For the longest time I would sort of guess what my revenue would be each month.  I’d look at sales reports, guess a date range and hope that it was close. Somewhere in the middle of the month I’d start getting emails stating that financial reports are available per region.  Eventually all regions would be in and I’d finally have a good idea of what my next check would be.

But this month Apple is just now starting to release revenue reports for February, and I had a good idea of what my next check would be about a month ago.  Of course it’s not that complicated and I just needed to open my eyes and look for the answer.

I always assumed that the information should be easy to obtain long before the report comes out, but I wasn’t sure how to calculate it.  Apple pays something like 45-60 days after the end of the sales month.  So by the time revenue reports come out the sales for that pay month have all concluded weeks earlier.  The key was in knowing the date ranges for each pay period.  Of course all of these date ranges are published by Apple and were not hard to find.

Once I finally found that I realized that it’s not that hard to model what a check will be even a week or two before the end of the fiscal month – as sales are happening, and 30-45 days before the automatic deposit happens.  It’s nice to be able to plan ahead and it can also be a good kick in the pants when you realize that sales are down.  You still have time to try to make whatever changes or adjustments you can to improve that next check.

Disclaimer #1: of course, the actual amount of the check will vary based on foreign exchange rates.  So everything we’re talking about here is estimates anyway.

Disclaimer #2: and of course, you do have to be a developer to do any of this, you have to be making revenue, and you have to have an account with App Annie and your back sales date in their system.

I do use appannie.com for sales analytics, and that helps a lot.  They log in daily for me and get my sales data and present it in a nice format.  You could also get the data from apple (iTunes connect).  But App Annie is a little nicer on the eyes and has a few more tools for charting your data.  And it’s easy to view only a specific date range – which is key to projecting revenue.

What App Annie doesn’t have yet is sales revenue modeling.  I suppose maybe there are some legal vulnerabilities or reservations they might have about telling people what kind of revenue they might make.  But for everyone else in business and sales, modeling is standard practice. Since they don’t do it, this part is something that you have to do manually.  But still, it’s very easy to do.

First, Apple lists their fiscal year calendar on their iTunes connect website.  This has their fiscal quarter dates and each pay period.  I’m not going to link to it here, but just log in to your iTunes connect account, go to Payments and Financial Reports and look for the link called “Fiscal Calendar”.  It been there the whole time and I had just never thought about it.  But each Period listed there is the date range to use in App Annie to see a close estimate of your revenue for that month.  I’ve prebuilt all the links for 2012 here.  The ones marked 5 weeks have an extra week of revenue and theoretically should be 25% higher.  You have to log in to App Annie first and then click away to see how close the past ones were to reality:

Oct 2011 = Sep 25 – Oct 29 (5 weeks)
Nov 2011 = Oct 30 – Nov 26
Dec 2011 = Nov 27 – Dec 31 (5 weeks)
Jan 2012 = Jan 1 – Feb 4 (5 weeks)
Feb 2012 = Feb 5 – Mar 3
Mar 2012 = Mar 4 – Mar 31
April 2012 = April 1 – May 5 (5 weeks)
May 2012 = May 6 – Jun 2
Jun 2012 = Jun 3 – Jun 30
Jul 2012 = July 1 – Aug 4 (5 weeks)
Aug 2012 = Aug 5 – Sep 1
Sep 2012  = Sep 2 – Sep 29

For my sales these numbers often come in about a day too much.  So I recommend subtracting about a typical day’s worth of revenue to plan correctly.  You can see how close yours are and come up with your own rule of thumb.

Of course it’s not a time machine and any dates in the future show as zero sales.  I don’t see a point in modeling for a period that you’re not currently in.  To me that would be utter guesswork.  But for the period that you are currently in you’ll see some sales and then a flatline in the last part of the month.  For these ranges you can model the remaining days.  You just have to decide how optimistic to be.  Of course sales go up and down all the time and there’s no way to know for sure.  But there are a couple of ways to approach it.  First you have to simply count days – how many days have reported so far this month and how many are left.  Divide the revenue total for the selected period by the number of days with sales numbers and that will give you the average per day so far.  Then multiply that average by the number of days left in the period and add it to the current total.  You’ll have a “reasonable” estimate of revenue for that period.  I did this about a month ago and got a very close estimate of what my next check will be – about 45 days before the check will arrive.  Alternatively, if you’d rather be more conservative, a very low estimate would be obtained by taking the lowest single day sales for the month during the period and use that as your model for the rest of the month.  The rest of the month is always a ‘guestimate.’

