A little musing, computers

Some bits, pieces, and suspicions all pulled into one little package. Well, little for me. Maybe I can stay under a few thousand words.

Let’s start with the most jaw-dropping, and at the same time least able to defend. I think Apple will quit doing computers within the next handful of years.

That is, computers as computers. I think they’ll keep running various successors to the iPad, both stronger and weaker, for some time. But laptops and desktops are, well, I would not be surprised to see the big fruit absent from that area.

Assuming I’m right I think they’re smart in the short run and I think they’ll regret it in the long run.

Here’s the smart: at this time desktop purchases are declining. And it appears laptops may be following them. So why stay in a dying market? No, stay with tablets and phones and such, that’s where everyone is headed. Or so it seems.

Here’s the long-term regret: it leaves Microsoft dominant in a field that’s got some potential resurgence. Now let’s face it, Windows already dominates. It is, whether people like it or not, The Standard for desktop and laptop computers. Absolutely so for enterprise (businesses) and primarily so for PC type games.

So where’s the potential resurgence? Call it AR – Augmented Reality. Google Glasses are a start, but I think there’s another thing coming. And for this I have to come from the other side.

I’ve written on here before of Dragon Naturally Speaking, and don’t have much else to say. It’s extremely good right out of the box, and with continued use it’s beginning to have fewer typos and mis-cues than my fingers at the keyboard and mouse. But it’s still a pain for some of the things for which I use keyboard and mouse. Enter something else I’ve mentioned a time or two.

Leap Motion, and to a lesser extent Kinect. Leap Motion is the one that impresses me more, but Kinect may turn out to be VHS to its Betamax. Both are good at detecting motion, and Kinect is already better than it was a year ago. Leap can detect fingers moving independently – say, ‘touch typing’ a virtual board, and Kinect is probably not more than a year or two from doing it as well.

Muse with me for a moment as I go sideways. Graphic tablets become magnificent with pressure-sensitive pens. They’re good anyway, but that’s when they become great. But the main thing about the tablet is detecting exactly where the pointer is located. So, what if your motion detector can do that? What if Leap Motion (which claims a resolution of 0.01 millimeter) or Kinect (claiming a fraction of an inch) can gain a boost in accuracy using a specially designed stylus? Suddenly any tabletop or wall is sufficient for this graphic manipulation. Add, if you wish, the bonus of pressure sensitive pens and suddenly you can write on paper so it feels like you’re writing on paper – and it’s drawing on your screen.

Now what if that’s the pointer for your screen?

fwiw you could probably use any pen or pencil for just pointing. Or even the tip of your finger (though as always the point just isn’t fine enough for small detail. maybe. Assuming you don’t zoom in.)

Thus far, of course, I’ve meandered before. But there’s always been one small problem: the display. The screen. And the google glass just doesn’t make it – not quite.

Allow me to point to a VR game device: Oculus Rift. It’s good. It solves a lot of the problems of VR. But it’s not the first to look good, so why am I dragging it in? First, and less significant, it’s naturally stereoscopic. It gives you ’3d’ vision naturally, by presenting separate screens to each eye. Second, there are companies already working on binary cameras for it. Dual small cameras, each feeding to an independent screen, giving the wearer the ability to look past the box without taking it off. Finally, and most significant, is that unlike almost all its predecessors it’s suddenly got a major backer. It was purchased by Facebook.

I don’t like facebook for a number of reasons. But nobody can doubt its deep pockets nor its breadth of distribution. OR will be pretty much worthless to the company for social networking. As an AR (Augmented Reality) device that allows constant social networking, however, it’s very large. And that’s really just one use of it.

Let me give some scenarios.

Joe is a computer professional, someone who these days is chained to a desk with large powerful screen plus keyboard plus input devices (mouse and graphic). Shift slightly: give him LM and OR plus perhaps a stylus. Now he can work anywhere, so long as the OR can connect (wired right now, wireless in progress) to a desktop. Touching the screen (if that’s what he wants or needs to do) does not require reaching out. Just reaching to the virtual – or telling his system that the table at which he’s sitting is now the screen.

Mary is a mechanic. She can use her OR as a reading screen for the maintenance manual while looking at the machine to be repaired. And point to something with the line diagram activated and tell her system to order ‘this’.

The engineer looking at something with the CAD diagram of his intended enhancement as overlay – which he can tweak on the spot.

That’s augmented reality, and it’s coming in hundreds of different ways.

