Limitless LEDs - Not So Limitless

I recently leapt into the wifi LED trend/fad with the cheapest wifi-controller LED replacement bulbs I could find, from limitlessled.com. The most important take away is that they are not, in fact, limitless, but at their price they may still be exactly what you're looking for.

The Pros are shorter than the Cons, but not because these are bad lights - pros are just uncomplicated and don't require any workarounds. That said they clearly need work - this seems to be version 2 or 3 for them. They need a v3 or 4 before I'd recommend them without any "buts," and before I'll buy more.

Pros:

• You can fade them from ~10% to 100% brightness with either your phone or a remote that comes with the lights. Judging by documentation it sounds like this is 1w-9w of power usage.
• They can be a short list of colors, though they're dimmer than white mode; those colors are also dimmable.
• They offer both cool white and warm white models - cool is flat white and warm is incandescent-style white.
• They're programmable via a Web API.
• If you really know what you're doing you can manipulate them across the internet, by punching a hole in your router's firewall.
• The lights don't get very hot.
• They fit a standard 40-150w US bulb fitting.
• They're about the same size as normal bulbs so fit isn't an issue.
• Their support staff (maybe the owner/inventor?) is very responsive via email.
• They don't break easily - all 4 of the ones shipped to me have worked without issue.
• The dimming is fairly precise, so you can dim to the exact lighting conditions you need - for example low ambient light while watching a projector.
• You can turn them off from bed, via your phone, when you realize you forgot to turn the lights off.

Cons:

• All of the light is projected forward, in a 180 degree hemisphere pointing out of the top of the bulb.
• The color options are very limited:
• The color options are all fully-saturated - you basically can have gaudy red, orange, yellow, etc, and no less-saturated colors like a warm white.
• Internally there are red, green, blue, and white LEDs, but you can't turn them on at the same time except in the predefined patterns the designer came up with. The whites can't turn on when any of the color LEDs are on.
• This path around the saturated color wheel is pretty bumpy, with cyan (blue-green) being near-white, and yellow being fairly close as well, while red is quite dim, as is pure blue.
• Internally there are 3 of each color (R, G, B) and 8 white LEDs. This means White Mode is much, much brighter than most colors. To get a well-lit room you need 2-3x the bulbs in color modes that you would in white mode, which further strands the RGB function as a gimmick.
• If you want to use these from the wider internet, rather than connected to your wifi router, you need to punch a hole in your firewall and discard security. Anyone with a basic understanding of the product can then manipulate your lights from anywhere on the internet. The FAQ on their site is disingenuous about this: "with 128bit encryption access is very secure." Yes... but that's the encryption on your home wifi signal. Once you punch a hole in your firewall that encryption's irrelevant.
• You can't control more than 4 bulb "groups" at a time, and the groups are defined by tapping a Group button on the remote within 3 seconds of turning on a bulb via a light switch. If you have more than 4 rooms to light, this puts you in an awkward spot.
• You can apparently resolve this by buying a second or third Wifi Bridge, although I'm skeptical of the resulting wifi noise levels in such a scenario.
• Because of the on-then-press pattern, it would be difficult to mix and match groups to make up for the RGB limitations.
• The Bulb Groups are Per-Controller, so the Remote (a Controller) seems to have one idea of what a group is, and each Wifi Bridge another. To use both the remote and your phone you need to go through the Group assignment process twice.
• You can't use bulbs individually/outside a Group so the Group Assignment process is mandatory, and probably confusing for new users.
• The wifi setup process is poorly documented and quite difficult. It's unlikely most customers would be able to figure it out. The manual isn't really for the hardware that arrives, nor is it for the app you'll find online (it appears to be for a much older version), and leaves out a lot of critical details. The app and hardware have some issues that if documented wouldn't be so bad:
• To access the lights from your phone you setup a Wifi Bridge. The Bridge is very confusing.
• The app does a terrible job of stepping you through the necessary steps, which are:
• Connecting to the Bridge as a Router
• Rebooting it into a different mode where it's only accessible via the app talking through your Router, and no longer offering itself up directly as a router.
• Accessing it via the app once your phone's wifi is talking to your router, and the Bridge has finished booting, and the app... gets around to noticing this.
• When the Bridge switches on, it does something really strange. It registers with the DHCP provider on your router ~100 times, consuming all the dynamic IPs on it, before finally settling on one. This can take several minutes, during which it's unavailable. If you're a new user with no documentation trying to figure out what you're even supposed to be doing here, this period of unavailability will lead you to believe it's simply not working.
• Any mistake with the setup process requires Reset to Hardware Defaults on the Bridge, which is a very hard to access recessed button needing a paperclip - I don't know about you but I don't own paper clips. I had to press this button ~10 times to sort out exactly what ways the Bridge was failing. The button should be a simple button.
   
• The API is a bit of a false lure. Things like Disco Modes and color control give the impression the lights hold a lot of potential waiting to be unlocked by usage of the API, but it turns out those limited colors and modes are in the actual bulbs, and can't be modified without the manufacturer coming out with better lights that have solved these problems and thought this through.

All in all I am still glad I bought these. I have wirelessly dimmable lights in my bedroom and living room. That means I can get just the right light level for watching TV or work. I can increasingly dim the lights when it's getting time to pass out, which surprisingly works better at getting me to sleep on time than it seems it should. And I can tap a button on my phone to turn all the lights off when I'm about to fall asleep, or, if I wake up and realize I didn't turn them off, I don't need to get up to resolve that. Finally, my power usage is down from 60w to at most 9w on each bulb - less when dimmed.