I actually sent this as a feature request to app annie but so far they don’t have plans to add it.  I think it would be nice to at least have these date ranges auto fill in with a pull down menu.  But in the mean time just come here and click the links. Good luck!

Gillian works on her awana grand prix car with a high speed dremel.

Our car did really well.  We had the 2nd fastest car in her division.  But it was the first year they did the races at the church.  So I expect the competition will be much more fierce this year.

I LOVE this picture: Gillian proudly holds up her 2nd place trophy.

Her car’s total race time over the course of all (seven?) races was only 0.1672 seconds slower than the first place car.  That means if we could have gotten her car to be just .025 seconds faster per race she would have won.  That’s a tight margin.

Of course our approach (from a physics and speed standpoint) is a tightly held family secret – just like a super hero’s secret identity.  We’ll never tell!

However our design influence this year is something we can freely talk about. This year Gillian wants to do a Green Arrow car, which I think is an awesome idea.  So I decided to do a thread of blog posts here to go over our ideas, design process, show some of the work, and eventually (March 10th) tell how well we did.

Block of Wood

Last year we started with something that already looked like a car.  It was a good intro to pinewood as it didn’t require much cutting.  But it limited our ability to be creative.  This year we’re starting with a block of wood.  So there will be a lot more work to get it there.

Green Arrow

Gillian first was introduced to the Green Arrow through Batman: The Brave and The Bold.  He’s a great counterpart to Batman because, like Batman, he doesn’t have any real super powers.  He uses gadgets, technology and skill to beat the bad guys.  But it’s really all about his attitude and personality.  He’s very competitive and has a good playful banter with Batman.  I think the reason Gillian likes him better than Batman is because of his playful sassiness.  He’s one of the few characters on the show that doesn’t take any heat from Batman.  BTW: There are also rumors of a Green Arrow TV show coming to CW.  Of course we’re always on the cutting edge and are in to things before the popular public is.  By the time that show is out we’ll be wearing Blue Lantern/Flash crossover gear.

Design Option #1: Cartoon Reference

Green Arrow’s car in the show looks very much like an arrow that is green.

It’s really more of a rocket looking thing and it has fire coming out the back.  The Awana rules are strict with regard to overall size, but the shape can really be anything we want it to be.  We could do a very similar design with a cylindrical-ish body and fins around the wheels, but would have to modify the design some to fit the whole thing within the size limitations.  And cutting out that design would be more difficult. Of course, unlike the car in the show, the wheels on our car have to be in the exact positions to work on the pinewood track.

Here’s a little video to give you a taste of Green Arrow’s car.  Don’t blink or you’ll miss the car.  This thing is a rocket:

Design Option #2: Vintage Toy

Another design for a Green Arrow car that I found online is a vintage toy car, which basically already looks like a Pinewood Derby car:

I think this approach would be simpler because it’s a fairly simple shape with wheels on the outside just like a pinewood derby car.  I think we could pull off this design a lot easier – minus the windshield.

Design Option #3: Spence’s Green Arrow Car

Finally, I found one other kid in the world who’s done a green arrow pinewood car…

At first glance I liked his design.  I think his car looks fast.  I think he came in 4th.

But I prefer one of the other two designs.  Or possibly a mixture between the two. Tomorrow night after school I’ll see what Gillian thinks and we’ll start plotting our return!

The dome comes in several pieces.  There is an outer dome, which has been laser cut for “easier” removal of all the panels, and an inner dome which comes uncut and must be manually cut if desired.  Eventually the two domes are bonded together to create the beveled look that is so familiar.

The R2D2 dome right out of the box.