Yes, you can lift your tablet or smart phone or glance at the postage stamp on your google glass. Or you can ‘just look’ and ‘just point’.

Now, maybe, you can see why I think Microsoft was very smart in creating Windows 8. And if Apple is, indeed, exiting the computer field why I think they’re making a mistake.

An indirect brag

My daughter is an engineering student at Georgia Tech. So I get these things that don’t pop up on most folk’s views.

Recently, Tech Crunch went to Georgia Tech to look at its makers lab – I mean, Invention Studio. The whole makers’ lab concept is bright on its own, and putting it in an engineering college just plain makes sense too. I don’t know about most, but I do know that GT’s lab is very well-stocked with equipment. Not just 3d printers but lathes and waterjet cutters and a wide range of tools designed to make things.

So anyway, here is Tech Crunch’s video of that lab tour.

But wait, there’s more.

If you’re watching you’re going to see there are two or three women shown making things in the studio. One of those women is my daughter. She’s the one working on gears for the clock she’s making for the studio.

I said I’d brag a bit, but indirectly: her, not me. She’s done the work to instruct and supervise people using every machine in the lab. She’s not The Expert for any of them, but she’s safe and competent with all of them. And she makes these neat things; hopefully someday she’ll have time to put more of them up on her blog.

I’m a bit proud of her, and suspect that as time goes on I’ll be even prouder.

I want an app

I think the first grocery chain that does this is going to come out a winner. I also think it’s going to happen.

I want an app that lets me find things in the grocery store. That I input my shopping list and it gives me a map or location.

Bonus points for letting me scan the item for a price check (with price per ounce/piece ability) that lets me track what I put in the basket and how much I’m spending. Toss in an ingredients label that lets me check what the crap carrageenan (as an example) is while I’m in the store and I may make you the only store I use.

But given how many stores right now have POS systems that track inventory and that make their cap-and-shelf guides for stocking electronically I’d think this could be just a matter of doing it.

On worldbuilding, first steps

(crossposted to my gaming blog)

One of the peeves I keep as a pet is the frustration of world building. Over and over I see capitals and cities and such placed in locations that make no sense beyond “Isn’t this a neat spot.” Or worse, because it’s in the center. So I’m going to scratch the surface of world-building. Today: population centers.

Population centers developed because there was a reason to be there. The most common reason for this is trade. This is why just about every major city in the world in the 19th century or earlier was a port – either river or ocean.

I cannot tell you how often I see, and get annoyed by, rivers without cities and cities in the middle of nowhere with a river a couple days travel away.

Now there’s an interesting case on rivers that’s worth also noting, and that’s the ‘portage’ cities. Rivers have hazard zones – rapids and falls. If the water both above and below that hazard is navigable for any length by what’s considered ‘typical’ river transport then a village or perhaps a city will develop there. Provided, that is, the break is only a mile or two in length. Because distance matters.

We’re back to trade. Let’s go back to our ports. Now you’re going to draw a path from one to another that are not on the same coast or river. There are two rules.

1) The course is the easiest path.
2) Each segment should be one day’s travel. At that segment end make an evaluation as to what’s there.

One of the really annoying traits I see in fantasy maps (and to be honest think it’s a flaw in science fiction as well) is that roads are straight. They shouldn’t be, not between villages and towns and cities. They follow the easiest path – the one with the fewest and gentlest climbs, the one that has the best landmarks to avoid getting lost.

Will other, faster paths develop? Yes, but later. They’ll be put in for both trade and war, often because getting from point A to point E can be done in 3 days instead of 5 if a straight line is followed (and yes it means some of the older points whither). But as a rule they’re not going to get used by traders. Because traders don’t want to be caught outside.

So unless we’re on flat plains the road wiggles and curves such that the distance traveled is anywhere from half to 3/4 the straight-line crow’s-flight distance. And distance…

Distance is how far the traders can get in one day. Somewhere near that point an enterprising person will put a station – an inn, a feed store, a place for last minute “crap I forgot” items. Food suppliers will gradually gather round because there’s a market. Likewise suppliers of goods for those food suppliers.


Sometimes the land just won’t support all those people. Or they can’t be protected from the dangers – both natural and man-made. Sometimes it’s just too far from the major city.

But we’ve got enough now to start drawing our maps. So let’s play a little.

Draw a wobbly line down one side of a sheet of paper. I’m going to recommend using hex or graph paper unless you just like calipers or a ruler but for our first run any paper will do. This wobbly line is the coast.

