In other news:

Computer programmers weary of optimizing code

Auto engineers weary of increasing fuel economy

Home owners weary of insulating their houses

Electricity costs money. Reducing the cooling costs of data centers isn't a green issue; it's a cost issue. TFA mentions this specifically:

Steven Brill and his analysts have pounded the table on the importance for IT to pay the electric bill so they understand just how much power they consume.

so I find it odd that the take-away is "green fatigue."

I've lost data with ext3.

I've had data corruptions with reiserfs.

I've lost data with ext4 (which happened to be the most frustrating, tedious, and complete failure of all).

Most recently, I had some HD failures on a fully RAID-1'd server running entirely on XFS, and had to re-install the OS from scratch and restore from backups. The new install was onto btrfs.

I've had partitions running on btrfs for a little over a year, and have not yet lost data on these, but it's just a matter of time; I will lose data. I used to blame it on cheap drives, but I've seen SMART failures on young Seagates so I'm now convinced there's no such thing as a high quality, high density drive. At the moment, I find btrfs easy to use (intuitive and simple), and full-featured, so it's what I'm currently using. But I suffer from no illusions; at some point, I will have FS corruptions and have to restore from backups, and I can only hope that any FS corruptions won't go undetected and be propagated to my backups for very long before that happens. Failures are inevitable no matter what I use, so now I value simplicity, convenience, speed ... and backups.

Right now, btrfs beats the alternatives for convenience and features. I put my trust in backups, not file systems, and value is in features and convenience, not some false perception of safety or reliability.

Amen.

I keep seeing this assertion. If applications aggressively grab memory and resources that they might use, and if I'm a user who uses the computer for more than just a single-domain application (say, web browsing), then I'm going to encounter a lot of OS swapping as I jump around between applications. If the OS has a free buffer of RAM, then new applications I open are going to open more quickly as the OS doesn't first have to swap opt currently in-use memory.

I can't help but think that this philosophy that unused RAM is wasted RAM is what's led to application bloat. I see this most in my career in situations such as frantic last-minute GC tuning resulting from UAT load testing; or when a developer discovers that some new feature is going to push the app over some threshold, and they have to go back and spend extra time analyzing and tuning other parts of the code base that don't, strictly, have anything to do with the feature they were task with developing.

It's as if, when it became widely recognized that eager optimization had consequences, the industry threw the baby out with the bathwater and took it as permission to entirely ignore resource use considerations during the initial design and development phase. This philosophy has apparently permeated through to the general computing zeitgeist, as evidenced by your (commonly held, and understandable) conviction.

Don't you know American companies are crooked, evil, liars?

Tip: sarcasm works best when you don't skirt so close to the truth.

In my limited experience judges don't find it clever if you violate the spirit of the law without violating the letter.

So, what you're saying is that we don't have rule of law in the US, just rule of judge's opinion? It's one thing for a judge to interpret in the case of ambiguity, but you can't convict somebody of violating the spirit of the law. That's why suspects "get off on technicalities."

The law must be a strict definition, or it is subject to being applied differently to different people, based usually on one person's personal biases. Historically, this is Not A Good Thing(tm), and you see evidence of bias abuse in sentencing.

Because unless the editor is truly idempotent, in the formatting, after you've reformatted, and then reformatted again, your version control system may think you changed lines that you didn't. This causes erroneous conflicts in merging, and renders history annotation useless.

Which allows me to rant a little: one of the best and worst things about Go is gofmt . It's nice to have such a tool; it's not so nice that it defaults to using the OS's line endings. If you're going to define whitespace rules-of-thumb, don't wimp out when it comes to line endings.

The iPhone also represented a huge effort ... radically different from other cell phones

Are you suggesting that there weren't full-screen, touch-sensitive slate phones prior to the iPhone? You can go back as far as 2000 (five years before Apple started designing the iPhone, and seven years before it was first sold) to the Ericsson R380; the Sony/Ericsson P800 was even closer -- if you removed the clip-on keyboard, it was the same form factor as the iPhone, with a touch screen and full PDA functions. So how, exactly, was the iPhone "radically different" from other cell phones?

