Almost a year ago, I built a new computer. It was a budget machine that cost around $800, and it's still going strong today. I can play all of my games at their highest detail, including a graphically intense one released this year, and they all run silky smooth. For details on my setup, as well as basic advice on PC building, refer to my first blog post.
Are you in the market for a new gaming rig? If so, read on. I've done some studying of what's currently available, and I have great news for you. First, prices have gone down. Second, AMD is back in the game, so you have more options for processors and video cards. Third, there's even a new, reasonable option on the Intel side for a slightly larger budget. Last year, there was really only one viable option for a budget gaming rig. Now, there are more.
In this post, I'm going to analyze five different combinations of processors, motherboards, and RAM. Remember, these are what make up the core of the system; everything else is cake. I'll get to the other parts in another post. These other parts include, but are not limited to:
- Video Card
- Hard Drive
- Case
- Power Supply
- Optical Drive (DVD)
- Heatsink and Fan
I estimate that these other parts might total about $450 if we're careful, so keep that in the back of your mind when considering the entire cost of your new machine. I'm assuming you already have a monitor, keyboard, mouse, and other external things.
This is a long post. If you don't care about all the details, I recommend you just take a look at the first build and the last build. Those are the two most viable options in my opinion.
Now, onto the meat!
My first CPU/MB/RAM combo is an AMD build. It's the least expensive, but it's still a competitive performer when compared to the others. Here's the breakdown:
Processor: AMD Phenom II X3 710 - $119
Motherboard: ASUS M3A76-CM - $65
RAM: Patriot Viper 4GB DDR2 800 - $27 (after $25 rebate)
Shipping: $11
Discount: $25 for Newegg's CPU/MB combo deal
Total: $197
Estimated entire build: $647
Like I said, AMD is back, and its Phenom II line is right up there with Intel's Core 2 line. The triple-core Phenom IIs are giving the Wolfdale processors a run for their money.
You'll have a hard time finding the Phenom II X3 710 on benchmark sites. The reason is that the 710's big brother, the 720 Black Edition, is more popular. However, the only differences between these two processors are the clock speed and the unlocked multiplier in the 720.
The 710 is clocked at 2.6GHz while the 720 is clocked at 2.8GHz. This gives the 720 a very slight performance advantage. AMD will unlock multipliers in some of their processors, and these are referred to as Black Editions. This makes BE processors a little easier to overclock. However, all the Phenom IIs are great overclockers and you don't need a Black Edition to get a modest boost in clock speed. So, what is the performance difference between the two? I found a spot an AnandTech that gives a nice comparison.
For the games, which are at the bottom, the difference is no more than 5%. If you want to get the 720 instead, you'll have to pay an extra $25. The processor itself costs $20 more, but your combo discount will also be $5 less. The Phenom II X3 710 is a gem that will give you way more bang for your buck than anything else I will discuss here.
Beware! Never forget that gaming performance is most heavily impacted by your video card, not your processor! The benchmark sites that measure CPU gaming performance do their tests in a way that is not practical for actual gameplay. They run the games at very low detail, shifting the bottleneck from the video card to the processors. This makes for good CPU comparisons, but it's not a realistic example of actual gameplay. When you get your shiny new video card and run these games at much higher detail, the difference in CPU performance will be practically non-existent.
If money is tight, anything more than the Phenom II X3 710 is a waste. You won't notice any gaming difference with other processors as long as you have a video card that's more than $70.
Regarding the motherboard, it's inexpensive but it has good reviews. (I ignore any product with less than 4/5 eggs.) Newegg's combo deal makes it an absolute steal. It is PCI-Express 2.0 compatible and has HD Audio, as does every motherboard in this post. It should also allow you to do some modest overclocking, as should every motherboard in this post. You should note that it does not support firewire, and it only has one PCI-E 2.0 x16 slot. It also does not support DDR3 RAM. If you want an AMD motherboard that has firewire, more than one PCI-E 2.0 slot, or DDR3 support, they are available. However, they also come with a higher price tag. Please leave a comment of write me an e-mail if you want recommendations that meet your specific needs.
The DDR2 RAM is an awesome value. It operates at 4-4-4-12 which is about the fastest you can get with DDR2 800. 4GB is more than enough for today's software. Simple and sweet. Moving on.
What does Intel have to offer for a budget gaming rig? Well, it still looks similar to my own setup. Here's the breakdown:
Processor: Intel Core 2 Duo E8400 Wolfdale - $168
Motherboard: GIGABYTE GA-EP45C-UD3R - $105 (after $15 rebate)
RAM: Patriot Viper 4GB DDR2 800 - $27 (same as RAM above)
Shipping: $3
Discount: $15 for CPU/MB combo deal
Total: $288
Estimated entire build: $738
Last year, I paid $190 for the same processor. The price has dropped a bit, but not significantly. It still seems to be the sweet spot on the price/performance curve. If you want something that costs less, you'll end up with a crippled CPU that has not only a lower clock speed but also a smaller cache. If you go up, you pay a premium for a very slight clock increase.
For performance comparisons, you can use the same tool I did for the 720/710 comparison. The E8400 seems to be about a 10% gaming improvement over the 710. However, I also noticed an interesting benchmark article that showed the Phenom II x3 720 outperforming the E8400 in Far Cry 2 and Unreal Tournament 3.
You'll also notice that the 710 and 720 outperform the E8400 on some things like media encoding. The E8400 is a dual-core processor, and software is slowly changing. It still has the edge in gaming, but other areas are not so clear. As more multi-threaded applications are used, triple-core and quad-core processors will have a greater advantage. Also, consider that these other applications will not be relying on your video card as much as games, if at all. Therefore, you're actually more likely to notice the performance increase.
There really aren't any viable motherboards with a price similar to the one in the AMD build. However, this GIGABYTE board does have some features that the other doesn't. It has firewire and supports both DDR2 and DDR3.
The RAM is the same as before. However, if you really want DDR3 you can get this set for a total of $25 extra. Since the price of DDR3 has dropped significantly from what it was last year, this actually isn't a terrible option. However, the performance increase won't be much more than 5% if you're lucky.