Pick eight spots on the wobbly line and mark them. These are harbors – natural places for boats to come to shore and be a bit protected from the full force of the ocean.

Draw two ‘rivers’ from somewhere on the non-coast side of the paper to a harbor – each to a different non-adjacent. On one river put a mark about 1/3 of the way up. On another make one about 2/3 of the way up. These are riverports, and we’re going to give each a reason to exist.

At the city closest to the coast we’re going to give one river a second tributary. Run it to just short of the non-coast side, and for this exercise make sure it’s toward one of the other rivers but doesn’t close more than 2/3 the distance.

A second river is going to be a ‘portage’ port. To reflect this, make a mark about an inch long that is roughly perpendicular to the river and which runs through the city mark. This is an escarpment mark so make it plainly different (shading, color, pattern, whatever works for you.

If you later do other cities on rivers there are other reasons to exist did other cities give them a reason as well. In addition to escarpment and crossroad there’s crossing point. If you do this make a mark upriver (and down if you want) that’s at least an inch long that reminds you it is /hard/ to cross the river there. Why is it hard? You can play with it later, but it might be speed or width or rough terrain or a bit of a canyon or, well, that’s for you and your story later. But we’re not doing that for now.

Now we can build several roads here. There’s the coastal road, the one that’s a bit inland of the coast but roughly parallels it. There is a road that parallels each river from seaport to riverport. This exists because while barging downriver is easy, sailing upriver is difficult and often either skipped or uses a tow. And there’s a road that will connect riverports.

Let’s make that last. We’re going to assume for giggles that it takes a day for a trade train to move an inch overland – in a perfect world. We’re going to make that world imperfect.

Let’s start with the port that’s at a river junction. Estimate the point upstream that is closest to the other riverport and make a mark. This is the ideal stopping point, whether wide spot or village to be determined later. There might be terrain reasons not to use it, however.

Now here’s where we enter a little randomness. Take two dice, each different colors. One is ‘distance’ and the other is ‘accuracy’. You’re going to build your road alternating from each end.

For distance it’s high-low. One inch or 3/4 inch. This applies whether following the tributory or cutting across the land.

For accuracy it depends on whether we’re following the tributory or not. If following the river, you will stop once your segment crosses the ‘closest approach’ point (don’t turn there, go through it) OR if you roll doubles with the distance/accuracy dice. For cross country you go straight unless you roll a 1 or a 6. If you do that move your end-point for that segment 1/4 inch toward the coast or inland respectively.

If you’re using graph or hex paper add one more wiggle – your line must go from center to center. 45 degree diagonals are acceptable, ‘jumping’ from center here to center that’s one over and two up is not. Of course if you’re doing this you can ‘explain’ the short by wiggling the line a bit more so it’s always a full inch.

Now as I said you’re going to alternate from each side of this route. And as a result unless you’re terribly lucky you’re going to find the ends don’t meet. No problem, keep going. You’re going to end up with a ‘split’ route as the two routes join twice.

The merge points WILL develop small communities. There will be a reason for two routes – lake, dense forest, rougher hills, etc. And since they grew organically they’ll make sense.

Congratulations, you have a road. You have places for inns and communities. And most important it is NOT a straight line.

Now, there are a lot more roads and reasons and I’ll be covering those in a bit. I’m also going to help ‘organically’ grow nations – though they’re going to tend to be more fantasy than sf. But we’ve got a first road technique for world building and the primary reason cities and villages exist, and I’m getting tired of hearing the sound of my own typing.

So for now, have fun.

Why I still buy tahini

I like hummus. And halva, and a host of other things that tend to rely on tahini. And I like to cook. Capping this, tahini is EASY to make.

So why the heck do I still buy tahini?

Because I haven’t mastered getting it smooth.

OK, here’s the simple deal of tahini. It’s sesame seed butter, made the same way you make any nut butter. You grind the nuts/seeds, you add some oil till it’s the right consistency, then you put it on the shelf till you need/want it. Roasting the seeds/nuts a bit before grinding will bring out more flavor.

My problems is that regardless of whether it’s food processor or mortar and pestle, my tahini tastes a bit gritty. In some things, say a tahini and lemon juice based salad dressing, that’s fine. But in halva and hummus it’s, well, it’s not fine. The texture is irksome.

So if you’ve got a recommendation feel free to toss it in here. Because the flavor of homemade is miles above that of store-bought, and as a bonus it only costs about half as much.

