For most non-gamers the choice will be the i3. For light gamers, HTPC, and notebooks the choice will be Llano. For more serious gamers the choice is obviously the i3 since the Llano CPU is too slow and the Llano onboard GPU isn't anywhere near good enough. These people will use higher end discrete graphics cards. I chose i3 for my gaming box for exactly that reason.

Just a nitpick here, but Sandy Bridge supports full h.264 hardware decoding up to 1080p, 3D TV support, and Bitstreaming of HD Audio formats. I can decode 1080p blu-ray video using the Sandy Bridge GPU and not spike over 10% CPU usage. That's in addition to the fact that there's a desktop-targeted 35W model available that is perfectly suited for noise- and power-sensitive HTPC accplications.

Unless HTPC also needs to include playing modern 3D games at 1080p (which it can, for some), the i3 is probably a better choice given the similar prices, the GPU that's way more than sufficient for hardware decoding, the (possible) lower power consumption, and the beefier CPU that can future proof your computer for new codecs that may not have hardware decode support right from the beginning.

In my experience (and I've been working as a cashier the last ~7 months at a Walmart), parents usually don't give a damn.

Sure, that's probably the biggest group. It's also why many stores require cashiers to ask for ID. If a parent doesn't care, well, there's not much you can do beyond follow your employer's policy and your own code of ethics.

The original post complained about how well a single letter rating could summarize the content. I pointed out that it isn't meant to, and that there's a fine-grained system in place to clarify where the rating came from. Someone else complains that Wal-Mart's anti-theft system prevents you from using the fine-grained content descriptors on a large scale, but those aren't really relevant until you've already selected a game and are trying to decide if it's appropriate. It's not really fair to fault what is a pretty robust rating system for apathy on the part of a parent or a lousy anti-theft system.

Assuming you can even find an electronics associate who isn't busy with another customer. Or knows anything about their department.

And that would be but one of the reasons why I won't even bother with Wal-Mart. Incidentally, they're the only major retailer (besides Toys R Us) that I'm aware of that still uses the glass case method. Most either use individual anti-theft cases, dummy boxes, or something like Target's new system (the game boxes are all tethered to a storage cabinet) so that customers can pick up, handle, and read the game box.

You may want to check your configuration. DTS-HD has a "core" 5.1 DTS track, and Dolby TrueHD has a similar "interleaved" Dolby Digital 5.1 track. Any blu-ray player is supposed to automatically select the appropriate sound stream for your equipment. On the PS3, the system configuration will allow you to tell it how your gear is hooked up. If you tell it audio is connected via Optical digital and that it only supports AC-3, DTS, and PCM then it should automatically output those sources rather than a down-mixed 2.0 stereo feed.

And the one letter summary is a great place to start (and stop in most cases). You're not going to find Sexual Content in anything below an M-rated game. You won't find blood in anything below T-rated, and even then they stipulate "Minimal Blood" in T-rated games. The content descriptors are there to help make fine selections after you've narrowed the field - If you're comfortable with your child hearing profanity but not seeing violence then they can help you make your choice.

Once you're looking at a particular game (or a choice among a few), it's pretty trivial to ask the Wal-Mart employee to open the case and let you flip over the box.

I worked in videogame stores or big box retailers in one way or another for 6 years through high school and college. Admittedly, it's been years since I worked in a retail environment selling games, and maybe parents have changed since I sold games. I ran into plenty of parents who were concerned about content, and even a good many who asked about the ratings. In all my time, though, I never encountered someone for whom the content descriptor was their first step in the purchasing decision. It was a final step, a check on whether the game(s) they'd selected were appropriate.