Thanks to AMD's options, Intel has a great competitor at this price point:
Processor: AMD Phenom II X4 940 Black Edition - $190
Motherboard: GIGABYTE GA-MA790X-UD4P - $110
RAM: Patriot Viper 4GB DDR2 800 - $27 (after $25 rebate)
Shipping: $3
Discount: $30 for CPU/MB combo deal
Total: $300
Estimated entire build: $750
This quad-core processor seems to be neck and neck across a spectrum of gaming benchmarks from different sites. However, since it has twice as many cores it seems to have the edge in other applications.
This motherboard doesn't support DDR3, but it does have firewire. It also has two PCI-Express 2.0 slots so you can run Crossfire or Physx. NVIDIA doesn't like AMD, though, so the AMD boards can't do SLI.
Overall the comparison between the X4 940 and E8400 is too close to call. It's really up to your personal preferences.
Now, let's suppose that you're on Dave Ramsey's Baby Step #5 or further. (You know who you are.) There's a new option that will give you a relatively future-proof machine for a price that has recently become somewhat affordable. The following may be worth considering:
Processor: Intel Core i7-920 (Nehalem) - $280
Motherboard: GIGABYTE GA-EX58-UD3R - $185 (after $15 rebate)
RAM: OCZ Platinum 6GB DDR3 1600 - $75 (after $20 rebate)
Shipping: $8
Discount: $25 for CPU/MB combo deal
Total: $523
Estimated entire build: $973
Core i7 is Intel's newest architecture, replacing the incredibly successful Core 2. These pricey processors weren't exactly designed for gaming, though the most expensive ones still give the best gaming performance. The 920 is the most inexpensive Core i7 CPU available right now, and its gaming performance is actually pretty close to the other processors listed already. Core i7 was really designed for multi-threaded performance, and that's where it excels.
If we look at the Tech Report benchmarks again we'll see that this processor is very slightly outperformed in games by the others I've mentioned. However, notice that they did comparisons with the Core i7-965 Extreme with Hyperthreading enabled and disabled. You can do the same thing with the 920. It looks like this can give you perhaps a 5% boost to your gaming performance, putting it about exactly on par with the others. Also, the 920 totally owns in most of the non-gaming benchmarks. If you plan to use your machine for more than just gaming, you may want to give this build some serious consideration.
Don't make a judgement just yet! If we look at Anandtech, they seem to show the i7 beating out the other processors. It's hard to tell just what's happening here, so I'll investigate it further and post an update.
Anyway, the motherboard is also more expensive than the others, but it definitely has more to offer. It supports firewire, triple channel DDR3, and has two PCI-E 2.0 x16 slots for Crossfire, SLI, or Physx. Yet again, a combo deal helps to curb the hefty price tag.
Since the motherboard uses triple channel, we're getting 6MB of RAM. Just like the 4GB of DDR3 I picked out for the E8400, this ram operates at 7-7-7-24. This is about the fastest you can get at this bandwidth. Make sure you use a 64-bit operating system so your computer uses all the memory! Also, note that this set of RAM has several combo deals with video cards. You may be able to save up to $25 if you want the right card. I'll consider this in a later post when I dig into the other hardware.
By the way, don't assume that you'll be able to upgrade your CPU if you get a more advanced motherboard right now. If you only upgrade about every 3 years, your board will probably be outdated no matter what you get. It's generally a bad reason to spend more on a motherboard.
Is there an AMD alternative at this price point? I'm glad you asked. Take a gander:
Processor: AMD Phenom II X4 955 Black Edition - $245
Motherboard: GIGABYTE GA-MA790XT-UD4P - $125 (after $15 rebate)
RAM: OCZ Reaper HPC 4GB DDR3 1600 - $55 (after $20 rebate)
Shipping: $8
Discount: $15 for CPU/RAM combo deal
Total: $418
Estimated entire build: $868
Well, what can I say? This build puts Intel to shame, at least as far as gaming is concerned. The motherboard supports firewire, dual channel DDR3, and Crossfire or Physx. This build is basically capable of anything the i7 build is capable of, it costs less, and it seems to usually beat out the 920 in gaming performance if you look at both the Tech Report and AnandTech. I the i7 still has a clear performance advantage in other applications, but that's not really our focus here.
By the way, you could even couple this processor with the motherboard and RAM from the X4 940 build for a total savings of $53. The only real difference would be a choice between DDR2 and DDR3.
So there you have it. These are five different builds based on the three main CPU architectures available. Each one has its own advantages and disadvantages. Each one is also highly customizable, and I'm willing to help you make adjustments to meet your specific desires.
All of these CPUs are manufactured using a 45nm process. They are all great for overclocking. I strongly suggest that you give it a try, and am willing to walk you through the process.
Tuesday, May 19, 2009
Thursday, April 2, 2009
Little Ones and Lockjaw
Some concerned family members have asked me to provide evidentiary support for our decision to avoid vaccinations, so far, for our children. Understanding this valid concern, I'm more than willing to provide some. There are many vaccinations that can be discussed, but I'll just focus on tetanus right now because it was mentioned specifically.
I have zero medical training. I'm not responsible for your medical decisions. If you make any judgements based on this post, your lockjaw isn't my fault.
This is hardly a scientific study, but rather my own analysis of other studies. I'm going to show a blatant disregard for things such as significant digits and formal terminology. Where assumptions need to be made, I will tend to grant the benefit of the doubt to support of vaccination. I'm open to corrections, so if you have any please leave a comment or write an e-mail. Finally, I'm going to avoid wasting time defining everything. There is plenty of information available on what tetanus is and what the DPT vaccination is.
Generally, when considering whether to receive medical treatment, an analysis of the benefits and the risks is prudent. If the treatment provides a great benefit while having little risk, it should probably be taken. However, if the benefit does not outweigh the risk associated with the treatment, it should probably not be taken. If the risk and benefit are considered to be equal, the treatment most likely isn't worth the trip to the doctor.
I'm going to roughly evaluate the benefits and risks associated with the DPT vaccine where the benefits apply to tetanus. The DPT vaccine is one of the most well documented. It's difficult to find anything reliable on DTaP. The only thing I read about that had any semblance of a scientific study regarding Pediarix (DTaP/IPV/Hep B, the five-in-one currently used in the U.S.) contained a sampling of only 400, 8 of which had questionable serious adverse events. The sample size is way too small for the probabilities we're considering here, so it's hardly worth mentioning.
