So with all of this attention to design, how did they screw up the usability so badly?

Show me ONE kid under 13 who, when presented with a closed XO-1, can get it open in under 3 minutes and who doesn't first remove the battery accidentally. For that matter, show me an adult who can do that. My girlfriend just tried the other night and I had to show her how. It took me a few minutes to figure out, and I took the battery off first, too.

As for the software, why the HELL is there not a consistent 'close' button across all applications?

I really WANTED to believe that this was going to be the slick, intuitive device that everyone said it would be. I tried to make allowances for the fact that it was still in development and might have some rough edges. But after a couple of days of playing with it (my kids gave up in under an hour) I had to admit that the overall user experience was just awful. I could give my kids a Commodore 64 with a text menu based interface to load a few programs (games, word processor, educational stuff, doodle, etc) and they'd have a better time with it than with the XO-1 and Sugar.

I was all for this project until I actually got my hands on the hardware. Only my piece of crap Windows Mobile based smart phone annoys me more than the XO-1, and that's mostly because I'm asking more of it. At least it has a consistent 'close' button, even if it doesn't really kill programs.

It was less invasive and less painful than some procedures I've had in the dentist's (or oral surgeon's) chair.

You can always freeze sperm beforehand if you're worried about reversability.

No one 'cuts your balls'. It's two little cuts in your scrotum under local anesthetic (granted, I didn't care for the injections but it was still better than the dentist) and a few minutes of tying and snipping. I spent the requisite weekend with a bag of frozen peas in my pants afterward, but I didn't even have any significant post-operative pain.

I'm not a super macho guy by any means (come on, I'm posting on Slashdot) but it didn't bother me at all to watch the whole procedure in a mirror.

Male injections aside, birth control usually means asking your SO to mess with their hormones and endure any side effects. Vasectomy isn't perfect, but it's easy, reliable, and has minimal long-term side effects.

I posted a Craigslist job listing (in an area with no Craigslist region of its own and relatively little local usage of the adjoining regions) for a 10-15 hour per week packing and shipping job at a buck an hour over minimum wage and had 75 resumes in 3 days, with most of those applicants willing to commute 15 miles or more for 2-3 hours of work a day!

As it turned out, just having 75 applicants doesn't mean it's easy to find someone who will actually show up on time, follow simple directions, and actually remember when I tell them three times in the same week that UPS next-day air packages always go in the express envelopes if they'll fit.

I think next time I try, I'm going to add some simple qualifying test, like "tell me what the USPS first class international rate is for an 8-ounce package going to the Czech Republic." That way I can just throw out all of the cut-and-paste responses that didn't even read the whole listing. I will most certainly include a note that when I say to email or fax and not call or visit, that means that a call or visit will automatically take you OUT of the running for the job.

So yeah, really really bad signal to noise ratio, even without $10M on the line.

In the embedded world, of course, you still get to use most of these tricks. Not to mention all sorts of low-level hardware hacks. I've written about that here before, but what I find really interesting is that despite the general trend toward ever-faster and more powerful embedded processors, there's at least one new core out there that's SMALLER than its 1970s predecessors.

Freescale's (relatively) new RS08 core is absolutely Spartan. IIRC, the directly addressable memory is 256 bytes. Everything beyond that requires paging. No index register, but it's emulated with a couple of RAM locations. Some single-byte instructions act on only a 4 or 5 bit address space. The advantage of all this minimalism is that the parts can cost under 50 cents each.

So not only is there still some need for all of the old tricks, but there are opportunities to learn NEW ugly hacks! I'm sure someone can come up with something suitably mind-blowing with an addressable index register and a processor architecture that supports self-modifying code.

How about industrial hardware? You'll probably pay at least twice as much as you would for a consumer desktop, but PCs made for industrial control applications tend to be a lot more rugged and build to serve for many years in harsh conditions. Sounds like you don't need a lot of processing power, so you could probably get by with a fanless system and eliminate a major failure (and noise) source.

I haven't bought anything from these guys, so I don't personally know anything about their quality, but SuperLogics has a barebones fanless Atom-based system for $315. Something like that might be a good start.

I agree with the $50,000+ price tag if you're looking to have someone else do everything.

As for building it in the US, have you tried? It *can* be done, but it seems to keep getting more difficult. I do have some PCBs assembled in the US, even though I could have them done in China for less. But for stuff like injection molded cables, I've gotten prices of $8+ each in the US when I can get them for around $1.50 in China. I'm willing to pay 50% to 100% more to keep production in the US, but 500%+ is a little hard to swallow.

There are also independent QA providers that operate in China who (for a price) will keep a close eye on outsourced manufacturing, and it's their job to know all of the little tricks the factories like to pull.

Most of the service-related companies I use in the US (printing, PCB assembly, metal fabrication, axial component sequencing) are small, usually family-owned businesses. They're the only ones who have been able to offer the prices and level of service that make it worth paying a premium for. And too many of the big companies, through arrogance or apathy, won't even touch something a little out of the ordinary. They're like the kids working at McDonald's - if it's not on one of the buttons in front of them, they can't do it.

Depends a lot on the exact value of 'mass market' and the sales potential. Really, a lot of this can be done with contract labor. Once the design is done, there are plenty of places out there (many of them reputable outfits that aren't likely to swipe your IP, even) that can handle component sourcing, injection molding, assembly, and even fulfillment.

Marketing should absolutely not be neglected. Us nerds tend to underestimate (or simply find distasteful) the amount of marketing and promotion that needs to be done to make a product successful.