In my experience, a parent was either concerned about the content of a particular game ("Joey said to get BloodSlaughter 5 for his birthday - is it appropriate?"), is asking for a recommendation from the staff ("What's good for a 13 year-old boy? Uh-huh, and is there a lot of adult stuff?") or was familiar enough with the games him/herself and not in need of running down each game. Maybe there's someone out there who starts their purchasing process by making a pile of games without Sexual Themes or Animated Violence descriptors, but it seems so backwards because it would result in a massive pile of things to sort through rather than choosing games that look appealing and deciding if they're appropriate. A Dora the Explorer licensed game is unlikely to have any objectionable content descriptors, but that doesn't mean that most parents will be actively considering it for their 12 year-old son.

In other words, no one's reading the back of every game in the Wal-Mart case because, even for the most game-clueless parents, half or more of the games in the case probably aren't potential purchases to begin with.

Games already do this. I can't speak to app store purchases, but retail boxed games that are ESRB rated contain a more detailed explanation on the back of the box.

Here is a list of the ratings and content descriptors. I'm not sure what a concise but detailed explanation would consist of, but the sexuality side of things is conveyed by a number of descriptors:

• Mature Humor
• Suggestive Themes
• Sexual Themes
• Sexual Content
• Partial Nudity
• Nudity
• Strong Sexual Content
• Sexual Violence

There are similar spectrums for violence, profanity, drugs, and gambling.

The only area in there where I can see any room for mistake on a parent's part might be Sexual Themes vs. Sexual Content. If someone knew that both existed, they could probably figure out the difference. You wouldn't see both on the same package since Sexual Content encompasses Sexual Themes. If someone just saw "Sexual Themes," though, I could imagine they might think it was worse than "References to sex or sexuality." Violence exists in a similar realm - if anything, the only ambiguity in the descriptors is that they sometimes sound worse than they really are.

$0. America's Army and Call of Duty are different games.

Maybe, maybe not. I've heard of scripted lessons implemented at the single subject or school level, but because of the price tag they're often rolled out at the district level. A principal could certainly voice his opposition to such a program while it's in the planning stage, but once the decision is made, it would be the district's way or the highway. You're absolutely right that such programs are foisted by administration, but you're absolutely wrong in assuming what level of administration makes the decision to use it.

Once the district decides to use scripted lesson plans, it's up to the OP's father whether he believes it's something worth taking a stand (and potentially losing his job)

What he's talking about are products I've seen referred to as "scripted lesson plans," and he's correct; they're not just textbooks and workbooks, and they're not the "seeds" of lessons.

I have never actually had to use these products in my own teaching experience, but I have seen them and we did work with some of them in my teaching classes in college. Imagine a general math concept such as fractions. There are companies who sell entire packets of lesson plans, designed to be implemented by every teacher in the district and to be used for X weeks for fractions. The packet is three hole punched so that it can be easily distributed in binder form, and really is a collection of "canned" lesson plans. The ones I encountered went so far as to break a day's worth of instruction down into a format like this:

Warm up: 10 Mins [use warmup transparency 11a]
Lesson: 12 Mins [use overhead transparency 11b]
Exercise: 25 Mins [use worksheet 11c]
Suggested homework: [worksheet 11d]
Sample modifications for students with disabilities: X, Y, Z
The real version is much more detailed, of course; the ones I saw for English classes typically consumed three pages for a 45 minute lesson.

Typically, a district would purchase an entire years' worth of lessons and put teachers through extensive in-service training to discuss the proper way to implement such programs.

It's appealing on one hand; as you probably know, planning lessons is difficult, time-consuming, and requires a lot of trial and error. I wasn't truly happy with most of my lessons until after the third or fourth time I'd taught and refined them. These products take out the guesswork. The lessons have been tested (the companies pushing them talk a lot about how much testing goes into their development), and their pacing honestly looked pretty good. On the other hand, of course, it's deeply insulting to the teachers involved; it reduces us to robots, removes the opportunities for creativity, and generally brings everyone down to the same level of mediocrity. I assume this is probably why his father's school had to go all the way to termination - if you let one person off the hook on canned lessons, then everyone will want to.

He's right though. Such products do exist.

