value, we get $625,000/year. Free money. And this number is the lowest possible since $25/month (actually it's per 30 days now) is their lowest price service, not counting all the surcharges customers incur from using data or long distance without the plan for example. This number could very well be multiplied by 2 or 3 if not more. Free money. Half a million bucks minimum. Each year. In Quebec, profits acquired illegally. Fuck you. You got caught. By a simpleton like me. And they call Public Mobile a "value" service.

Who owns Public Mobile?




Pour qu’un changement soit valide

Que votre contrat soit à durée déterminée ou indéterminée, le fournisseur de services doit respecter les règles suivantes pour pouvoir modifier votre contrat.
Il doit avoir inclus une clause dans votre contrat. Ce que doit contenir cette clause est prévu à l’article 11.2 de la Loi sur la protection du consommateur Vous allez quitter le site de l'Office de la protection du consommateur. Consultez la page Accessibilité pour plus d'information sur les contenus non conformes.. La clause doit notamment mentionner les éléments du contrat qui pourront être modifiés.
Le fournisseur doit aussi vous transmettre un avis écrit au moins 30 jours avant l’entrée en vigueur de la modification. Cet avis doit uniquement comprendre les renseignements suivants :
  • la clause ajoutée ou modifiée (le changement);
  • la clause avant sa modification;
  • la date d’entrée en vigueur du changement;
  • vos droits, si vous refusez le changement.

Requirements for an amendment to be valid

Whether your contract is of fixed length or open-ended, the service provider must respect the following rules in order to amend your contract.
The service provider must have included a clause in your contract. Article 11.2 of the Consumer Protection Act Vous allez quitter le site de l'Office de la protection du consommateur. Consultez la page Accessibilité pour plus d'information sur les contenus non conformes. indicates what this clause must contain. In particular, the clause must specify which elements of the contract may be amended.
The service provider must also provide you with written notice at least 30 days before the amendment takes effect. This notice must include only the following information:
  • the clause added or amended (the changes);
  • the clause as it appeared prior to the amendment;
  • the date the change will take effect;
  • your rights, should you refuse the amendment. 

Friday, January 2, 2015

Academic Performance

Academic Performance

Kids at school learn new stuff. This happens with three concurrent factors. Growth, brain function, and opportunity and support.

Growth is primarily a matter of proper diet. Brain function is also a matter of proper diet. Opportunity and support is all about the money and time devoted to academic logistics.

Growth happens because of growth hormone controlled by genes and all that. This can be interfered with - or allowed to - by proper or improper diet. An improper diet will stunt growth through interfering with GH primarily by the action of hyperglycemia and hyperinsulinemia. Take those out and we get normal growth. With normal growth we get normal brain function potential.

Brain function depends on growth first, but then a proper diet will allow normal function. For example, when people adopt a low-carb diet, they talk about this brain fog lifting and suddenly then can think clearly. Well, this also happens with kids. Imagine a bunch of kids walking around with a brain fog, then we lift the fog.

Opportunity and support is the stuff we do for kids to get them to school, buy books, clothes, good food, safety, study and practice time, etc. It combines with growth and brain function to allow academic performance.

It's a simple idea, really.

Martin Levac  Copyright 2015

Friday, May 16, 2014

Argument For Repeat Dieting - LC Included

Argument For Repeat Dieting - LC Included

Take a look at the graph showing the effectiveness-over-time in the A-TO-Z study. There's an initial sharp drop, then a slow rise back up. This means it's most effective initially, then effectiveness grows worse over time. Note the timeline. Peak is at around 6 months, and worst is at around 12+ months. That's for peak static weight, not peak on-going weight loss. The two are not the same. Peak on-going weight loss is at around 0-2 months, then grows worse after that. It's understandable. We have only so much weight to lose, and once we lose the bulk of that weight and get closer to minimum achievable weight, on-going weight loss slows down or stops altogether.

So how do I make the argument that repeat dieting is a good thing here? Simple. Initial intervention is effective only up to a point, then loses effectiveness over time, both for peak on-going weight loss and for peak static weight. Repetition is how we get better at anything. Therefore repeating the initial intervention allows us to get better at it every time we repeat it, therefore should allow this same intervention to maintain its effectiveness.

For those who have lots of fat to lose, repeat the instructions at interval to maintain peak on-going weight loss, so every 2 months. For those who have reached goal, and want to maintain peak stable weight, repeat the instructions at interval, so every 6 months.

That's the argument. It makes sense, right? Unfortunately, I've never seen it tested experimentally, so I can't actually say whether it's going to work or not. All the dietary experiments I've seen only do an initial intervention, they never repeat this initial intervention. And that's why all those experiments show the same effectiveness curve, sharp initial drop, slow rise later on. It's this curve which gave me the idea of the argument for repeat dieting.