For those who’ve not made it themselves, it’s the standard nut-butter method. Lightly roast some sesame seeds till they start smelling good (because they’re releasing some of the oil), and pull them out to cool. Add them to the grinder or food processor or whatever, and add 1/4 the nuts or seeds’ weight in oil – in the case of tahini, olive oil. Add it as you would to mayonnaise, just a drizzle at a time, and stop when it’s the consistency you want. If using a food processor or blender you’ll want to pause frequently to scrape the sides down.

Cream of Mushroom Soup

So let me start with the fact that this recipe Makes A Lot, and it is NOT LOW-FAT. I make it for three and we each eat two bowls – which on a health level is probably not that good. You’re free to try and reduce.

1 1/2 pounds fresh mushrooms (meatier types like button or crimini, though portabella would also work), sliced and/or chunked.
1/4 cup minced onion
1/2 cup (1 stick, 8 tablespoons) butter
1/4 cup flour
2 cups chicken broth
2 cups milk
2 cups heavy cream
1/4 teaspoon nutmeg
1/2 cup grated parmesan

Melt half the butter in a large soup pan over medium low heat.

Add the onions and sweat till softened.

Add the mushrooms, stir so all have some butter. Cover the soup pot and let the mushrooms cook down – it will take 10 to 15 minutes (check so they don’t burn.)

Add the broth, and turn the heat up to medium.

In a separate small pan melt the other half of the butter. Add the flour, and cook for 1-2 minutes to make a light roux. Stir roux into mushroom and broth mix. Bring the broth to a simmer.

Add the milk and bring the soup to a simmer, stirring frequently. (Tip – I pre-heat this and the cream for about 2 minutes in the microwave before adding to the soup. It speeds the heating.)

Add the nutmeg and parmesan and stir till incorporated.

Add the cream and heat, stirring frequently, till hot and the thickness desired. It won’t hurt if you take it to simmering but don’t go to a full boil. Note this is a heavy soup – it is a meal, not an appetizer.

I tend to serve this paired with a salad and some fresh-baked bread.


Peanut Butter Meringue Pie

I got this recipe when I was visiting some folk in Georgia. I’ve since seen it in restaurants in other southern states but it’s rather uncommon. Which means you can pull it out as ‘different but good’ at all sorts of places if that’s your goal, but I digress.

The young lady I got this from, well, I was with a co-worker who needed to to stop and see his grandmother. She gave us ‘just a bite to carry us over’, and when she discovered I knew skillet from saucepan she shared. So her version first, then I’ll get to the translation.

Bake a crust.
Take enough peanut butter to cover the bottom of the crust and cream it with double that of sugar, line the crust with that.
Make a three yolk custard to fill the pie.
Use the whites to make a meringue.

So before I expand that a bit, I’ll add the variation I created (though I doubt I’m first). Before you add the peanut butter, put down a lining of chocolate. I’ll note that in the two following guides (too sloppy to be recipes) as [optional].

Two guides. One for the non-cook, one for the cook.

For both guides it’s staged recipe in 4 [5] parts. Don’t let that throw you – except for prebaking the crust I’ve done the non-cook way in less than 10 minutes, and if you time everything right you can do the cooked way in under half an hour. (If I were practiced I think it could be one in 20 minutes.)

The non-cook guide.
Buy a crust from the store.
1/2 cup peanut butter
3/4 cup sugar (ideally confectioners but granular works)
[half a bag of chocolate chips]
small box instant vanilla pudding.
Small container instant whipped cream.

Bake the pie crust according to instructions. Usually that’s five to ten minutes in a hot (around 400F) oven.

[Add the chocolate chips to the hot crust. As soon as they melt carefully spread the chocolate to cover the bottom and as far up the side as it'll go. Yes, you can use the whole bag of chips if you want, but you may have to stick it in the oven for another 30 seconds or so to melt enough.]

Blend the peanut butter and sugar (the cooking snobs call it creaming) till it’s fully mixed – pretty much a paste by the time you’re done. Put it in the crust [on top of the chocolate] and spread it, pressing lightly to pack it, so it forms an inner lining for the pie crust.

Mix up the vanilla pudding and fill the pie.

Top with whipped cream.

That’s the fast version for people who “can’t” cook. As you’ll see it’s surprisingly similar to the ‘cooking’ version, except we’ll do a meringue and a little different pudding/custard.