Both the inner and outer dome has been hand made by spinning a sheet of aluminum and curving it around a mold… Check out the youtube video to see how it’s done:

(If the video doesn’t load try: http://www.youtube.com/watch?v=U3ZWu9IURW4)

Luckily some gracious person has done this for us and provides us with something that is as close to a kit as you can get (well, at least for an aluminum dome).  And, like I said, before it gets to us the parts runner has that outer dome laser cut.  Without this step it would be VERY hard to get an accurate cut on all the dome panel pieces.  Here’s an animated gif of that process:

Dome Work

The first steps are on the outer dome.  When the laser cuts the dome there are pieces left behind that must be removed by hand.  And there are also semi-sharp metal burrs that need to be removed.

Gillian watches as I work on cleaning up the dome.

Here is the underside of the dome.  Numbers are written on all the panels so that we can tell where they came from.  I went a little overboard on labeling because I didn’t want to leave anything to chance.

Numbers are written on the inside of the dome so we can match the panels back to where they go after they are removed.

While waiting for time to do more work, I put the dome on a side table with a light underneath.  looks cool!  All the gaps where the blue light is coming through were the pieces that had to be removed.  You can also see the small metal tabs that are holding the panel pieces in place.

The dome before most panels are removed looks cool with a colored light under it.

One afternoon I decided to try to make some progress on the dome.  I used a hack saw to remove some of the larger tabs.  Usually you can remove the bottom tab this way which leaves only two side tabs.  These will friction apart by twisting the panel back and forth.

The dome with a few more panels removed.

A warning to any builders out there… aluminum is not a strong metal.  While working with the file on one side of the dome I didn’t notice that I was accidentally bending one of the thinner rings on the other side of the dome (the one above in the bottom of the picture), but it easily bent back into shape.  I don’t want to do that again.

The removed pie panel pieces and the bottom ring

The hack saw was a quick and easy way to remove the metal tabs.

I decided to get a decent file kit.  One of the files was curved on one side and flat on the other and worked really well to remove the metal burrs.

This file kit had a good mix of sizes and a handle.. about $26.

This is the file I used the most to remove all the metal burrs left behind from the laser cut.

The above file was perfect for working in the curved panels.  Next I’ll need to go over all the cuts with a finer grit sand paper.  I have some 200 emery cloth that I’m going to try.  I don’t want to mess up the spin lines if possible.

Potential Issues So Far

The outer dome should find a perfect fit on the inner dome so that it can be bonded (glued) together.  But you can see here that there is a gap.  Perhaps the outer dome needs to be forced down, but I’m not sure yet.  Perhaps when the inner dome panels are cut out it will fit together better.

The outer dome fit on the inner dome leaves a small gap that will need to be tightened up.

I also suspect it could be because the laser cutting wasn’t completely centered on the dome spin.  This picture below shows the top panel from the dome.  The hole was the center (circumcenter?) of the spin.  This hole is off center from the cut around it, signifying that the laser cutting wasn’t perfectly centered around this.  Meaning that the curvature of the dome wouldn’t be perfectly symmetrical, but rather is offset slightly.  My guess is that all the domes are like this to some degree.  It’s probably impossible to get it perfectly centered.  And I doubt it would be noticed in the finished product.  Still I wonder if this would make the domes not fit perfectly together.

The center ring shows that the laser cutting was slightly off center from the dome spin.

Finally.. while waiting for the next stage of work, I have the inner dome resting on the outer one.

The safest place for the now cut outer dome to sit is resting on the solid inner dome.

The To Do List

Next on the dome work adgenda:

• File and sand all the removed panel pieces to remove the burrs
• Do sanding and finishing work on the laser cut holes on the outer dome
• After finding the best fit, mark the holes on the inner dome and then mark the holes that need to be cut on the inner dome.
• Overcome my fear and cut the inner dome where needed.  And don’t mess that part up (or cut myself in the process).  This will be a big one.
• File/sand the holes where some other part need to fit in.. like the HP’s or the Logic Surrounds.
• Still need to buy a rockler bearing and an aluminum Radar Eye.
• Assemble the electronics for the dome lighting kit.