If someone wants to argue that the drugs have improved, put forth some serious studies that are readily available to interested parents. In this information age, there's absolutely no excuse for doing otherwise. The onus is not on me to prove how safe a treatment is. If they can't make a decent study readily available to me, they have no business sticking it in my kids.
Before looking at the probabilities of getting tetanus, I would like to establish one fact: the vaccination is not the only thing that prevents tetanus. Sanitation and wound hygeine are also factors. All three of these factors have become more significant over time - the number of vaccinations has increased, sanitation has improved, and wound hygeine has improved.
A Population-Based Serologic Survey of Immunity to Tetanus in the United States by the New England Journal of Medicine found that 69.7% of people at least six years old surveyed from 1988 to 1991 were immune to tetanus. For the sake of simple argument, let's assume that 69.7% of the U.S. population was immune in 1990.
Now, refer to Figure 1 near the bottom of the Tetanus Surveillance report by the Centers for Disease Control. It shows the decline of tetanus cases and tetanus deaths from 1947 to 1997. Tetanus cases began to be recorded in a consistent manner at 1947 when the vaccine was beginning to be used in a widespread manner.
According to the graph, 0.4 of every 100,000 people in the U.S got tetanus in 1947. In 1990, it was 0.025. 0.025 is 6.25% of 0.4. This suggests that the chance of getting tetanus in 1990 was 6.25% of what it was in 1947.
Now, let's assume that 100% of the population was not immune in 1947. This isn't true (though the percentage immune was probably very small), but it doesn't matter for this argument. With that assumption, the percentage of vulnerable people in 1990 was 30.3% of what it was in 1947.
Compared to 1947, the number of non-immune people in 1990 was 30.3% while the number of tetanus cases was 6.25%. In other words, the percentage of tetanus cases decreased faster than tetanus immunity. Yes, these are very rough estimates, but I think they still make one thing clear. Immunity could not have been the only factor in the reduced cases of tetanus.
Now, let's consider the benefits of the tetanus vaccination:
Even though immunity isn't the only factor, let's be conservative and assume that an unvaccinated person today will have the same probability of getting tetanus that people did in 1947 when the vaccination was rare. (I'd go back to a time before the vaccination even existed, but the numbers are less reliable. I want to compare apples to apples as much as possible.) The probability of getting tetanus for that year was .000,4%. Add the fact that the life expectancy of a United States female today is 81 years. This means that, for our argument, the probability of an unvaccinated person getting tetanus today is 1 out of 3,086.
What about death? Let's use the data in the graph again. Consider that, in 1997, about 0.000,016% of the population got tetanus while 0.000,002% died from it. This means that, with 1997 medical care, 1 out of 8 tetanus patients died. If the chance of getting tetanus is 1 in 3,086, then the chance of dying to it with 1997 care is 1 in 24,688. Rough, yes, but good enough for now.
It is not the case that being vaccinated means you will not get tetanus, but let's just assume that it does anyway. This means that the benefit of getting vaccinated is the prevention of a 1 in 3,086 chance of getting tetanus, or the prevention of a 1 in 24,688 chance of death due to tetanus.
On to the risks:
One source of data for the DPT vaccination is a UCLA study published in 1981. According to the National Vaccine Information Center:
"The UCLA/FDA study showed that 1 in 875 DPT shots results in a convulsion or collapse/shock reaction. Two babies in the study died following DPT reaction symptoms, but their deaths were classified as sudden infant death syndrome (SIDS or crib death)."
There were 15,752 DPT immunizations in that study, so if you associate the deaths with the vaccination that's a 1 in 7,876 probability. The cause of those deaths is questionable, so we'll leave them out.
1 in 875 vaccinations result in a serious adverse reaction, but a person's chances of a bad reaction increase if they have more than one shot. According to the typical DPT vaccination schedule, it was administered five times to young children. (It is recommeded that an adult receive a tetanus booster shot every ten years, but this is different from DPT or its variants.) This gives us a 1 in 175 chance of convulsion or collapse/shock in a lifetime.
In 1994, the NVIC reported on a year-long investigation of the Vaccine Adverse Reaction Reporting System (all data available for download at http://vaers.hhs.gov). It found that "A total of 54,072 reports of adverse events following vaccination were listed in a 39-month period from July 1990 to November 1993 with 12,504 reports being associated with DPT vaccine, including 471 deaths." (http://www.whale.to/vaccine/nvic4.html) Keep in mind that these are only the results of the national reporting system. The article goes on to explain, "At the end of February 1994, NVIC/DPT also conducted a survey of 159 doctors' offices in seven states, including Arkansas, California, Georgia, Illinois, Maryland, New York, and Texas. When asked the question, "In case of an adverse event after vaccination, does the doctor report it and, if yes, to whom?" only 28 out of 159, or 18 percent said they make a report to the FDA, CDC or state health department. In New York, only one out of 40 doctors' offices confirmed that they report a death or injury following vaccination."
Ick. Let's be generous and assume that the VAERS database contained 20% of the total number of incidents. That means there were really 62,520 adverse events from DPT, 2,355 of which were deaths. So, how many DPT vaccinations were given during this time period?
I can't seem to find any information that even comes close to estimating the number of actual DPT vaccinations in the U.S. during this period, but I think we can still come up with a very generous estimate. According to the typical DPT vaccination schedule, it was administered five times to young children. I'm going to assume that the CDC's DTaP schedule resembles the old DPT schedule. I'm going to assume that every child born during the time periods that qualify them as being due for a shot during the study's time period lived long enough to get all five shots, and that every single one of them did in fact get all five shots. We can estimate this number of children using the annual birth numbers listed at http://www.johnstonsarchive.net/policy/abortion/ab-unitedstates.html. This estimate is 65,499,055.
Something close to 65,499,055 children were born during five 39-month periods that make them due for DPT shots during the time period analyzed for the NVIC study. I'll assume that 65,499,055 shots actually were given even though it realistically would have been fewer. With 62,520 adverse events and 2,355 deaths, this suggests a 1 in 1,048 chance of adverse event per shot and a 1 in 27,813 chance of death per shot. Multiply those by five shots per person, and you get a 1 in 210 chance of adverse event in a lifetime and a 1 in 5,563 chance of death in a lifetime.
