The power grid is only conceptually (and billing wise) treated as a battery. It isn't electrically. The grid doesn't have a set of batteries (and AC-to-DC converters) storing excess solar panel output for later use. Instead, the excess power is consumed by nearby homes that don't have solar panels (the path of least resistance). Billing wise, it is treated as a battery (in a majority of areas), because that makes the billing simple.

I see roof-top solar as a convienence for the generators. It effectively removes load during daylight (ie peak) times, and transfers the load to nighttime (ie, non-peak) times. It smooths out the day/night variances in generation. That only works as long as the roof-top solar production remains smaller than day/night variance, but that's the case we're dealing with now.

Aside from that, I (living in SoCal) do pay a delivery charge on my monthly bill, regardless of my generation or consumption of power. I've no idea how it's computed, and it varies month to month. It doesn't seem to be related to how much I generate or consume in a given month. It's a couple dollars, so I don't really care.

I don't do this often, but I do occasionally use indentation that follows the data's logic, not the program's logic.  For example, I've written some XML generation code like:

$xml = new Xml();
$foo = $xml->addChild( 'foo');
  $foo->addAttribute( 'attr', 'bar');
  $foo->addNode( 'foonode', 'foonodevalue');
  $fc1->addChild( 'foochild1');
    $fc1->addNode( 'foochild1node', 'value1');
  $fc2->addChild( 'foochild2');
    $fc2->addAttribute( 'foochild2attr', 'value2');
    $fc2->addNode( 'foochild2node', 'value2');
$xml->addNode( 'xmlnode', 'value3');

Code like this gets large comment blocks above and below stating that I'm using a non-standard indentation, and why.

When I later had to refactor the XML schema, it was incredibly useful to have it indented this way.

It depends on the library and language though.  When I was using nested C structures, or nested object, it wasn't necessary, because the data structure was a lot more visible.  This code is easier to understand using traditional indentation:

$xml = new Xml();
$xml->addChild( 'foo');
$xml->foo->addAttribute( 'attr', 'bar');
$xml->foo->addNode( 'foonode', 'foonodevalue');

$xml->foo->addChild( 'foochild1');
$xml->foo->foochild1->addNode( 'foochild1node', 'value1');

$xml->foo->addChild( 'foochild2');
$xml->foo->foochild2->addAttribute( 'foochild2attr', 'value2');
$xml->foo->foochild2->addNode( 'foochild2node', 'value2');

$xml->addNode( 'xmlnode', 'value3');

See my other post in this thread.

Agreed, I'm not worrying about it, but I will in a couple years.

If you owe more than $1,000, you owe a penalty. Unless you got a refund last year, or you owe less than you did the previous year. The wording is awkward there. Basically the first underpayment is free, and you get a pass while you're trying to fix the problem. I'm pretty sure TurboTax told me this.

It is legal, and the IRS has a calculator for you. Intuit has one online somewhere. TurboTax will generate a W4 for you after you do your taxes. However, I hate all of them. They all try to take into account what's been withheld year-to-date, and figure out the magic number to get a $0 refund at the end of the year. And every single one gives me a wildly different number, without showing their work. Which means they're all wrong, and I'll have to try again next year. I finally sat down and figured it out, because fixing it was a couple hundred bucks a paycheck. I hate myself, but I made the largest spreadsheet of my life using the IRS Employer's Withholding publication. I started coding it, but the publication tables mentally mapped better to a spreadsheet. My HR department wasn't willing to help me, and I can't blame them once I finished. Now I just tweak the W4 number up or down by 1, based on if my refund was larger or smaller than last year.

I need to upgrade the shared family desktop from 10GB to 16GB. Every user logs in, then switches users instead of logging out. A Chrome process for every user eats up a lot of RAM. For a few extra bucks, I save everybody in the household tens of minutes per day (my kids are constantly switching users... I should probably get more machines).

Only one player uses the tablet. The other 4 players use standard Wii remotes.

It actually makes for some interesting game mechanics. Nintendo Land (the game that came with my WiiU) is basically 12 demo games bundled together, and they all do something with the tablet. In Mario Chase, you play hide and seek. The guy with the tablet hides, the guys with the wiimotes seek. The seekers can't see the hider's screen, but they can see each others'. In the Zelda game, the guy with the tablet is an archer, and uses the tablet to look around. The guy(s) with the wiimotes are swordsmen. In the racing game, you steer with the tablet's accelerometers. None of the games are very complicated, but Nintendo Land has replaced Wii Sports as the party game.