Recently, I've come to the conclusion that products are irrelevant; popularity is all in branding and marketing. Us developers (of hardware and software) like to kid ourselves into thinking that we're the ones who do the "real work," but really, it's the sales and marketing people who are the backbone. Apple didn't "invent" the smartphone, any more than they invented the MP3 player (they were three years late on that), or the laptop, or the slate PC (again, late by several years), or any of the other stuff they've been successful with in the past ten years.. They've just been able to corner the "sexy" market, through good advertising and branding. I think that since Jobs returned to the company, they also payed more attention to quality and product polish, and were willing to sacrifice volume in the increased costs that often incurred. But I really think what makes a successful product is the cult of personality.

Other examples:

• Microsoft. There's almost always been a better competing product to whatever Microsoft is selling, but Microsoft managed to capture the Business sector by its early and intimate association with IBM. Even OS/2, an arguably better OS, couldn't wrestle that crown away, and that's because they didn't have Bill Gates, not because it was a technically inferior product.
• Linux. Minix predates Linux, and had the potential to be as successful as Linux, and can be argued to have a better architecture, but Tanenbaum had different priorities and isn't, I dare suggest, the personality that Linus is. Or, if you don't like microkernels, BSD. Same thing: they don't lack technology, they lack Linus.
• Java. There are a lot of at least equivalent languages out there, even if you restrict yourself to the OO space, but none of them had Sun behind it, pushing Java so aggressively. I'm not going to give McNeally or Gosling credit for that; I don't think there was a personality behind that one, just aggressive and persistent marketing.

The only advantage to base 10 is that you've been forced to learn the multiplication tables in it. That's the only advantage.

However, there are provable mathematical advantages to base 12; mainly, that it's highly composite and is the smallest such number to include the numbers between 1 and 4 as factors. We use it in our clocks, in counting our eggs, in dividing feet into inches, in the zodiak, and in dividing a year into months. We see it in nature in the number of full moons in a year and in the platonic solids -- there is a dodecahedron (and tetrahedron and cube, and 3 and 4 both divide evenly into 12), but there are no platonic solids with either 5 or 10 sides. The only thing 10 has going for it -- the only reason why you count in base 10 -- is that you have 10 fingers. However, you do have 12 segments in your fingers (excluding those in your thumb, which you'd use for place-holding), which means that you can count up to 12 * 12 = 144 on your two hands.

The logical thing to do would be to retrain everybody to count in base-12, re-calibrate metric to use base 12, and then use that. Failing that, I'd suggest that leaving things in the US in imperial is preferable: it's not perfect, but it's superior to metric.

It hasn't failed me yet, and has the additional advantage of providing a sort of upgradability that's sorely lacking from most devices these days. I can swap out my carrier, or my cellular technology, without having to buy a whole new tablet. I just wish the rest of the components were as easily upgradable.

IMO, CDMA's biggest disadvantage is that the US is the only country in the world that uses it. I've never had a problem roaming internationally with my GSM (and now UMTS) phones, and that's important to me. This is no reflection on the technical merits of CDMA, obviously. Just an observation.

Not because I wanted to buy Android instead of a dedicated GPS, but because I've already got four Android phones in the house and didn't want to buy anything.

In my case, I'm going back-packing for 5 days; we have paper maps and compasses, but I want to bring my phone along and see how it does. I have a small, portable solar charger that I'm bringing as well. Here's what I've discovered:

First, I'm taking my Nexus One. I have to take the phone with me anyway; I just won't leave it in the car when we hit the trail. With the screen off most of the time, but with the GPS on and a tracking application running, I got about 7 hours of continuous running before the battery hit critical. All wireless was off; theoretically, the only things running during that time were the CPU and the GPS chip. I used the display for maybe 20 minutes during that whole time. I expect that, with some coddling, this amount of time would be serviceable -- and it'd certainly be a fair emergency device.