In practical terms, this means reading the whole book again every 2 or 6 months, depending on where you are in your progress whether it's on-going weight loss or weight maintenance. This means doing all the things the instructions tell you to do, including induction phase for Atkins, carb ladder, etc. Granted, Atkins includes a maintenance phase, but it's different from what I'm suggesting. It's different because it's part of the initial intervention. Nowhere in the Atkins book does it say to read the entire book again, and to repeat all the instructions as if you did it all for the first time.

The A-TO-Z experimental study. See figure 2 for effectiveness curve.

Martin Levac

Copyright 2014 Martin Levac

Friday, February 7, 2014

War Thunder - Aimbot Official Statement is BS

War Thunder - Aimbot Official Statement is BS

I started playing that game a few days ago, and soon enough sometimes I had the distinct impression that one or two players were really really good, or they used an aimbot. The official statement from Gaijin:

If I was completely computer-illiterate, I'd believe that statement. But I'm not. In fact, I know exactly how those aimbots work, and I'm not a coder. I mean, yes, part of the official statement is true, a local program cannot interfere with server-side calculations, but that's not the only way to cheat. To figure it out, you gotta know how your PC paints the pretty planes on your screen. How does your PC know where to paint the enemy planes, what direction they're going, how fast, etc? Your PC must have access to this information, or you wouldn't see any enemy planes on your screen. What form does this information take? TCP-IP packets, but since it's a real-time game, it's mostly UDP packets. The point is your PC has access to all the information required to make an aimbot program, which would literally intercept that information, which when combined with the ability to take over your own plane's controls (i.e. mouse, keyboard), gives you a fully functional aimbot, and the server is none the wiser, because like the official statement says, it's impossible to interfere with server-side calculations, therefore there is no incentive to program the server with anti-aimbot functions.

Think about it. If you control your plane locally with your own mouse and your own keyboard (or joystick and whatnots), then you can use an automated program to do it locally as well. Even Windows has built-in functions that interfere with the mouse and keyboard, but Microsoft calls it "Ease of Access Center". It allows you to set certain options for the mouse so that it moves more slowly when you hover over an icon, or the keyboard with sticky keys, etc. It's all done locally, and Microsoft is none the wiser either. All of this is possible with War Thunder as well, since between the server and your PC, there's data packets that contain all the information required by an automated aimbot program which would aim for you, shoot for you, and literally make you the best player of the battle, every battle, all the time. Unless you play against another such player, then skill and a bit of luck will make a difference, just like it does when nobody's cheating, ironically.

The official statement only applies to hard-coded things like damage, armor, bullet spread (it's a cone of accuracy, which depending on specific algorithm used by the game, determines exactly where the bullet will fly to, and hit), plane speed, collisions, etc. But when it comes to aiming your mouse and joystick, that's all you, and whatever you can do on your PC can be automated to be done for you. So yeah, the official statement is BS, because of this part:

"...these "cheats" are based on local reverse engineering of the game client, and could not possibly affect actual multiplayer gameplay."

A program can be made to intercept data packets that contain information on enemy planes' position, direction of motion, speed, etc. A program can be made to automate mouse, joystick and keyboard functions. Combine the two and we get an aimbot. The server doesn't know a thing about it. For such a program to be detected, a local program must be used to scan the player's PC (the PC's main memory mostly) while he's playing the game. It's been done in many other games for years now with Punkbuster for example. It's a local program that scans the player's PC for certain things that are typically used by cheat programs, or it scans for the cheat programs themselves. Other methods include taking screenshots locally at intervals, and analyzing the screenshots for abnormalities like target boxes for example.

The point is it's totally possible to cheat in War Thunder, specifically to aim, which when you think about it, aiming is the single most important part of the game. You aim for everything, to shoot, to fly, to land, to avoid collisions, to stay in formations, to follow your wing men, to drop bombs, to evade enemies, or just to fly around for kicks.

So, were they really really good, or did they use an aimbot? Dunno, I'd have to see their screen when they play. Even without any abnormal visual clues (like target boxes for example) an aimbot produces a typical behavior, and once you know what to look for, it's unmistakable. This is mostly due to human reflexes, they're just not as quick as a computer. A program can react in single-digit milliseconds, while the quickest human on the planet can probably react in about 60ms. Furthermore, a program can predict just as quickly and with uncanny accuracy, so that when you look at a video of an aimbot, it's like there's no delay and it's always on target, and unlike a human it's super smooth. An aimbot uses information contained in data packets, and there's anywhere between 30 and a few hundred data packets per second, and with each data packet the aimbot will react quicker than the data packets come in, and make accurate predictions and corrections before the next data packet comes in. If I can see their screen while they play, I'm absolutely certain I will know, it's just too obvious. However, since it is really too obvious, smart coders will introduce random elements in their aimbots that simulate human behavior so that to the casual observer it's not so obvious. But to me these random elements are also just as obvious. Think of it this way, it's really really really hard (if not outright impossible) for a computer to simulate human behavior, and conversely it's really really really hard (if not outright impossible) for a human to react as quickly and as accurately as a computer. Actually, think of it this way, the best modern war planes (actually all planes, even jumbo jets and stuff) today all use highly sophisticated computers to aim, shoot, fly, take off and land, and basically do everything the pilot used to do, and does it a thousand times quicker and more accurately than any pilot. That's how good an aimbot is, and you can run it on your home PC. Modern planes even have automated collision avoidance systems that communicate with other planes, and adjust altitude of one of the planes so they pass each other at a safe distance.