I likely won't be buying more Limitless LEDs until the following issues are fixed:

• Lights whose individual LEDs (both white and color) can be manipulated by the API, instead of fixed patterns.
• A wireless bridge that can handle infinite, or practically-infinite, lights, individually, Groups should be a layer on top of that handled in outside software, not something I'm locked into when setting up the lights. If I want to group the Living Room and Kitchen for now and later the Kitchen and Garage, I should be able to, even if those rooms each have 4 bulbs.
• The Bridge setup process should be at least documented and ideally, greatly improved.

Some nice to haves:

• It would be nice if there were more models. In particular if the bulb can do 9w 800 lumens, why not offer one that can do 27w 2400 lumens? They're dimmable so it's not like that brightness level would be mandatory for buying the bulb.
• It'd be nice if there were LED strips, omni bulbs (not just the current 180), and spotlights that spoke the same protocol as the rest of the bulbs.

The American Mafia

When people think "Mafia" they think men in suits firing machine guns, but the real mafia is about making millions by establishing a pseudo-government which collects taxes in the form of bribes, corruption, over-priced rent and protection payments. Italy's mafia, Afghanistan's Taliban, and Pakistan's ISI all operate this way, and by this definition, the US has a mafia of its own, that occupies a portion of Wall Street.

The good news is some portion of US mafia went to prison. The bad news is they got right back out, seemingly due to mistakes made by the government prosecution, maybe due to bribes. More bad news: The case shows clear evidence of the convicts bribing government officials, including former New Mexico Governor Bill Richardson for $100,000, in that case in exchange for $1.5million of New Mexico's money.

There are too many schemes employed by those in the trial to cover here - see the article - but generally, someone would hire them to invest money, and they'd take a large chunk of it for themselves. To keep states hiring them, they'd bribe politicians - and by bribe I mean fund their SuperPACs, which the US defines as legal and not a bribe.

The list of things we need to change to stop this secondary government to operate is pretty long. That politicians require campaign funding (bribes) to operate is a major factor. Another is the way US states continue to award contracts worth more than $1 million to a single vendor, in a confused belief it will allow them to bid it out and get the best deal. Inevitably any large contract from a single provider leads to incredible corruption, because even a .1% kickback on that sort of contract is worth 10s of thousands of dollars, enough to get a lot of iffy people paid off.

Maybe the worst part is that everyone involved in all the schemes uncovered continue to operate banks and other parts of Wall Street - they didn't lose their jobs and once they escape prison they're often promoted or given a bonus. The author of this article is clearly frustrated when he proposes we have them all killed. But we should at least ban them from the Finance industry, politics, and lobbying.

Rolling Stone: The Scam Wall Street Learned from the Mafia

Bloomberg: Bankers Win Reversal of Convictions for Bid Rigging

Adventures in E-Voting Precede State Experiments

Although voting for Congress remains behind the times, the Academy Awards began experimenting with e-voting in 2013. Unfortunately, it didn't go very well. They set the security on it so high that many members couldn't figure out how to prove who they were and login. Then again, the median age of its 5765 members is 62.

Your average Oscars E-Voter doing the deed

Despite some older members predicting most people would give up and not vote, it actually turned out to be their highest voter turnout on record. The 2014 e-voting process went by without complaint, so they appear to have fixed the below issues.

The e-voting process involved a 6 step process:

1) Pay your dues, which ties your credit card and that verification system to you.

2) Receive a Voter ID number in the mail, which adds your address to the level of verification.

3) User your Voter ID to go online and setup a Voter Password, with numerous annoying requirements like caps, special characters, etc  You also enter your cellphone number.

4) When the voting period opens, login with your Member Password, then your Voter Password.

5) Once you enter the second password, you receive a text with a security code. You then type that in, and finally, can vote. This would be the final step, but it seems most members with trouble forgot their second password.

6) The reset password process was very badly designed. To get a reset you had to wait about 24hrs.

On step 3, it's long been shown complexity requirements tend not to make passwords more secure. Only long-length requirements like 12+ chars with no max length have been shown to significantly improve security of a password; complexity requirements usually encourage more patterned passwords, which make them easier to hack, and more passwords forgotten.

In step 3, they also complained that the password box is a standard password box, with asterisks hiding what you're typing. I have to say I agree. The idea that most people are passwording into things in someplace where someone could be looking over their shoulder, stealing their password is a poor assumption. Hiding passwords should be opt-in, not assumed, and if you want a long, complex password - you've gotta make it easy to see that long, complex gobbledygood you're typing in.

Step 2, where the voter is notified by paper mail of an electronic voting system seems to be an unnecessary inconvenience; if they were texted or emailed a link, they could skip this and potentially the second password, at least if you could reasonably believe/verify the person you were texting/emailing was the right person.

But, by involving their physical address, the Oscars' system provides some insight into what would be involved in e-voting in local and federal elections, since legally all that's required to prove you're you in US elections is your address. A ballot is sent to your home, or you provide your address at a polling place, good enough - we believe you. Given the low legal burden, this e-voting system - with some simplifications - could plausibly be used in US elections. Worth watching to see if they cope with any fraud or hacks.

http://www.npr.org/2013/01/03/168560518/e-vote-hiccups-delay-oscar-balloting

http://www.hollywoodreporter.com/race/oscars-e-voting-problems-worse-406417

http://www.huffingtonpost.com/2013/01/10/oscar-voting-concerns_n_2451160.html

http://www.hollywoodreporter.com/race/audience-heckling-disneys-legacy-a-669305