Effective immediately, if we receive information indicating that someone is using our services to traffic in stolen intellectual or physical property from Microsoft, we will not inspect a customer’s private content ourselves. Instead, we will refer the matter to law enforcement if further action is required.
This is a far cry from:
As part of the investigation, we undertook a limited review of this third party’s Microsoft operated accounts. While Microsoft’s terms of service make clear our permission for this type of review, this happens only in the most exceptional circumstances. We applied a rigorous process before reviewing such content. In this case, there was a thorough review by a legal team separate from the investigating team and strong evidence of a criminal act that met a standard comparable to that required to obtain a legal order to search other sites. In fact, as noted above, such a court order was issued in other aspects of the investigation.
They were well within their right to do as they did, but what they did was “wrong” in the moral sense of how the general public feels that they should handle things. Going from a “we’ll police this” to a “let’s let law enforcement police this” position in a week or so is a great result from a huge company like Microsoft. If this kind of agility is something that will be a characteristic going forward, I’m optimistic about their chances for relevance in 3-5 years.
One of the two classes that I’m taking this semester is in design of experiments, which is more fun that it might sound. I’ve done these before, and I think that the one that I’m doing now is pretty cool.
I ran an OFAT1 experiment back in middle school using paper airplanes and a contraption that ensured that they’d be propelled with the same force each time. Dad and I built a wooden box, entrapped a rubber band inside of a staple on the front of the box, and placed a clothespin at the back of the box that would hold the tail of the plane such that the nose of the plane a standard length from the front edge of the box. Opening the clothespin launches the plane. I ran five trials for each plane, averaged them, found the variance, and tabulated my standings. I didn’t get first place,2 but I did have fun.
This experiment is more interesting and advanced. I’m running a fractional factorial experiment, which means that I can test four factors in just eight runs without losing anything but higher-order interactions that aren’t likely to matter very much. I’m using the coffeeshop that I spend a lot of time in, largely because I have the trust of the baristas that I’m not going to screw them over. I have four factors:
Time of day of the shift. There are two shifts each day.
The size of the tip jar.
The opacity of the tip jar. These two factors require that we have four tip jars.
Whether the tip jar is seeded or not.
My premise is this: tipping baristas is a social phenomenon. “Do I tip her? All she did was make me a latté,” is a valid question. Anything that we can do to shake up the social norm and show that, yes, people tip baristas is a good thing. But I have no idea what factors will work. That’s why you experiment, people.
I’m sitting here waiting on the end of the third shift under experiment. #4 and #5 happen tomorrow, and the rest will conclude by Thursday. I’m having fun, and so are the baristas, especially since I’m giving them money to participate. I got some very interesting results yesterday, and I’m getting some predictable ones today. Data! I want more data!
I’ll get to do this one more time, either as a full 2^3 factorial that would test every factor combination, or by doing the alternate fraction of this 2^(4-1), which would test the other eight combinations. I’d get to do the former if I have an obvious factor that has no value in the analysis of the first experimental run; if I don’t have conclusive data, I’ll do the latter. I could also run another fractional factorial by dropping a factor and adding another one.
The ladies already want me to run more experiments. I guess that I know what I’m going to be doing on my own time this summer.
One Factor at A Time. In this case, it was the plane itself that changed. ↩
What I really want is for my online backup systems to be smart. If I’m off of my home networks, I should be allowed to say, “Don’t back up again until I get home,” rather than having to look at the three services I now use and say, “Pause for N minutes.” I forgot this until just now, and I’ve been soaking the meager transfer capacity of the Cafe 153 network with my backups.
Why three services? I’m turning my knowledge of backups into a micro-business, and that means that I need to test all of the big players. Fun times.
You got new stuff for Christmas, and now is the time to start backing it up so you can quit worrying about it.
A pretty young woman and I had a conversation about this at the Apple Store earlier today, and she wanted to know more about it. This is published because of that conversation; I’ve been hacking away at it for a while now.
So hi, nice young lady whose name I didn’t get! I feel like an idiot right now.
Let me get a little bit of this out of the way: I have written about backups in 2009 and 2011. In the first one, I talk about a belt-and-suspenders approach; in the second, I talk about belt-and-suspenders supported by close air support. Now I have belt, suspenders, close air support, and Navy SEALs. Or something — I don’t know, I’m an Air Force brat, and we don’t know crap about the real military. [Sorry, Dad.]