Here's the argument for the official statement: The aimbot videos are fake. Well, OK. If they're fake, this means somebody did a pretty good job with a video editor, so it would be just as easy to reproduce that kind of video, and nobody's actually using an aimbot at any time. So I'd like to see a video that's been edited to make it look like an aimbot. But like I said, an aimbot produces a typical behavior, and this behavior is very very very hard to reproduce, whether in real-time by a human, or by editing a video. There is one way to edit a video to make it look like an aimbot: Using a program that detects the planes in the video with direct image analysis, and adds the typical things an aimbot would add (but the aimbot does it through data packet analysis instead). I mean if you do it manually, you'd have to be very very very good at editing video in real time, or frame by frame, to make it look like an aimbot. Better to make a program that automates the whole thing. Now if you can make such a program that edits a video to make it look like an aimbot, how hard would it be for you to make an actual aimbot instead? And if you made a program to edit videos for the purpose of extorting money from dumb folks who'd like to cheat, how long is that gig going to last? No, better to make an actual aimbot so you keep your customers happy and paying. One last thing about video editing to make it look like an aimbot. How can editing a video modify the human behavior of the player who recorded the video? It can't. And so the content of the video belies its nature, either it's just a human who plays and all the signs are there to prove it, or it's a human plus an aimbot and it's equally obvious.

The real reason Gaijin says it's impossible to cheat in War Thunder is because a) they can't prevent it yet (but hopefully they're working on it) and acknowledging it's possible to cheat scares paying customers away, and b) they're scaring would-be cheaters with FUD, i.e. the aimbot videos are fake so if you pay for the aimbot you'll lose money. I find it a bit paradoxical that they acknowledge they tried those aimbot programs and they work (i.e. reverse-engineering of local game client, i.e. intercepting data packets that contain information on enemy plane position, speed, direction of motion, etc), but then try to scare would-be cheaters by saying the videos are fake. Funny, that just occurred to me as I was typing this. Funny even more, the favorite expression of doubters on WT forum about this is "facepalm". Well, *facepalm*, Gaijin, you got caught in an embarrassing lie.

Aimbot Official Statement is BS

After all that said, the game itself is a lot of fun and it's free for almost everything except a few things like premium planes for example. Oh yeah, the matchmaker is broken but then that's just an inherent part of all matchmakers. Call it the unmatchmaker.

Martin Levac

Friday, January 17, 2014

RBC Liars and Crooks

RBC Liars and Crooks

I have an account with RBC, and they lie about the costs of the account, and they steal my money. The account costs $4/month, and allows 10 payments/month. However, the monthly fee is part of this limit, therefore if the monthly fee payment is the 11th, this one costs $1. So, they lie about the total number of payments included with the account, and they lie about the cost of the account. The real cost for the account is $5/month when making 10 payments on top of the monthly fee payment. The real number of monthly payments allowed is 9 + monthly fee payment.

I'm done fighting for my own goddamn money. I ain't gonna do shit about it. I'm just gonna make sure everybody I know hears about RBC. Good publicity for RBC.

Think about this for a second. Multiply this fraud by the number of clients. It ain't much for each client, but it's probably thousands for the bank.

Fuck em.

Martin Levac

Saturday, January 11, 2014

Update on my condition

Update on my condition

Today, I consulted doctor Lienchi Tonnu (spelling? I don't care at this point) to discuss various test results. I'm going to make this a short story, cuz a long one is gonna depress you too much.

To summarize, the doc did not listen to half of what I said (gave me the face you do when you "don't want to hear it" with both hands in the air and shaking the head "no"), did some token exam of one specific symptom and concluded "it's not an infection", then wrote a slip to see a doc in internal medicine. That's the second time I got a slip for a doc in internal medicine. The first one was so grossly incompetent that even a moron could see it. So even if I'm a moron, I could see the incompetence exude from that doc. Anyway, doc Tonnu's incompetence was slightly more subtle. You have to be very much aware of the signs. And after 5+ years and 8+ docs, I am very much aware of the signs of incompetence. Tonnu said something like "I don't know what you have. I don't understand your symptoms." (In French, though she could read the document I gave her, which I wrote in English)

I'm depressed. I'm angry. I'm desperate. I feel powerless. I have lost all hope. I feel like there's nobody on my side. I'm seriously considering slitting my throat. I never said that before today.