The Nexus One compass -- the magnetic one -- is way accurate! I walked around a bunch with a Suunto Global magnetic compass, and the Nexus kept up admirably!

I think I tried every free or demo GPS map program in the market, and the one I settled on was OruxMaps. RMaps and Maveric are interesting and have useful features, but OruxMaps turned out to be the easiest to build up maps of my destination with, and it provided all of the basic features that I wanted. It has a built-in map builder which takes a little fiddling to figure out, but is pretty easy to use once you do. I did this all over Wifi (which is going to be faster than cell data, anyway), so no cell plan is required (although a WAP and internet access still is).

The display is the biggest battery drain, obviously. With that on constantly, you're not going to get more than an hour of battery out of it. However, the Nexus is smaller than any GPS with a color screen that I've seen; attach enough external battery pack (through USB cable) to make it as big as your average Garmin, and I think the battery life would be comparable. As others have said, the quickest and easiest thing to do is just buy a dedicated GPS; you'll get less for your money, but if that's all he wants to use it for, I don't think it's worth the extra effort to set an Android device up as a dedicated GPS.

Ditto. Actually, I was using ShapeWriter until Swype came out... they both have their nice points, but both of them are nice, and a lot of the time, I actually prefer them to hard keyboards.

ShapeWriter has a really clever capitalization mechanism that I miss in Swype. Often. Swype has slightly better matching. ShapeWriter will insert a space between a period and the next word; Swype doesn't (grrr!). ShapeWriter has an annoying feature where, if the text entry ends with punctuation, the editing of misspelled words doesn't work. Swype requires you to actually swipe over apostrophes to get them (it doesn't recognize "its" as potentially being "it's"). They both have "alternate" keyboards, but ShapeWriter's alternate is much more useful (bigger keys, focused on numeric entry) -- although, both make getting to some common keys (:, /) uncommonly difficult. For some reason, I find Swype much easier to use if I'm tap-typing -- and ShapeWriter is almost impossible to use for password entry (if you're like me and use mixed-case passwords), whereas Swype is useful.

They're both good. I don't know about Swype, but I get regular updates from ShapeWriter. I've been using Swype for the past couple of weeks; I think it annoys me less, but they're pretty darned close.

It doesn't get any better than that. But feeding the trolls is fun!

I don't know about other people with 4 year CS degrees, but I took three years of calculus (in addition to numerous other math classes); is the poster suggesting that either vocational schools cram four years of math into a two year program, or that math isn't an important part of computer science? Probably the latter. Which would explain a lot of things I've seen in industry over the years, actually.

--- SER

I've been running a Transcend 64GB SSD (ca. $200, PATA -- not high-end, definitely) in my laptop for 10 months. It's on all the time, except when I suspend it for transportation. It is running Ubuntu, and I've got a current uptime of 30 days. I'm a software developer; I download and install betas of OpenOffice, I upgrade Netbeans and Eclipse regularly, update and build software (including one work project that's over 1GB built), and generally trash the hard drive. I haven't had any trouble with it, at all.

I also installed an OCZ 64GB SATA SSD in my wife's laptop since mid-June (so, 4.5 months). Hers is more often in sleep mode than in use, since she has a separate, work, laptop. She uses it for writing, homework, browsing, and so on -- light duty. No problems there, either.

Neither laptop is configured to run /var/log or /tmp in RAM, or anything fancy. Both are configured with ext3 (although mine has a BTRFS partition, for play) with normal journalling.

I'm happy with mine. I don't notice the speed increase, if there is any; I mostly went this route to (a) reduce the heat, (b) reduce power consumption, and (c) reduce noise. My wife's Acer Timeline is particularly silent, as the CPU fan never comes on. I don't know if I'd put SSDs in my server; HDs are too ridiculously cheap, and I don't need extra speed for my modest music/file/web server uses. But, so far, I've been entirely satisfied with their reliability.

I do back both machines up nightly, just in case.