Here's a quick review of my estimates:
If you are unvaccinated:
1:3,086 to get tetanus in lifetime
1:24,688 to die from tetanus in lifetime
1:2 chance your dad's fingers will bleed from too much typing
If you are vaccinated:
1:875 to have convulsion or collapse/shock per shot
1:1,048 to have adverse event per shot
1:27,813 to die per shot
1:175 to have convulsion or collapse/shock in lifetime
1:210 to have adverse event in lifetime
1:5,563 to die in lifetime
So far, it's looking like DPT isn't worth the effort to pull up my child's sleeve. Even if we assume the VAERS database contains every single incident, it still looks bad for DPT. Also, consider that this does not even take into account the tetanus booster everyone is supposed to get every ten years. I don't even need to bother with it.
Based on all of these extremely conservative estimates, if my concern is Tetanus, it would be irresponsible of me to have my children vaccinated.
I have zero medical training. I'm not responsible for your medical decisions. If you make any judgements based on this post, your lockjaw isn't my fault.
This is hardly a scientific study, but rather my own analysis of other studies. I'm going to show a blatant disregard for things such as significant digits and formal terminology. Where assumptions need to be made, I will tend to grant the benefit of the doubt to support of vaccination. I'm open to corrections, so if you have any please leave a comment or write an e-mail. Finally, I'm going to avoid wasting time defining everything. There is plenty of information available on what tetanus is and what the DPT vaccination is.
Generally, when considering whether to receive medical treatment, an analysis of the benefits and the risks is prudent. If the treatment provides a great benefit while having little risk, it should probably be taken. However, if the benefit does not outweigh the risk associated with the treatment, it should probably not be taken. If the risk and benefit are considered to be equal, the treatment most likely isn't worth the trip to the doctor.
I'm going to roughly evaluate the benefits and risks associated with the DPT vaccine where the benefits apply to tetanus. The DPT vaccine is one of the most well documented. It's difficult to find anything reliable on DTaP. The only thing I read about that had any semblance of a scientific study regarding Pediarix (DTaP/IPV/Hep B, the five-in-one currently used in the U.S.) contained a sampling of only 400, 8 of which had questionable serious adverse events. The sample size is way too small for the probabilities we're considering here, so it's hardly worth mentioning.
If someone wants to argue that the drugs have improved, put forth some serious studies that are readily available to interested parents. In this information age, there's absolutely no excuse for doing otherwise. The onus is not on me to prove how safe a treatment is. If they can't make a decent study readily available to me, they have no business sticking it in my kids.
Before looking at the probabilities of getting tetanus, I would like to establish one fact: the vaccination is not the only thing that prevents tetanus. Sanitation and wound hygeine are also factors. All three of these factors have become more significant over time - the number of vaccinations has increased, sanitation has improved, and wound hygeine has improved.
A Population-Based Serologic Survey of Immunity to Tetanus in the United States by the New England Journal of Medicine found that 69.7% of people at least six years old surveyed from 1988 to 1991 were immune to tetanus. For the sake of simple argument, let's assume that 69.7% of the U.S. population was immune in 1990.
Now, refer to Figure 1 near the bottom of the Tetanus Surveillance report by the Centers for Disease Control. It shows the decline of tetanus cases and tetanus deaths from 1947 to 1997. Tetanus cases began to be recorded in a consistent manner at 1947 when the vaccine was beginning to be used in a widespread manner.
According to the graph, 0.4 of every 100,000 people in the U.S got tetanus in 1947. In 1990, it was 0.025. 0.025 is 6.25% of 0.4. This suggests that the chance of getting tetanus in 1990 was 6.25% of what it was in 1947.
Now, let's assume that 100% of the population was not immune in 1947. This isn't true (though the percentage immune was probably very small), but it doesn't matter for this argument. With that assumption, the percentage of vulnerable people in 1990 was 30.3% of what it was in 1947.
Compared to 1947, the number of non-immune people in 1990 was 30.3% while the number of tetanus cases was 6.25%. In other words, the percentage of tetanus cases decreased faster than tetanus immunity. Yes, these are very rough estimates, but I think they still make one thing clear. Immunity could not have been the only factor in the reduced cases of tetanus.
Now, let's consider the benefits of the tetanus vaccination:
Even though immunity isn't the only factor, let's be conservative and assume that an unvaccinated person today will have the same probability of getting tetanus that people did in 1947 when the vaccination was rare. (I'd go back to a time before the vaccination even existed, but the numbers are less reliable. I want to compare apples to apples as much as possible.) The probability of getting tetanus for that year was .000,4%. Add the fact that the life expectancy of a United States female today is 81 years. This means that, for our argument, the probability of an unvaccinated person getting tetanus today is 1 out of 3,086.
What about death? Let's use the data in the graph again. Consider that, in 1997, about 0.000,016% of the population got tetanus while 0.000,002% died from it. This means that, with 1997 medical care, 1 out of 8 tetanus patients died. If the chance of getting tetanus is 1 in 3,086, then the chance of dying to it with 1997 care is 1 in 24,688. Rough, yes, but good enough for now.
It is not the case that being vaccinated means you will not get tetanus, but let's just assume that it does anyway. This means that the benefit of getting vaccinated is the prevention of a 1 in 3,086 chance of getting tetanus, or the prevention of a 1 in 24,688 chance of death due to tetanus.
On to the risks:
One source of data for the DPT vaccination is a UCLA study published in 1981. According to the National Vaccine Information Center:
"The UCLA/FDA study showed that 1 in 875 DPT shots results in a convulsion or collapse/shock reaction. Two babies in the study died following DPT reaction symptoms, but their deaths were classified as sudden infant death syndrome (SIDS or crib death)."
There were 15,752 DPT immunizations in that study, so if you associate the deaths with the vaccination that's a 1 in 7,876 probability. The cause of those deaths is questionable, so we'll leave them out.
1 in 875 vaccinations result in a serious adverse reaction, but a person's chances of a bad reaction increase if they have more than one shot. According to the typical DPT vaccination schedule, it was administered five times to young children. (It is recommeded that an adult receive a tetanus booster shot every ten years, but this is different from DPT or its variants.) This gives us a 1 in 175 chance of convulsion or collapse/shock in a lifetime.