Oh, and then there was the time that I was sitting in the floor of my downstairs bathroom on April 27th, 2011. With tornadoes all around me, I said aloud, “Hey, at least my data is in a data center far from Huntsville.” All of my data would be safe: photos, music, you have it. I could rebuild my computers from the last known safe state. It would’ve worked: the cavalry would’ve come over the hill.
So, back up your shit. Here’s how.
I use what’s called a 4-2-2 system: 4 copies of my data, 2 of which are local, 2 of which are offsite.
The first local copy is a cloned backup of my boot volume via Shirt Pocket’s SuperDuper! This gets me something that I can boot from at any time. This was prominently featured in 2011 and December 2013.
The second local copy is Time Machine. I’ve written before that Time Machine Just Works, but that’s really not true. Time Machine has lots of problems, and I wouldn’t trust it by itself. But for A) a second local copy that I don’t have to be primarily reliant upon and B) the value that it provides in being incremental backup for file retrieval, etc., it has value.
Seeding an online backup with either one of those services is going to take a long time. Backing up 28GB off of my Macbook Air took the better part of a week. It took a couple of weeks for the main boot volume on my iMac to seed, and I was uploading the first (of three) Drobo volumes when my iMac started thrashing around on the floor like an angry toddler.
Since I mentioned the Air, I want to make a point: I use Time Machine (to a Time Capsule on my network, a replacement for one that got zapped a while back) for incremental backups as well as CrashPlan and now Arq. For some time, I said that I didn’t need cloned backup of the Air. For one, it’s a problem on my end, because I would have to dock a hard drive1 to make it work. SuperDuper! will backup-on-mount, meaning that it will work … when I remember to connect the drive. I am the weak link there, because I forget to back up even though I know that I should. However, I have a reminder in my GTD setup2 that gets me a record of when I’ve last backed up as well as a nudge to do so. I’ve generally stayed current, but not always.
So what do I recommend?
First off, I recommend all four backup solutions. The two local backup solutions have different features and restoration times; the two offline backup solutions are both good, and I’ll leave it to you to pick. But I really do recommend having one local and one offsite at a minimum, and if you’re going to pick three, have both local copies. Restoring from an online backup is very time-intensive, especially if you live in a bandwidth backwater like North Alabama.3
I recognize that not everyone had the resources that I had to bring to bear when I got started with this. I recommend a cloned drive that is automated or well-maintained (but preferably the former). You do not want to be thinking the following when you have a computer emergency: “So when was the last time that I backed up to this drive?”
For desktops, you have no excuse to not have a clone attached at all times. For laptops, you do, but consider this as well: don’t carry your clone with you everywhere. Leave it at home. A solid scenario for using that clone is, “Someone stole my laptop and I need to get back up and running with this replacement laptop paid for with insurance money!” Do you feel smarter? You should.
If you have just two backups, I recommend a clone over Time Machine. TM is convenient but can be flaky. If I’m concerned with backup first, I’m going offsite. Also, if you’re going to have just two copies of your data, one of them should be off-site.
I recommend CrashPlan over Glacier, because most of you aren’t going to need/want something in Glacier’s sphere. Those who do were probably criticizing me for this choice, but they probably weren’t reading this in the first place other than purely to criticize me. Quit trying to be John Siracusa. You are probably not John Siracusa; if you are, John, this is crazy: here’s my number4, call me maybe?
I’ve got the comment box below for comments. I’ll also be posting this to Facebook and Twitter per uzhe, so if you comment there, I’ll see it, too. You can also email me at email@example.com, and John, my number is in that footnote.5
I have a spare on the floor in my office waiting for this to happen. It’s been there for a few months. ↩
I will write about that at some point, I promise. ↩
Seriously, we put men on the moon and then get treated like this? “Come on!” —GOB ↩
My only response was: “I’m doubling your [hosting] fees.” Four minutes later, Stephen told me that Misty might be in labor. It was a bit of a day.
I was mainly excited that my friends would be bringing another awesome kid into the world (and Liza is indeed awesome), but I was also worried about my server going into massive heat death. Single server, four CPUs, moderate amount of RAM, SCSI hard drives, Web and SQL stored on the same drives. You can imagine how that went.