I have lost all my health. I was in perfect health 5 years ago. It's not my fault that I'm sick. It's not my fault that 8+ docs can't find what's wrong with me. It's not my fault. They don't listen. At least that's the impression I get, with one exception. The doc who did the stomach exam. I forget his name. Whatever, doesn't matter, he can't do fuck all about it. He did write a slip for an infection specialist at Hotel Dieu. I'm waiting. That's all I do. Wait. 5 years. 8+ docs. My health declines each day. I feel like crying. I only feel this way when I feel totally powerless.

Help me please.

If you can get things moving quickly so I can consult docs in days instead of months, you can help me. If you can get me prescription meds without a prescription, you can help me. If you can make this public to get things moving, you can help me. If all you can do is tell me "don't give up hope" or some soothing words that have exactly zero tangible effect, you can't fucking help me.

I was so happy 5+ years ago. I was playing golf every day. I was lifting heavy weights at the gym. I was eating with a healthy appetite. I had zero health problem. I was strong, bright, slept like a baby. I did not smoke (I quit in 2006 to improve my golf game, started again a couple years ago due to depression and declining health and stupid docs). I was much leaner than I am today. I took care of myself because I could.

I have this conversation in my head with the docs. I explain and they listen and they do something about it. Doesn't work in real life.

Are you depressed yet? How many more years do I have to wait before I get my health back? How many more docs do I have to consult before I find the smart one who will find what's wrong with me? You gotta understand that every time a doc gives me a slip for a specialist, that's another 3-to-6 months to wait, and then another month or two to discuss the test results, and so forth. Can you get things moving quickly? Do you have the power to do anything for me?

Would you offer me a job? I'm unreliable. I can't sleep properly. I wake up at random hours of the day. I get sleepy at equally random hours. I suffer from chronic fatigue. I probably have a contagious infection of an unknown nature. I'm cranky. I have no patience. You know my symptoms. I asked doc Tonnu to fill out a form for Securite du Revenu for temporary medical problem. She refused. Her reason is "there must be a diagnosis of something specific". Well, why don't you fucking find out what's wrong with me so you can fill out the fucking form. But you don't listen to what I say. You don't do proper exams. You don't fucking care about my health. You still get paid for the consultation. That's how the medical system works in Canada. The doc sees you, he gets paid, regardless of whether he finds what's wrong with you or not.

I consulted a ORL, he referred me to another ORL. That's a fucking gimmick. He's just giving money to his buddy. And his buddy couldn't find his own foot, but he wrote me a slip to that incompetent doc in internal medicine I was talking about. That stupid doc got paid too. They all got paid. I'm still sick. My health continues to decline. They still got paid. Is there even one of you who wants to find out what's wrong with me so you can fix it? Email me. Give me an appointment next day. Give me a real solution. Or fuck off. I got no money. I can't pay you. You gotta do it for free, or get your money from the system somehow.

I want to try one treatment.

Clarithromicin or equivalent


Doxycycline or Minocycline


Metronidazole or Rifampin

All 3 together for at least a month to see if it has any effect. If it does, then discontinue after the month and see if symptoms return. If symptoms return, continue treatment for at least 1 year.

Do any of you docs out there want to prescribe me this treatment? Do you understand why I want this treatment? Do you know what condition warrants this treatment? If yes to all, email me. Or fuck off. I'm done with tests. I'm done with waiting for you to get paid and for my health to decline. I want that treatment today. Call it a trial by fire. The risks? Compared to all the symptoms I suffer from now? Ya, I'm OK with the risk of dying. In fact, I decided that I want to become a professional golfer, or die. I did everything I could to get better. I practiced hours every day. I became the best player at the club. I scored par for the course, better than everybody except an actual honest-to-goodness pro who did the mini-tours in the US. He scored 2 under. I'm that good. I have the potential to be much better. When it comes to golf, I know my shit. I wrote lessons based on my own technique. I know how to put it into words so it's easy to teach and easy to learn. My technique is so good that even though I played only twice last season with zero practice, I still had it, though putting and chipping requires constant practice so I fucked it up with those shots. I know how to get good and stay good. I taught myself. I'm not bragging. I'm telling the story of how I decided to become a professional golfer, or die. So I'm OK with the risk of dying from a treatment if that treatment has the potential to give me back my health.

And if that treatment has no effect, then I'll figure out another treatment to try. I'm OK with the risk of dying. I'm OK with death.

Help me please.

Martin Levac  15:50 2014-01-11