In 1994, the NVIC reported on a year-long investigation of the Vaccine Adverse Reaction Reporting System (all data available for download at http://vaers.hhs.gov). It found that "A total of 54,072 reports of adverse events following vaccination were listed in a 39-month period from July 1990 to November 1993 with 12,504 reports being associated with DPT vaccine, including 471 deaths." (http://www.whale.to/vaccine/nvic4.html) Keep in mind that these are only the results of the national reporting system. The article goes on to explain, "At the end of February 1994, NVIC/DPT also conducted a survey of 159 doctors' offices in seven states, including Arkansas, California, Georgia, Illinois, Maryland, New York, and Texas. When asked the question, "In case of an adverse event after vaccination, does the doctor report it and, if yes, to whom?" only 28 out of 159, or 18 percent said they make a report to the FDA, CDC or state health department. In New York, only one out of 40 doctors' offices confirmed that they report a death or injury following vaccination."
Ick. Let's be generous and assume that the VAERS database contained 20% of the total number of incidents. That means there were really 62,520 adverse events from DPT, 2,355 of which were deaths. So, how many DPT vaccinations were given during this time period?
I can't seem to find any information that even comes close to estimating the number of actual DPT vaccinations in the U.S. during this period, but I think we can still come up with a very generous estimate. According to the typical DPT vaccination schedule, it was administered five times to young children. I'm going to assume that the CDC's DTaP schedule resembles the old DPT schedule. I'm going to assume that every child born during the time periods that qualify them as being due for a shot during the study's time period lived long enough to get all five shots, and that every single one of them did in fact get all five shots. We can estimate this number of children using the annual birth numbers listed at http://www.johnstonsarchive.net/policy/abortion/ab-unitedstates.html. This estimate is 65,499,055.
Something close to 65,499,055 children were born during five 39-month periods that make them due for DPT shots during the time period analyzed for the NVIC study. I'll assume that 65,499,055 shots actually were given even though it realistically would have been fewer. With 62,520 adverse events and 2,355 deaths, this suggests a 1 in 1,048 chance of adverse event per shot and a 1 in 27,813 chance of death per shot. Multiply those by five shots per person, and you get a 1 in 210 chance of adverse event in a lifetime and a 1 in 5,563 chance of death in a lifetime.
Here's a quick review of my estimates:
If you are unvaccinated:
1:3,086 to get tetanus in lifetime
1:24,688 to die from tetanus in lifetime
1:2 chance your dad's fingers will bleed from too much typing
If you are vaccinated:
1:875 to have convulsion or collapse/shock per shot
1:1,048 to have adverse event per shot
1:27,813 to die per shot
1:175 to have convulsion or collapse/shock in lifetime
1:210 to have adverse event in lifetime
1:5,563 to die in lifetime
So far, it's looking like DPT isn't worth the effort to pull up my child's sleeve. Even if we assume the VAERS database contains every single incident, it still looks bad for DPT. Also, consider that this does not even take into account the tetanus booster everyone is supposed to get every ten years. I don't even need to bother with it.
Based on all of these extremely conservative estimates, if my concern is Tetanus, it would be irresponsible of me to have my children vaccinated.
Wednesday, February 18, 2009
A Stick of Mint
One of the neat things about Linux is that you can install an entire operating system on a little USB stick. This allows you to run your personal operating system from just about any computer at any location. As you can imagine, there are many reasons for doing this. I'll explain mine below.
I have access to a laptop with an encrypted hard drive with Windows installed on it. I won't get in to the reasons why, but I should not be installing any programs on it. Use your imagination. If I can't install programs on Windows, I could normally resize the Windows partition and install Linux on the cleared space. However, I can't just resize the partition because the hard drive is encrypted. The one good thing is that the bios is not password protected, so I am free to change the boot order. This allows me to boot the laptop from things other than hard drives, such as CDs or USB drives.
I put an 8GB thumb drive on my Christmas list, and my brother-in-law was kind enough to get me one. Let the experimentation begin!
You can do this with as little as 1GB, but that only leaves you with about 200 MB of extra data storage. 4GB should give you plenty of wiggle room for installing updates, configuration changes, and adding several extra programs. With 8GB, I have plenty of extra room for things like music and videos.
I tried this twice, and I'll explain why later. The first method is simpler but limits your storage space to about 4GB. The second method unlocks all the available space on your USB drive.
Ubuntu has a program called usb-creator, and I used it for both methods. It is installed by default on Ubuntu 8.10, but it also works on Ubuntu 8.04 if you download it. To use it, look for the option to create a USB startup disk under the Administration menu. You might be doing this from a Live CD, which is just fine. I heard that you can even run usb-creator from Windows, so you are free to do this from your Linux or Windows desktop if you wish.
In order to boot from a live CD or a USB drive, you may need to change the boot order in your bios. This method can vary depending on your motherboard, so I won't get into that here.
The First Method:
For my first go, I formatted my entire USB drive with a FAT32 partition before using usb-creator. To do this, you can use the program, GParted, which can be found in the administration menu as "Format Hard Disk." You will need at least one FAT32 partition because the USB drive will be running Linux similar to the way your Live CD does. Don't worry, the program will enable a way to save your files and changes as I mentioned earlier.
When you run usb-creator, the program will ask you to select a .iso file to use for the installation. This can be the same .iso file you used to create your live CD. I used a second thumb drive to hold this .iso file while the program running off the live CD installed to my new thumb drive. I suppose you could copy the .iso file to the home folder of the live CD, which means it would be stored in RAM and erased when you rebooted. That's fine because you only need it once, and it allows you to get away with using only one thumb drive for the entire process.
I used a .iso file for the Ubuntu 8.10 64-bit Desktop version the first time. This worked right "out of the box" with only one small issue that I will discuss later.
One of the other options the program gives you is the ability to enable persistency. This allows your changes to the operating system to be saved, making it much nicer than just running from the Live CD. This option presents you with a slider, allowing you to designate an amount of "persistent space" on your thumb drive. This space can be as small as 128MB or as large as 4GB. Here's what's really happening behind the scenes:
The usb-creator program will copy the Live CD files from the .iso image to your thumb drive just like it does for CDs, and mark the FAT32 partition as bootable. This takes up about 700MB of the parition's space. Then, the program will create what is called a persistency file. This file is named "casper-rw" and is stored on the root of your FAT32 partition right along with the Live CD files. This persistency file is used by your Linux installation to store all the changes you make to the operating system. All your configuration changes, software updates, and added data files such as music or office documents are magically stored inside this persistency file. Think of it as a big, complex .zip file. The FAT32 file system cannot handle files that are larger than 4GB, no matter what operating system you are using. You cannot create a persistency file larger than 4GB, and as a result, you cannot store more than 4GB of extra data in it. There is a way around this, and I used it for my second method which I will discuss later.