I only came across this anniversary the other day when one of my choir kids posted the following:
I responded with a link to LOLTrek, and when I did so, I noticed the date stamp. I thought I’d bring it up, because it still makes me laugh. The funniest thing for me are the commercials.
Take, for instance, a self-driving car. One of the assumptions we have is that allowing computers to drive cars will allow a lot more cars to be on the road, since computers are better drivers than humans (a fact I don’t want to dispute). But imagine we do fit 30% more cars on the road. Imagine a traffic disruption. There will surely be far fewer traffic disruptions because computers are better drivers than humans. But when they do occur, they will cause massively more congestion than now, because the system will have been optimised that much further.
A driverless car will be best implemented when it communicates with its peers in a networked way that mimics the old CB network band: “Get off at Exit 351 and take US 31 north; I-65 is a parking lot.” But there’s fragility, of course: not all cars will have humans out of the loop, not everyone will have a car that communicates in the same way, there will be network outages, etc. That’s why peer-to-peer on open technologies will make that work.
See, my technological bias is showing. But I will also admit my own bias against driverless cars: I’d rather drive, and if not, I’d rather take mass transit to have it be worthwhile.
This is a nerdy post. I’m telling you that before you get sucked in. In short, I use my computer, which resides in my office, to send a continuous alarm to a set of speakers in my bedroom. I do this with a couple of AppleScripts and a cronjob. If what I said made you wonder if I was speaking a second language, this is not the post for you. However, as some of you have expressed a little bit of wonder at my Rube Goldberg alarm clock, I decided that I’d write it up. Here goes!
Why All the Trouble?
I’m a night owl. My most productive hours are 2200-0200, as shown by the timestamp on this post. I sleep better in the morning after the sun has come up than I do most of the time at night. My best sleep hours start at 0400. I have more than a year’s worth of data to prove this, as I’ve been tracking my sleep with Sleep Cycle alarm clock for more than two years. My ideal work schedule would be 1200-2100, as I could sleep in until 1030 or so.
The world does not live on my schedule.
My mother can tell you that waking me up is not easy. [I told her about this setup and she laughed for like 10 seconds. “Does it work? I bet it doesn’t work.” THANKS, MOM.] When I am at their house, she will stand at the door and repeat my name for a minute or two before I sleepily wonder just what in tarnation is going on. My MSMS roommates will tell you that I can get out of bed, walk across the room, turn the alarm off, get back in bed, and go right back to sleep. I can tell you that I’ve moved my alarm clock any number of times. I used to re-arrange my bedroom furniture every sixth months to fight this.
Yet what I’m doing right now is working. I explained it a couple of weeks ago on Facebook, and they were stunned to see the process. As such, I feel that I owe you an explanation.
What’s Happening Here?
I am piping audio around my house. This is starting to come into vogue with hardware and software solutions that replace things like Sonus systems, setups that run into the high hundreds and low thousands of dollars. This isn’t necessary anymore, especially if you use a Mac.
Any old Mac
Airport Express (AE)
Stereo speakers, self-powered, that accept 1/8″ input.
To make this work, connect the speakers to the AE. The AE serves as an AirPlay point that can be used for all sorts of things, including this alarm system. I pipe all sorts of audio to my bedroom, mainly a Web stream of BBC World Service and radio captured on my radioSHARK. [I’ll talk about these later, especially if there’s interest. The AppleScript that I have for the BBCWS stream is kinda fun.]
Meridian. [This is shareware/abandonware that is no longer in development.]
The night before, I use Audacity record an M4A of what I want to hear the next morning when I am waking up. I Export this file to ~/Documents/Alarms/. I name the file YYYYMMDD of the date I’ll be waking up. I found out tonight that using the same datestamp twice will cause the next step to fail.
I fire off meridian-alarms.scpt using Launchbar. [If you’re trying to do this and haven’t been using Launchbar or Quicksilver: WTF, yo.] This 1) quits Meridian then 2) moves the file from ~/Documents/Alarms/ to ~/Library/Sounds/ and 3) re-activates Meridian. The quit/restart option is required for Meridian to know that the new alarm exists. [I’d like to thank the dev for telling me how to do this.] The only hitch is that you do have to click on a dialog box to quit.