Your ECC system is substantially more expensive; it's not just the RAM. $50 more for comparable RAM, $100-$150 more for a board capable of supporting it, and $100-$150 more for slower CPUs. It's the total cost of the platform, not just the RAM, that causes a sticking point. $250 or $300 goes a long way in performance. ECC is all of a sudden the difference between a single $200 card and a pair, and it doesn't provide any tangible benefit to the guy who wanted to game.

5-10% is important once it represents the difference between playable and not. You're absolutely right that the difference between 110 and 120fps on a 60hz LCD is pointless and comical, but mid-range systems don't tend to do that. If you can keep it at 60-70fps on today's games, then next year's start dropping a little below that point. I'll bet the person asking for a gaming computer will want every performance gain he can get to keep him in the playable range for the next few years.

Random crashes? I hate to fight your strawman with anecdotal evidence, but it's not really a problem. The flipped bit has to be one in use at the time it flips, and it has to be one that's actually important (and not just changing color on a texture or something). I haven't used ECC since the EDO RAM era, and I've never had a problem with system instability other than in the dark days of Windows 9x.

My best may be a webpage that's a little vague (who benchmarks products without trying to keep things as similar as possible?), but you haven't even gone that far. Seems as if you're sticking to handwaving for this one. Your argument has boiled down to two points thus far:

A) UPTIME! Besides months of uptime being useless to someone gaming (due to frequent reboots for Windows patches), I hope you understand that there isn't some mass "crash fest" afoot for the majority of the users not running ECC. Seriously, your system's uptime and stability, like any other well-built computer, likely owes far more to a good PSU, reputable brand components, and good cooling than it does ECC.

B) It's good enough; you might pay a bit more and get a little less performance, but that's not such a big deal. Except that it is a big deal, because every dollar you advocate spending on hardware with minimal benefit is a dollar not spent on hardware with an actual tangible benefit.

You and I both know that the best you could honestly tell someone in this position is "It might one day keep your OS or game from crashing, but I can't really tell you the likelihood that it will."

What's the best I've got? For starters, not dismissing non-substantial differences as meaningless? The end result of our entire discussion is the same as it was at the beginning: The original poster wants a gaming machine, and you would have him select something that is either slower at the same price or more expensive. Assuming whatever budget you're working with, burning money on pricier mainboards and RAM (even if you don't care about the CPUs) still takes money away from the portion of the budget that you can devote to GPUs. I'll completely agree that GPUs make up the majority of the performance in the computer, so I'm very confused about why you seem focused on getting the poster to drop cash in areas other than his GPU.

If you want to casually dismiss any performance difference as meaningless, go right ahead; don't be surprised when someone calls you on it. 1-2% is pointless and should be ignored, but 5-10% is not. The guy plays games, and wants to know why he would want ECC RAM in his gaming computer. Frames per second is a pretty damned important feature in a gaming computer, and is certainly a hell of a lot more compelling to someone who says they play games than months of uptime. ECC has its place, and clearly you value it, but its place is not in a gaming computer unless your goal is simply to throw money about.

So, what's the best you've got? The poster wants to play games, and presumably he's interested in the best product for doing so. Please, tell me how dropping loads more money on a mainboard that supports ECC RAM, the ECC RAM itself, and the slower processors to run the whole system will benefit him rather than focusing that budget on videocards. Alternately, feel free to resort to condescension and insults, as it's about all you appear capable of mustering to bolster your argument.

Easily done on a home PC, and with no gain or benefit for that user. OP wants a gaming machine, and you want to spout off on things that won't benefit him.

"CPU can have an impact" is a bit of an understatement. According to Tom's Hardware's Far Cry 2 benchmark a $280 core i7 920 will spank a pair of comparably priced 2.5Ghz Opterons ($360 for the pair) to the tune of almost 30 additional frames per second. The opteron isn't on that chart, mind you, but you can ballpark it pretty reasonably among the comparably clocked Phenom chips.