Anyway, once you tell the program how large you want the persistency file to be, it's ready for you to click, "Make Startup Disk." Writing to flash isn't exactly speedy, so this may take a few minutes. Once it's finished, shut down operating system, remove the CD when it tells you to, and boot up again with the USB stick in the computer of your choice.
When I did this the first time, I had an almost fully functional Ubuntu operating system. It takes a while to boot up because flash drives are slower than hard drives. However, once the operating system is up and running, most things seem to run quite smoothly. The one thing that did not work right away was the wireless card. All I had to do was connect the laptop to my wired network, go to Hardware Management in the Administration menu, and select the wireless driver I wanted to enable. It downloaded the driver and turned it on after I rebooted. This is a Dell Latitude D630.
I noticed that whenever I would boot off the thumb drive, it would come up with a menu similar to the one you see on the Live CD. If you just select the default or let it timeout to the default for you, it will continue to boot to your persistent Ubuntu installation. I also noticed that Ubuntu did not require any login credentials by default, just like the Live CD. You can enable this if you want, and I did the first time, but I won't get into that right now.
One of the most fun things about having Linux on the laptop was that I was able to install Adobe Flash and watch TV shows on Hulu.com via my 32" LCD TV. I didn't have to worry about hauling a larger, noisier desktop machine over by the TV and running network cables to the family room. There were two drawbacks to this, however. First, the only sound output on the laptop is meant for microphones. In order to hear the shows we had to turn the volume all the way up on the TV and computer. I am sure a desktop or other laptop with standard sound output would not have this problem. Second, the laptop doesn't like displaying at full resolution on my TV, so the picture is smaller than it could be if the show is in widescreen. I also think that a desktop would fix this issue, because the operating system won't have to worry about the laptop's built-in display. I may still figure out how to get this to work.
Anyway, I did have one small problem when I tried installing updates from Synaptic Package Manager. I selected "Mark All Upgrades" and let it do its thing, as usual. However, when it was done, it gave me an error about a broken symbolic link. This error would pop up every time I did an update in the future. After a bit of research, I found out that it was because of an issue downloading Linux kernel updates onto Live CD types of installations. Remember, the usb-creator method of making a bootable thumb drive treats your installation much like a Live CD. The update that caused the bug wasn't really necessary, and the bug itself seemed to be harmless. I also discovered a way to fix it, but I won't go into that here.
Remember that my persistency file was only about 4GB in size. The core Linux installation was about 700MB. This left me with around 3.5GB of wasted space. I used GParted to turn that extra space into a separate FAT32 partition so I could store other files on it and access them while Linux was running. However, in the end I decided I really wanted to have all of my storage space on a single partition so I could install as many programs as I wanted in the future. This brings me to...
The Second Method:
I also had recently done some research on another popular Linux distribution called Linux Mint. According to DistroWatch.com, it is the third most popular Linux distro after Ubuntu and openSUSE. Linux Mint is based off of Ubuntu similar to the way Ubuntu is based off of Debian. However, they seem to focus on a much more user friendly experience (though I would definitely not complain about Ubuntu being unfriendly) and a better "out of the box" experience. The primary version of the operating system comes with media codecs and other free proprietary software (such as Adobe Flash) preinstalled. It also comes with some applications called Mint Tools, which have been added to make life a little easier. The graphical interface is also attractive, and more familiar to people who are used to Windows.
Since changing how persistent data was stored on my thumb drive would require me to basically start over from the beginning anyway, I decided to use Linux Mint 6 64-bit for my .iso this time. But why did I need to start over?
I found out that instead of using the casper-rw persistence file, it is also possible to simply create a separate partition with the label "casper-rw" and it will store all of the persistent data. You just need to delete the casper-rw file so Linux knows to use the partition instead.
The recommended file system for this partition is ext2. Why not ext3? Well, ext3 does a lot of extra writing to the drive for things like indexing, which isn't necessary and will actually wear out your flash drive faster. (Flash drives are supposed to have a more limited lifespan than hard drives, so keep this in mind if you ever use them for important data.) Also, it is my understanding that some of the features of ext3 aren't even enabled when running like a Live CD, so much of the benefit of ext3 is wasted anyway. So, I used GParted to make a 900MB FAT32 partition for the Live CD portion of the operating system combined with a temporary minimal casper-rw file, and used the rest of the space for ext2.
After the partitions were set up, I went to usb-creator and had it make my Linux Mint bootable drive with a minimum persistence file of 128MB. After that was finished, I deleted the casper-rw file from the FAT32 partition. In order to make use of that extra space, I went back to GParted and resized the FAT partition to 700MB and added the extra to my ext2 partition. Then, I had to label the ext2 partition as "casper-rw" by using the following command in the terminal:
e2label /dev/ casper-rw
The exact location of the ext2 partition should be listed in GParted, so be sure to use that in the command.
I rebooted, took out the CD, booted from the thumb drive, and had Linux Mint running. There was just one problem: no persistence. Of course, it was the first thing I tested. I did some research and learned a little about the Live CD menu that pops up before you actually boot into Linux. On the Live CD, and on the thumb drive, there is a file that tells the menu what options to display and how to behave. This file is /syslinux/syslinux.cfg. This file needs to tell Linux to boot into persistent mode, not just normal Live mode. I am assuming that for whatever reason, usb-creator was able to properly modify this file for Ubuntu but not for Mint, so I had to do it myself. Here's how:
Above the line that says "label live" you need to insert the lines:
label custom
menu label Start Linux Mint in ^persistent mode
kernel /casper/vmlinuz
append file=/cdrom/preseed/mint.seed boot=casper persistent initrd=/casper/initrd.gz quiet splash --
You'll notice that the only difference between the custom section and the live section is the word "persistent." After you insert those lines, you need to go to the top line and change it to say:
default custom
I guess this tells the menu to use the information in the "custom" label as the default option. For the heck of it, I also changed the "timeout 100" line to say:
timeout 50
After doing this, two things happened. First, I noticed that I no longer got the boot menu at all! The thumb drive just went straight into Linux Mint. I'm not sure why this happened, but I might have messed up the syntax in the config file. After all, this was all new to me. I actually thought this was a good thing, however, because I really don't need that menu anyway. The other thing that happened: persistent mode was now working. I guess I killed two birds with one stone.