Go into the preferences for the Alarm and set the new wakeup sound to the file that you’ve just made. Failure to do this gets you the previous day’s alarm. Make sure that the Continuous option is checked, or you’ll hear yourself for 12 seconds and nevermore.
Set up a fire-off time with Cronnix to run meridian-pipe.scpt. This will set you up for piped audio from Meridian to those speakers at alarm time.
Come wakey time, your cronjob will fire. Airfoil will be routing sound from Meridian to your bedroom speakers. Once your alarm fires, you’ll be hearing yourself from the night before.
Here are some screenshots that may help you understand what’s happening:
To kill the alarm, you’re going to have to get out of bed, walk out of your bedroom, find the computer in whatever room you keep, sit down, and turn the alarm off. I find that the combination of a) hearing why I need to be awake and b) having to do a lot of work to stay asleep makes things work for me.
Scan the release into Delicious Library. This works for purchased CDs only, of course. Concert recordings don’t go in here, as I didn’t pay to take ownership of them. I put things into DL so I can loan them out and know who has what. Also, I have this as a record for insurance purposes.
Rip the CD in Apple Lossless. I’d use FLAC for maximum interoperability, but Apple only uses their lossless format for iTunes, and as I use iTunes Match to move music onto my iDevices, I knuckle under and use their format. I don’t see it going away anytime soon, so I don’t feel like I’m investing time ripping into a format that I won’t use. When I compare this to the 3-4 different lossy encodings I used from 2003-2011, it’s not a big concern.2 Now that I have two large HDD arrays, I really don’t worry about storage space.
Run the rip/encode through MusicBrainz’s Picard tagger.
Add in the highest-quality cover art I can find. I really should be making my own with the scanner I have, but I’m lame.
I’d put in lyrics—and I care about that in a theoretical way—but there’s little practical value in doing so.
That’s where I’m at these days. I don’t see this methodology changing much given that I’m using a stable lossless codec.
Why am I writing this update? I referenced something the other day that referenced the original post. Holy nine years, Batman. ↩
That I kept moving the target was a big part of the problem. ↩
If someone buys my music on iTunes, Amazon, or in a record store (remember those?), let alone streams it on Spotify, it’s all short-term money. That might be the last interaction I have with that particular fan. But if I give that fan the same record for free in exchange for a connection (an e-mail and a zip code), I can make that same money, if not double or triple that amount, over time. And “over time” is key, since the ultimate career success is sustainability. Longevity. See, the reality is that out of a $10 iTunes album sale, I probably net around a dollar. So if I give that record away, and as a result am able to get that fan out to a concert (I can use their zip code to specifically promote my shows in their area), I make approximately $10 back, and twice that if they visit the merch table. I can sell them an older/newer album and make approximately $10 back. The point is, if I can find some organic way to creatively engage them in a paid follow-up transaction, I increase my revenue 10 times on any one of these interactions.
I’m a fan of Mpix‘s photographic products. I’ve gotten good results with every single order from them, and the prices are within what I would expect to pay. For display purposes, I really like the standout: your photographic print mounted on gatorfoam, banded as you like, ready to hang.
I clumsily knocked one of my standouts off the wall the other day, ripping the hanger out from the wall in the process. I’ve got to re-hang it, so it made me think of what it takes to hang them. I’ve learned this by trial and error, but you don’t have to do that. Here are the steps you’ll want to use:
Measure the standout. In my experience, all of the standouts are short in both dimensions, just like dimensional lumber. The 11″x14″ shown above is short by 1/8″ in both directions. This will let you set up an envelope.
Regardless of size, the edges of the hangar holes are 2″ in from the corner. Note that this is not the hole center but the tangential edges of the hole.
The hangar holes are 0.5″ in diameter. You can do the math and say, “So that makes the hole center 2.25″ in from each edge, right?” Indeed it does: circle gets the square!
From there, it’s like any other picture-hanging extravaganza: find a level, determine your placement, mark a level line, mark your holes.
When it comes to hanging, don’t use a traditional nail or picture hanger. I’ve had the best success using push pins. Standouts are really light, even both of the 24″x30″ ones I’ve bought and hung. Four push pins are going to do a great job. You can also do some tweaks to get the fit that you want by skewing them toward or away the corners of the standout. Because the head of a push pin is closer to 0.25″ in diameter, there is some play in the placement.