Winter Tuneup

The new year is upon us. The tech previews have been thrown out there. It’s time to turn back to programming. First thing I did, was purchase 4Gb of ram for an old iMac I rescued from my mom’s house a couple of years ago. My mom has been a lifelong “I’m a PC” person, but a few years back my brother thought it would be nicer if she got into the Mac slipstream. Switching costs between the two platforms at the time was huge. None of her apps worked, and the whole thing was just way too different.

So, I bought her an HP desktop, and carted away the Mac, with it’s 1Gb of RAM. Recently, I wanted to do some development using that machine, and I found that I needed at least 2Gb of RAM. Well, RAM is cheap, so I went for 4Gb instead, as that’s the max, and I’ll never do another upgrade on that machine again.

Then there’s this desktop machine I’ve been using for the past 4 years. AMD Phenom II X4 965 Processor, 3.4 HHz, 8GB RAM, 64-bit Windows 7. This machine has been bread and butter, with more crap installed and uninstalled than I care to remember. It could probably benefit from a fresh install of some OS or another. Possibly a combination of Windows 7, Windows 8 and Ubuntu, would be beneficial. I really don’t have any complaints about this machine, but I did recently purchase a nice wireless adapter for it. The OurLink USB dongle was giving decent performance, but it kept disconnecting and asking for the router’s password all the time.

I purchased a Rosewill Dual Band PCIE car 802.11 a/b/g/n (N900PCE). It says right on the box that it will give me 450mbpsX2. That means 450Mbps both dowload and upload! That’s nice as the ASUS router promises the same thing. So, I did a on this setup and:

5ms (ping>      103mbps (download)     128mbps (upload)

This is not 450mpbs, but speedtest is internet speed, not local router speed. It’s communicating to my home router plenty fast enough, and I’m getting the promised 100Mbps to the rest of the internet. So, this is a good thing. Ideally I’d be using the wired 1Gbps connectors on my motherboard, but it would not improve my internet speed at all, since that’s capped at 100mbps. I’m sure that if I’m communicating with other devices in my home network, I’ll get the fastest speeds possible, probably close to the advertised 450mbps.

One thing that was very interesting when I went to put this wireless card into the machine, I noticed tons of dust inside the case. So, I vacuumed it all out, figuring having clogged pores and fans can’t be good for the thermal profile of the machine.

One place that was particularly nasty was underneath the fan that sits atop the CPU. The fan sits atop a grill like heat sink, which in turns is thermally pasted to the CPU. Well, lifting that fan, there was at least 1/2″ of caked on dust/dirt sitting underneath the fan, and atop the grill.

Last summer I was having some problems with thermal shutdown when I was running some particularly heavy piece of code. Now I think I understand why. The CPU was just getting too hot, and it wasn’t really getting any real cool down assist from the active cooling system.

After doing all the cleanout, the machine is running quieter, as the fans aren’t trying to work so hard. That’s a nice relief.

In addition, I’m happy to not have to purchase a new desktop machine any time soon. In fact, this might be the last desktop machine I ever purchase. It may turn out that much smaller form factors, and gangs of them, will be my future instead. The only real benefit this desktop machine has is that I can place really beefy GPU cards into it. Well, given the ongoing price levels of those things, I’m not likely to do that any time soon either. So, there you go.

Well, new year, winter under way, PC tuned up, I’m ready to go!

Supporting the Next Industrial Evolution

A long time ago in a galaxy far far away… I was in business with my brother.  We did software, on odball platforms such as NeXT and BeOS, until I left for a job in the Pacific Northwest.  In the past couple of years, we have been having discussions about the economy, and where jobs have gone, and what can be done to create new jobs.

One of the thoughts that I’ve had on the matter is that certain types of jobs are rapidly disappearing, like heavy manufacturing.  A lot of these jobs are disappearing because of automation, or shifting labor rates, or shifting raw materials, or simply the needs of societies are changing.  Moving out of coal, more into natural gas, so less coal miners needed.

One of the things that’s been changing most recently is what I’ll call the democratization of manufacturing.  When you can purchase a “computer” for $250 (lastest Samsung ChromeBook), and you can purchase an assembled 3D printer for less than $1,000, you’re suddenly got in your hands the tools to begin micro manufacturing.  You can both create the appropriate design files, and with a reel of plastic, you can print off your own versions of things, locally.

This sort of scenario is being played out time and again in Maker Faires, NYTimes articles, Wall Street Journal and the like.  There is a lively community of makers in the world, and they’re all kind of chanting in the same way “We want to make our own stuff”.

One aspect of the maker community that is true for me is that my design skills are somewhat limited.  I’m not a SketchUp user, nor will I ever master SolidWorks, or anything remotely like that.  But, I know people who are experts in that stuff.  I’m an expert in certain kinds of software development.  Of late, I’ve been focused on things related to cryptography, identity, and network services.