After I installed the wireless driver exactly as before, everything worked perfectly. I didn't have to install Adobe Flash because it was already installed, and Hulu worked the same as before. I also heard that DVDs play just fine right out of the box.
Instead of using Synaptic Package Manager for my updates, I used mintUpdate, which was located in the lower right corner of the tray. One of the neat things about mintUpdate is that it ranks each update on a scale of 1 to 5. If something is very important and/or very safe to install, it gets a 1. If something is unnecessary and/or unstable, it gets a 5. I've heard that these ranks can change based on user feedback, so even if a package gets through initial testing but causes problems later, it can be moved down the list. By default, mintUpdate only displays and installs updates with a rank of 1 to 3. I changed it to display all five ranks, but only install 1 to 3. Guess what package showed up in the list of unsafe updates? The kernel update that gave me that error under Ubuntu. One of Linux Mint's tools just prevented me from getting an unnecessary error that Ubuntu would have given me (and did). I am a happy camper.
Linux Mint looks like the best option for Windows users that want to get away from Windows. I'm already talking it up to my friends. It's free, easy to use, and has a nice graphical interface that feels a bit more like what we're used to under Windows. (As with most Linux distros, you customize the GUI to look very different if you wish.) It comes with a complete office suite and e-mail client among other useful software, and it's completely free. Check it out at http://www.linuxmint.com/.
I have access to a laptop with an encrypted hard drive with Windows installed on it. I won't get in to the reasons why, but I should not be installing any programs on it. Use your imagination. If I can't install programs on Windows, I could normally resize the Windows partition and install Linux on the cleared space. However, I can't just resize the partition because the hard drive is encrypted. The one good thing is that the bios is not password protected, so I am free to change the boot order. This allows me to boot the laptop from things other than hard drives, such as CDs or USB drives.
I put an 8GB thumb drive on my Christmas list, and my brother-in-law was kind enough to get me one. Let the experimentation begin!
You can do this with as little as 1GB, but that only leaves you with about 200 MB of extra data storage. 4GB should give you plenty of wiggle room for installing updates, configuration changes, and adding several extra programs. With 8GB, I have plenty of extra room for things like music and videos.
I tried this twice, and I'll explain why later. The first method is simpler but limits your storage space to about 4GB. The second method unlocks all the available space on your USB drive.
Ubuntu has a program called usb-creator, and I used it for both methods. It is installed by default on Ubuntu 8.10, but it also works on Ubuntu 8.04 if you download it. To use it, look for the option to create a USB startup disk under the Administration menu. You might be doing this from a Live CD, which is just fine. I heard that you can even run usb-creator from Windows, so you are free to do this from your Linux or Windows desktop if you wish.
In order to boot from a live CD or a USB drive, you may need to change the boot order in your bios. This method can vary depending on your motherboard, so I won't get into that here.
The First Method:
For my first go, I formatted my entire USB drive with a FAT32 partition before using usb-creator. To do this, you can use the program, GParted, which can be found in the administration menu as "Format Hard Disk." You will need at least one FAT32 partition because the USB drive will be running Linux similar to the way your Live CD does. Don't worry, the program will enable a way to save your files and changes as I mentioned earlier.
When you run usb-creator, the program will ask you to select a .iso file to use for the installation. This can be the same .iso file you used to create your live CD. I used a second thumb drive to hold this .iso file while the program running off the live CD installed to my new thumb drive. I suppose you could copy the .iso file to the home folder of the live CD, which means it would be stored in RAM and erased when you rebooted. That's fine because you only need it once, and it allows you to get away with using only one thumb drive for the entire process.
I used a .iso file for the Ubuntu 8.10 64-bit Desktop version the first time. This worked right "out of the box" with only one small issue that I will discuss later.
One of the other options the program gives you is the ability to enable persistency. This allows your changes to the operating system to be saved, making it much nicer than just running from the Live CD. This option presents you with a slider, allowing you to designate an amount of "persistent space" on your thumb drive. This space can be as small as 128MB or as large as 4GB. Here's what's really happening behind the scenes:
The usb-creator program will copy the Live CD files from the .iso image to your thumb drive just like it does for CDs, and mark the FAT32 partition as bootable. This takes up about 700MB of the parition's space. Then, the program will create what is called a persistency file. This file is named "casper-rw" and is stored on the root of your FAT32 partition right along with the Live CD files. This persistency file is used by your Linux installation to store all the changes you make to the operating system. All your configuration changes, software updates, and added data files such as music or office documents are magically stored inside this persistency file. Think of it as a big, complex .zip file. The FAT32 file system cannot handle files that are larger than 4GB, no matter what operating system you are using. You cannot create a persistency file larger than 4GB, and as a result, you cannot store more than 4GB of extra data in it. There is a way around this, and I used it for my second method which I will discuss later.
Anyway, once you tell the program how large you want the persistency file to be, it's ready for you to click, "Make Startup Disk." Writing to flash isn't exactly speedy, so this may take a few minutes. Once it's finished, shut down operating system, remove the CD when it tells you to, and boot up again with the USB stick in the computer of your choice.
When I did this the first time, I had an almost fully functional Ubuntu operating system. It takes a while to boot up because flash drives are slower than hard drives. However, once the operating system is up and running, most things seem to run quite smoothly. The one thing that did not work right away was the wireless card. All I had to do was connect the laptop to my wired network, go to Hardware Management in the Administration menu, and select the wireless driver I wanted to enable. It downloaded the driver and turned it on after I rebooted. This is a Dell Latitude D630.
I noticed that whenever I would boot off the thumb drive, it would come up with a menu similar to the one you see on the Live CD. If you just select the default or let it timeout to the default for you, it will continue to boot to your persistent Ubuntu installation. I also noticed that Ubuntu did not require any login credentials by default, just like the Live CD. You can enable this if you want, and I did the first time, but I won't get into that right now.