1 cup flour
3 tablespoons butter
3 tablespoons shortening
3 tablespoons (cold) water
4 ounces sweetened chocolate
4 ounces heavy cream
{peanut butter}
1/2 cup peanut butter (your choice chunky or smooth)
3/4 cup sugar (confectioner’s preferred, granulated works)
1 1/2 cup milk
1 teaspoon vanilla extract OR 1 vanilla bean
3 tablespoons corn starch
3 egg yolks
3 egg whites
1/4 teaspoon cream of tartar
1/2 cup sugar (again confectioner or granulated)

Timing notes. This recipe will progress in order of assembly. However if you’re trying to assemble quickly, cook the custard while the crust is resting and baking. Cream the peanut butter, [make the ganache,] and whip the meringue next – depending on space and equipment (stand mixer, whisks, bowls) you can overlap their production, and pour everything in only a couple of minutes after the crust comes out of the oven and you’ve removed any pie weights. With that off to the recipe.

Using a heavy fork, work the butter and shortening into the flour until it’s a consistent but coarse texture. Add the water, stir just till it appears to be absorbed, cover, and put the bowl into the refrigerator for fifteen minutes. Pull it out, knead the dough gently until the moisture is evenly distributed, and roll between two sheets of waxed paper until it’s large enough to cover your pie pan.

Lay into the pie pan without stretching. Add pie weights if desired, and bake for 10-15 minutes in a hot (400-425F) oven until lightly browned.

{comments. The more you work the dough the more gluten it will develop and the tougher the crust will be. If you’re planning to eat the pie slices out of hand work that crust. Otherwise do as little as possible – and you should know that even rolling it do part of the kneading and working.

Also, the thing about pie weights is that opinions are mixed on their usefulness. The crust will puff due to steam, but it will compress again from filling. Some of us swear by them, the rest of us swear at them. Feel free to take a side as works best for you.}

[ the whole optional chocolate layer
For all the fanciness, a ganache is one of the easiest things I know to make. In this case I'm making it very chocolatey - the norm is 2 parts cream to 1 part chocolate.

Break up the chocolate into chips and put in a mixing bowl. Heat the cream to just short of boiling. Pour over the chocolate and stir till the chocolate is melted and the product is smooth and glossy. Pour the whole into the pie crust and tilt to cover all the sides, then let the rest settle into the bottom.

{At this point you have a choice. If you let the crust and ganache cool it'll be a clearly separate layer. If you go immediately to the next step you'll still have layers but there will be some mingling. Both ways are good, go with what you prefer.}]

{Peanut butter cream}

Cream 1/2 cup peanut butter and 3/4 sugar together. If you’re using confectioner’s sugar you’ll end up with something about the texture of what you find in peanut butter cups. If you use granulated sugar you’ll have that crystal texture.

Reserve about 2-3 tablespoons of the mix, and spread the rest inside the crust to form a full lining. It can be loose or smooth and lightly packed as you prefer.

Not a true custard but close.

Combine milk, sugar, cornstarch, and vanilla, and heat until just before it boils (it will thicken to a very soft pudding consistency). Lightly beat egg yolks in a separate container till they yellow, temper with the milk mixture, and add to the custard. Simmer for 3-5 minutes stirring constantly. Remove, and add immediately to the pie.

{comment time again, this time about tempering. If you know what that is, go on. For those who don’t, if you just add the egg yolks directly you end up with scrambled egg chunks in your pudding. What you do to prevent this is add a spoonful of hot mixture to the eggs, stir briskly till it’s all mixed in, and repeat. You’re done when you’ve about doubled what’s in the egg yolk container and it’s all really hot – not as hot as what’s in the pan, but definitely hot. Then you can stir this mixture back into the pudding and it won’t turn into little egg lumps.}

Put the egg whites and the cream of tartar into a mixing bowl and start whipping them. When they reach a heavy froth stage start adding sugar a teaspoon or so at a time, adding more when the previous spoon full is absorbed. Take the whites to a stiff peak. This means when you lift out the (stopped, please) blender the peaks form sharp points.

Spoon the mixture over the top of the pie, and spread so it covers the entire pie and crust. Shape if desired (pat the top with a spoon so it forms those pretty waves and spikes you see in all the fancy cookbooks.)

Put this in a hot (400-425) oven for 5-10 minutes until the meringue is lightly browned. Those waves and spikes will be browner – watch out for burning, but that medium to dark brown is caramelization of the sugar.

You can serve this hot but it works a lot better after being chilled for about an hour.