I really like Rayman Legends on the WiiU. Some of the levels really make use of the tablet feature, especially in two player mode. The guy with the remote runs the dungeon, and the guy with the tablet uses the touch screen to manipulate the dungeon. It requires a lot of communication between both players, and I think it's really well done. This is the game I point to when people ask why a console needs a tablet. Penny Arcade talked about this game a couple of times, and I agree.

I did end up buying one WiiU Pro controller. All of the functionality of the tablet controller, without a screen. For some of the games my kids play (Transformers), the tablet would always beat the wiimote. The Pro controller restored the balance between the players. It's also more comfortable for long game play than the wiimotes.

TL;DR: I wasn't planning to upgrade to the Wii to a WiiU, but I'm glad I did.

Seconded. I've been doing the same, since 1999. Web spiders are responsible for most of my upstream bandwidth, and I only notice when I'm looking at the log files. None of the 4 ISPs I've had over the years have complained or blocked my service.

The only actual problem I've had is email deliverability. Most destinations would bounce my emails because they came from a Dynamic IP. I configured Postfix to forward everything through Time Warner's mail servers, and I haven't had problems since.

The CA fine print has the same info.

If you don't agree, you can can ask the provider for a RF shielding bag. This comes with a warning that you're liable for fines if you forget to take the pass out of the bag before using a toll.

I know I'm late to this thread, but I didn't see anybody else say this.

After my son had corrective surgery for a crossed eye, the surgeon warned us that artificial 3D would inhibit his development of real 3D. He was born with the crossed eye, so he never had stereoscopic vision. It took about 6 months after the surgery to get a bit of depth perception, and about 18 months before he could pass all of the 3D vision tests.

Once he passed all the vision tests, the doc said to avoid artificial 3D, because it could cause the eye to re-cross. Now he's at an in between age when a re-crossed eye could cause him to lose stereoscopic vision permanently. If it re-crossed, and was left untreated for long enough, there's a risk that his brain is flexible enough to drop the neural paths, but not retrain when stereoscopic vision is surgically restored. ie, a very small risk. At some point (16 I think?), he'll be old enough that it's not likely to happen anymore. Given the relative risk/reward of artificial 3D, it's not worth even the tiny probabilities involved.

Yes, there are technical work arounds (ThinkGeek sells some "2D" 3D glasses). If it was something useful, I'd do it, but artificial 3D isn't worth the effort.

Full disclosure: I don't like artificial 3D. I can see full 3D, and I'd still buy a 2DS over a 3DS.

That is true for both HDD and SSD, it's just less common to use it for HDDs. IIRC, 10% of the HDD platter is reserved for sector re-mappings. HDDs usually reserve re-mapping events for things like starting-to-fail sectors and bad sectors created during manufacturing. SSDs use re-mapping to prevent flash wear.

When I wipe an HDD or SSD, it's because they've been replaced after failing SMART. The SMART attribute Reallocated_Sector_Ct tells me how many HDD sectors had my data, but are no longer accessible to me. Some of my older SSDs have a low double-digit Reallocated_Sector_Ct value. That indicates that these SSDs only increment that attribute for re-maps due to failing sectors rather than wear leveling.

For what it's worth, I had that problem until I added much more RAM than I thought I needed.

My original setup, I had 10GB of RAM, used about 4GB in various Mac apps. If I ran a VM that needed 1GB of RAM, the machine was sluggish. Pause the VM, the host sped back up. I tried adding a second HDD, and moving all the VMs onto it. It helped some. The BBOD didn't stick around as long, but the host OS was still annoyingly slow.

Later I needed to run a couple more VMs full time, using an additional 4GB of RAM. I added 16GB of RAM (had to remove 2GB, ending up with 24GB of RAM). Now I can run at least 4 VMs full time, all of them using about 10GB of RAM, without any noticeable host slow down.

I don't remember that part of the Lewinski scandal.

Tri-state booleans FTW.

Forearm strength is a side effect of being 8 hours away from your girlfriend.