In this case, I think building a working prototype needs to be the first step, and then you've got something more than just big ideas if you do go looking for partners or VC.

And I have to point out that not everyone WANTS an IPO. The best thing about owning my own company is that I don't have to answer to anyone except my customers. Investors are the last thing I want right now. Yeah, I'll probably never become a billionaire with a privately held company, but I could be happy enough as a millionaire. I have simple tastes; I just want a nice house somewhere, an airstrip, and a T-38 Talon with a Robotech paint job.

Good luck getting source code from Chinese companies. I've tried a few times, and even with a company that I already buy tens of thousands of dollars of equipment from, the answer has always been an emphatic 'no'.

In some cases, the problem may be that the source code isn't theirs. Take two way radios, for example. There are many, many different models on the market that all share the same basic firmware. Each of the companies licenses it from one design house, probably along with some of the hardware designs, too.

It's often hard to tell who's even a manufacturer and who's just a trading company, unless you go and personally tour the factory. Even then they can make it difficult to figure out who's who.

Where I HAVE had a measure of success is in buying partial products. For example, if you look on SparkFun Electronics' website, you'll see a weather sensor assembly. I bought those from a weather station manufacturer in China, and since their usual wireless interface wasn't FCC approved and wasn't needed for my application anyway, I negotiated a deal to buy the bare sensors at a significantly reduced price that still gives them enough extra profit margin to make it worth the hassle (the unneeded touch screen display is the expensive part), while still being far cheaper than designing and producing my own hardware. The reason they're at SparkFun now is that I'm way behind schedule on designing my own electronics package and I got tired of them taking up warehouse space, and sold part of the lot at wholesale.

The good news is that lots of Chinese companies are basically family-owned, and you CAN sometimes get through to the people who make the decisions, where with US companies you might not. But again, I've never had any success getting firmware source (even relatively simple stuff that I could recreate myself in a week) from any of them. Cable assemblies, housings, and so forth, sure. But not a single line of code.

If you're serious about making it happen, consider catching a flight to Hong Kong next month. The Hong Kong Electronics Fair, electronicAsia, and the China Sourcing Fair are all there at the same time around the 12-16th, and the massive Canton Fair (this will be my first year there) is right after that in Guangzhou, but that takes a little more planning.

Just showing up in person and leaving business cards (bring a few hundred, seriously) will get you much better responses later in email. They know you're serious enough to make the trip, at least. That was a benefit I hadn't foreseen my first trip. Also, allow a couple of days extra after the fairs for meetings with vendors if you do make some good contacts.

Also, one book I've found particularly useful in understanding the business culture in China is "The Essential Guide for Buying from China's Manufacturers" by James Lord, ISBN 1419628461. Wish I'd read that before my first trip there. (Tip: Beware the phrase "no problem". =])

If you do make it to Hong Kong, drop me a note and I'll meet you for a beer some time.

scott@argentdata.com

Then again, some of us are very well aware of it and just don't care so much. If I want to post thoughts to a blog that I don't want linked back to me (and I've done so in the past), I'll set up something entirely separate, with a name I've never used before, linked to a new gmail account.

Anyone with half a brain can figure out exactly who I am, where I live, and what I do for a living, starting from this post, in about 20 seconds. Medical conditions and sexual preference might take a little more work, but I'm sure some of it is out there.

Frankly, I don't care. I'm self-employed and don't worry about what an employer might think of me. My friends and family seem to like me well enough despite already knowing that stuff. So long as it's not information that's going to result in identity theft (account numbers and such), there's not much that's worth the effort to conceal.

I hired my first employee about two months ago. I chose him over 70 other applicants (he was my second choice, #1 got an offer with more hours from someone else) largely because his experience with music and computers told me that he was a creative person with a good attitude, which I thought was important despite the fact that the job is not a particularly technical one.

Turns out that creativity doesn't automatically translate to an ability to follow verbal directions, pay attention to detail, and work efficiently without supervision. He's getting better, but his work's still not quite up to par.

Next time I think I'll be looking for a different set of qualifications. Maybe I'll have to give more weight to other keywords, like "varsity cheerleading." If I'm going to pay someone to work slowly and make costly mistakes, they should at least be easy on the eyes. =]

Generally true. Sometimes clever is necessary, though. I do most of my programming these days on embedded systems, where size and speed are absolutely critical. I'll occasionally do something horribly non-standard and convoluted (usually to avoid writing even more annoying inline assembly code), but I've learned to allow about a 3-to-1 comment to code ratio in those cases. Even something not that complicated but just unusual (casting a char array to a function pointer and calling it because that particular buffer is the only one available to hold the flash programming code that has to be copied down to RAM, for example) warrants a clear, concise description of what the hell is going on.

No matter how sure I am that I'll remember how something works and why I did it, I still try to always comment it. I'm sure everyone here (who's been programming for more than 3 years anyway) has gone back to code they wrote 3 years ago and thought "what the hell is this, and what was I thinking?". In my experience that's usually followed by a quick correction, and then after a few hours of chasing down some obscure bug that subsequently appeared, remembering why you did that in the first place and putting it back the way it was.

Planes will have navigation strobes on (they blink), though if you're looking right at the landing lights they can be hard to see sometimes. Satellites will be steady, not blinking, and not changing course. You also won't ever see one flying east-to-west. West-to-east and north-south either way, but not east-west.

Come on, do you really think the native Americans would have ever developed casinos on their own?

<ducks>