One of the most fun things about having Linux on the laptop was that I was able to install Adobe Flash and watch TV shows on Hulu.com via my 32" LCD TV. I didn't have to worry about hauling a larger, noisier desktop machine over by the TV and running network cables to the family room. There were two drawbacks to this, however. First, the only sound output on the laptop is meant for microphones. In order to hear the shows we had to turn the volume all the way up on the TV and computer. I am sure a desktop or other laptop with standard sound output would not have this problem. Second, the laptop doesn't like displaying at full resolution on my TV, so the picture is smaller than it could be if the show is in widescreen. I also think that a desktop would fix this issue, because the operating system won't have to worry about the laptop's built-in display. I may still figure out how to get this to work.
Anyway, I did have one small problem when I tried installing updates from Synaptic Package Manager. I selected "Mark All Upgrades" and let it do its thing, as usual. However, when it was done, it gave me an error about a broken symbolic link. This error would pop up every time I did an update in the future. After a bit of research, I found out that it was because of an issue downloading Linux kernel updates onto Live CD types of installations. Remember, the usb-creator method of making a bootable thumb drive treats your installation much like a Live CD. The update that caused the bug wasn't really necessary, and the bug itself seemed to be harmless. I also discovered a way to fix it, but I won't go into that here.
Remember that my persistency file was only about 4GB in size. The core Linux installation was about 700MB. This left me with around 3.5GB of wasted space. I used GParted to turn that extra space into a separate FAT32 partition so I could store other files on it and access them while Linux was running. However, in the end I decided I really wanted to have all of my storage space on a single partition so I could install as many programs as I wanted in the future. This brings me to...
The Second Method:
I also had recently done some research on another popular Linux distribution called Linux Mint. According to DistroWatch.com, it is the third most popular Linux distro after Ubuntu and openSUSE. Linux Mint is based off of Ubuntu similar to the way Ubuntu is based off of Debian. However, they seem to focus on a much more user friendly experience (though I would definitely not complain about Ubuntu being unfriendly) and a better "out of the box" experience. The primary version of the operating system comes with media codecs and other free proprietary software (such as Adobe Flash) preinstalled. It also comes with some applications called Mint Tools, which have been added to make life a little easier. The graphical interface is also attractive, and more familiar to people who are used to Windows.
Since changing how persistent data was stored on my thumb drive would require me to basically start over from the beginning anyway, I decided to use Linux Mint 6 64-bit for my .iso this time. But why did I need to start over?
I found out that instead of using the casper-rw persistence file, it is also possible to simply create a separate partition with the label "casper-rw" and it will store all of the persistent data. You just need to delete the casper-rw file so Linux knows to use the partition instead.
The recommended file system for this partition is ext2. Why not ext3? Well, ext3 does a lot of extra writing to the drive for things like indexing, which isn't necessary and will actually wear out your flash drive faster. (Flash drives are supposed to have a more limited lifespan than hard drives, so keep this in mind if you ever use them for important data.) Also, it is my understanding that some of the features of ext3 aren't even enabled when running like a Live CD, so much of the benefit of ext3 is wasted anyway. So, I used GParted to make a 900MB FAT32 partition for the Live CD portion of the operating system combined with a temporary minimal casper-rw file, and used the rest of the space for ext2.
After the partitions were set up, I went to usb-creator and had it make my Linux Mint bootable drive with a minimum persistence file of 128MB. After that was finished, I deleted the casper-rw file from the FAT32 partition. In order to make use of that extra space, I went back to GParted and resized the FAT partition to 700MB and added the extra to my ext2 partition. Then, I had to label the ext2 partition as "casper-rw" by using the following command in the terminal:
e2label /dev/
The exact location of the ext2 partition should be listed in GParted, so be sure to use that in the command.
I rebooted, took out the CD, booted from the thumb drive, and had Linux Mint running. There was just one problem: no persistence. Of course, it was the first thing I tested. I did some research and learned a little about the Live CD menu that pops up before you actually boot into Linux. On the Live CD, and on the thumb drive, there is a file that tells the menu what options to display and how to behave. This file is /syslinux/syslinux.cfg. This file needs to tell Linux to boot into persistent mode, not just normal Live mode. I am assuming that for whatever reason, usb-creator was able to properly modify this file for Ubuntu but not for Mint, so I had to do it myself. Here's how:
Above the line that says "label live" you need to insert the lines:
label custom
menu label Start Linux Mint in ^persistent mode
kernel /casper/vmlinuz
append file=/cdrom/preseed/mint.seed boot=casper persistent initrd=/casper/initrd.gz quiet splash --
You'll notice that the only difference between the custom section and the live section is the word "persistent." After you insert those lines, you need to go to the top line and change it to say:
default custom
I guess this tells the menu to use the information in the "custom" label as the default option. For the heck of it, I also changed the "timeout 100" line to say:
timeout 50
After doing this, two things happened. First, I noticed that I no longer got the boot menu at all! The thumb drive just went straight into Linux Mint. I'm not sure why this happened, but I might have messed up the syntax in the config file. After all, this was all new to me. I actually thought this was a good thing, however, because I really don't need that menu anyway. The other thing that happened: persistent mode was now working. I guess I killed two birds with one stone.
After I installed the wireless driver exactly as before, everything worked perfectly. I didn't have to install Adobe Flash because it was already installed, and Hulu worked the same as before. I also heard that DVDs play just fine right out of the box.
Instead of using Synaptic Package Manager for my updates, I used mintUpdate, which was located in the lower right corner of the tray. One of the neat things about mintUpdate is that it ranks each update on a scale of 1 to 5. If something is very important and/or very safe to install, it gets a 1. If something is unnecessary and/or unstable, it gets a 5. I've heard that these ranks can change based on user feedback, so even if a package gets through initial testing but causes problems later, it can be moved down the list. By default, mintUpdate only displays and installs updates with a rank of 1 to 3. I changed it to display all five ranks, but only install 1 to 3. Guess what package showed up in the list of unsafe updates? The kernel update that gave me that error under Ubuntu. One of Linux Mint's tools just prevented me from getting an unnecessary error that Ubuntu would have given me (and did). I am a happy camper.
Linux Mint looks like the best option for Windows users that want to get away from Windows. I'm already talking it up to my friends. It's free, easy to use, and has a nice graphical interface that feels a bit more like what we're used to under Windows. (As with most Linux distros, you customize the GUI to look very different if you wish.) It comes with a complete office suite and e-mail client among other useful software, and it's completely free. Check it out at http://www.linuxmint.com/.
Subscribe to:
Posts (Atom)