I'm glad that you've realized the folly of claiming that a slow CPU isn't going to make a difference, but I'd encourage you to go check benchmarks on RAM speed and x8 vs x16 GPUs on games (the latter might be harder to find as it's been a while since mid-range boards likely to be used for gaming have dropped speed on second/both PCI-E slots). It's not but a few frames per second - maybe 10 or 12 in aggregate between the RAM and GPU bandwidth- but that's quite a bit when it costs nothing more and the original poster wants a gaming machine.

FYI: calling people "kid" on the internet doesn't bolster your argument one bit. It just makes you look rude.

Hell, let's grant your claim that GPU speed is all that matters. It's not, but we'll roll with it. You're going to end up with a lousier GPU when you blow all of your budget on more expensive mainboards, RAM, and CPUs. Enjoy your sub-par (but not really any more stable) gaming experience!

The OP is building a gaming computer. He specifically asked about it. Not a "dev box," not a DB server, not a home theater PC, but something for gaming. Before you go spouting off like a jack-ass about not building gaming computers, understand that I'm talking about them because the original poster asked why he wouldn't put ECC in a gaming machine. For someone who doesn't build gaming computers and never claimed to, though, you sure can spout bullshit about games not benefiting from CPU speed, RAM speed, or GPU bandwidth.

Different needs for different apps; you can continue to pretend that RAM and GPU bandwidth and CPU speed have no effect on gaming. The rest of us will stay in reality, thanks.

You can get a dual CPU board for about $300 (Why dual CPU for gaming, again, when a multi-core chip will do fine for the few games that truly benefit from multiple CPUs?) You can spend more money on equal speed parts or kid yourself that your CPU speed won't have any effect - I'll give you that it affects some types of games less than others, but it always has some relevance to the argument.

You can pay more money for slower RAM that will be guaranteed against a flipped bit, even if that bit doesn't effect what you're playing at the moment anyway.

You can get 2 x16 slots that cut down to x8 speed when used in paired mode and further cripple your performance as compared to a regular ol' $150 semi-premium mainboard. Either computer may last for many years, but the presence of ECC RAM has no bearing on that case.

Staying up for months? I had a craptastic K6-2 on ALi chipset computer that had no problem doing that. If someone's building a gaming computer, they're probably using/dual-booting into windows, and the updates or dual-booting is going to negate their "amazing" uptime anyway.

In other words, you advocate spending a good deal more money for less performance than a non-server product would deliver on gaming, with "reliability" features that may never be relevant given the stated use. OP wants to know why he wouldn't want ECC RAM for gaming, and I explained it.

ECC is a server-targeted feature. Newegg has 18 mainboards that support ECC listed in the Dual LGA 1366 category alone, and I'd imagine plenty more scattered throughout their server board categories.

As you've already discovered, though, it's not terribly common on home-targeted boards. You're welcome to use one of those boards for gaming, but you'll probably have to use a pricier Xeon or Opteron processor, more expensive ECC RAM, and suffer with slower PCI-E links for your videocards. Higher prices and similar or slower gaming performance is probably not what you're interested in.

You'll also have to assume that a bit will flip in an area of RAM that's actually holding information that's important at the moment that bit flips; it's a useless feature if nothing's in the bit of RAM that accidentally flips. It's extremely useful on servers that are on 24/7, always stressed, and likely to have the RAM completely filled with important information. For home users, it falls on the wrong side of the cost/benefit test.

The Z3 hasn't been made in almost 7 years, and the first models rolled off the line in 1996. You can certainly debate the safety of giving the car to a new driver, but the odds are that the cost wasn't prohibitively more than many 6 or 7 year old cars. An extremely cursory internet search shows several 70K mile examples in my area for $10,000 to $11,000, and those were all 2000 or later models. Older stuff is likely closer to the $7,000 mark.

If you had enough cash to buy your 16 year-old a Z3, you also have enough cash to buy them a 6 or 7 year-old Toyota Camry. Getting a good condition used Camry as your first car is a pretty sweet deal, mind you, but it's also not the sort of carg you'd bat an eye at if you saw a young driver in one.