What does this have to do with democratization?  Well, it occurs to me that the way the world is headed is more about mash ups of different skill sets, combined with a proliferation of relatively inexpensive manufacturing tools.  Once anyone on the street can get something designed to their specifications, and get that object manufactured within a reasonable amount of money, and a reasonable amount of time, we will be engaged in a new paradigm for manufacturing.

My brother’s latest venture (without me) is the pursuit of a first step along this path.  His company (Adamation) is a design firm, which is focused on the rapid development of customized products that can be rapidly manufactured.  By ‘rapid manufacture’, I mean 3D printed in most cases, at least to start.  They’ve just launched their new web site, which is a collaboration of artists from the gaming industry, along with visual artists, web designers and the like.

What they have to start is a series of completely original figurines.  There are 20 new figurines in total, which are bunched into 4 groups, with 5 characters in each group.  Each group has a unique story, and all the characters have individual bios.  You can go to the site, click on stuff, see some visuals, rotating animations, and the like, and ultimately make a purchase.

Now, I think this is really cool, not just because it’s my brother doing it, but because this is how I think goods should be.  Right now, each of the figurines comes in 3 different poses, but I know they have plans to allow the user to customize them further, creating a figurine that is as unique as the user desires.  This is customized manufacturing.  The figurines are printed ‘just in time’, so there’s no inventory.

I could gush on about these things, but I’ll just encourage you to go check out the site if you’re interested in such things.

This does bring up another thought in my mind though.  There are a lot of people thinking about how best to bring about this industrial revolution.  There are tons of little companies working on 3D printers.  Lots of people working 3D printable models, and lots of individuals dreaming about how to pull it all together.  The pieces are out there.

When I look at the current landscape, I see things like small electronics companies (SparkFun, SeeedStudio, AdaFruit), and their ability to crank out small scale electronics kit.  Then there’s the oh so popular for the moment Raspberry Pi, which brings some compute to the table.  Then there’s tons of people focused on creating cases for electronics, and these 3D printer people, who can make just about anything to pull stuff together.

Perhaps what’s missing is the spark, a true vision to help guide things.  Perhaps there’s some amount of funding that needs to be flung into some of the dimmer corners to bring out true innovations.  Some of this happens with KickStarter projects, but there’s probably more to be done.

I was happy to help my brother get started in this little venture by providing capital so that he could purchase the means of production.  Not a fortune, but more than I had in my piggy bank.  Perhaps what’s needed here is more people coming together collectively to fund the creation of new design firms, new design software, new micro manufacturers.  Just a thought.

At any rate, I’m happy my brother’s company has reached the stage where they can start talking about it openly in the world.  I’m sure it will be a success, or at the very least, very provocative.

These are a few of my favorite things

I’ve always dreamt of being an “industrialist”.  Unlimited capital, applied to projects of my choosing, transforming the world as we know it for the betterment of mankind…  and then life intervened.  But, thank goodness for KickStarter.  Since I discovered KickStarter a couple years back, I’ve had the pleasure of investing in a few projects.

I’ve backed simple things that I like, such as the because they improved on a couple of designs that I pioneered.  Then I invested in the InkShield, because I’m pretty sure this will help make 3D powder printing cheaper/easier some day.

I like this form of investing.  It allows me to lay out small amounts of money, less than $1,000 and get something in return, and possibly fund a new startup.

My latest funding effort is going to be the Teensy 3.0.  Then Teensy is a tiny little dev board that contains a 32 bit ARM Cortex-M4 CPU.  You can look at the kickstarter to get all the fabulous details.  What’s so special about this device?  Well, first of all, the Teensy can run ‘Arduino’ code, which is good for all those Arduino hackers out there in the world.  It can act as a USB client device, which means if you want to do automated testing, you could program this thing to be a mouse or keyboard, plug it into your machine, and have at it.  But wait, there’s more.  It’s got IO pins up the wazoo!  It’s got A/D and D/A, with at least 13-bit usable resolution.  You can overclock the thing to 96Mhz, which will make it useful for some things, like generating audio if you like.  It’s got timers, SPI, I2C, I2S, IR modulator, touch sensor inputs.  What more could you ask for in a $22 package!!  Forget the 8-bit Arduino at $35.  This is a much better starting point for many projects.

One of the cool things is you can program it in straight C, which is one of my platform choices.

I can imagine teaming this little thing with a Raspberry Pi for some awesome real world interaction.

One odd curiosity that’s in the device is a real authentic ethernet hardware address.  what’s the big deal?  Well, having an ethernet MAC address is necessary to do any sort of IP based networking.  Although this is not a cryptographically secure means of having an identity, it’s certainly a start.  Assuming the user has not changed the MAC address, you can use this little bit of information to identify the device, which might come in handy for some applications.

It might be very interesting to have some nice simple crypto routines readily available for this thing, possibly written in some nice tight assembly?

There are tons of other little boards that I’ve purchased and played with over the years.  These are a few of my most recent favorites.  It strikes me that with the capability and low price points that we’re seeing these days, my software efforts are going to increasingly be about connecting these small specialized modules together in interesting ways, rather than trying to build giant monolithic systems.