4S, that’s it? So, when can we expect iOS5 for iPad?

I responded on facebook and decided to also post a response here…

First off, iOS5 for iPad is releasing on October 12th… and it’s going to be awesome.  My parents both have iPads and I’m looking forward to the “PC Free” feature which will be an big advantage of them.  Never again having to plug either iPad in to the mac yet staying in sync with each other will be great.  Now when one of them buys a new Angry Birds game they both get it instantly.  And their data will be backed up automatically to iCloud every day when they plug in to wall to recharge. It’s more than a sales pitch, I’ve seen it work.

Sidebar: this Angry Birds example is real world.  I bought the app for them on one of the iPads and had to plug that one in to the mac to sync and then the plug the other one and sync it.  This isn’t something that they were going to figure out on their own.  It’s annoying.  But with iCloud it’s instantly a non-issue.

About yesterday’s announcement:

I have to admit that I was bummed when I first heard about there not being an iPhone 5. 9to5mac.com had a source and outed this info a few days ago.  Like many other people I didn’t want to believe it.  After 16 months we all hoped for an exterior redesign.

So what was this iPhone 5 we were teased?  I believe it is the iPhone5 that will be released year from now. I think it will be LTE and have that larger screen and the teardrop design. In previous releases we’ve heard rumors of features that weren’t released until the following year.  Apple plans their product roadmap years in advance, and very often some of those plans get leaked.  This leads to excitement over rumors and disappointment after the announcement.  If it weren’t for the rumors, this announcement would have been looked at very differently.

I think it is true that Apple sometimes holds back features so that you’ll want to buy another new iDevice again the next year. As an Apple fan boy it does get frustrating, like when we had to wait the extra year to get those front facing cameras. But the tactic works. People keep buying their stuff.

I actually don’t think they held much back this time. Keep this in mind: Apple historically only changes the outside design of the phone every 2 years, probably in part because people have 2 year contracts and only get to buy a new phone every other year. Case in point, I was talking to a different friend who didn’t get a new iPhone last year.  He is excited because he still has a 3GS. So an upgrade to an iPhone 4S will be new to him – and will actually be a very big upgrade.

Many people will be upgrading from a 3GS to a 4S this year.  Think about what that upgrade will be like, because for all of them it will be huge:

1. Much faster processor.  The 3GS has an ARM 11, the 4 has an A4 and the 4S has an A5.  So it’s two generations faster processor, or about 4-5 times faster processing.  Also in this category is much faster graphics processing.

2. Retina display.  Going from 3GS to 4 or 4S means twice the pixel density.

3. Front facing camera and facetime. The 3GS didn’t have this.

4. Way better camera.  The 3GS camera had no flash and was 3megapixels.  Also going from 480P to 1080P on video recording.

5. Much faster downloading over AT&T via the new dual antenna design.

It’s really big.  For anyone going from a 3GS to to a 4S it’s going to seem like a quantum leap. Every experience in the phone will feel a ton snappier.

Of course for me, it’s a single generation jump.  But after coming to terms with the same exterior, I’m very happy with the announcement overall.  I’m really looking forward to my upgrade.  I’ve seen the difference between the A4 and the A5 when I upgraded from an iPad to an iPad2. There will be a noticeable difference in the speed.  Plus there are several features on the list I’m looking forward to, including the better camera, the faster downloading, and Siri.

The 4S might look like a 4 but it’s very different inside. It’s twice as fast, has faster graphics than the tower computer I had before the one I have now, and their redesigned antenna system solves the antenna-gate problems and more by (theoretically) doubling the download speed (on AT&T only).

What if… they had introduced a different physical design but all the same innards and named it iPhone5 instead of iPhone4S?  People would have thought it was amazing. (Or they would have found something else to be bummed about.)  So the only bummer thing about it is that physically it looks the same as the 4?  How shallow are we?  I still think the 4 has a cool look – and almost everyone covers up the design with a case anyway. So what’s to be bummed about?

On Friday I’ll be clicking refresh over and over again until the order screen comes up, and preordering my iPhone 4S.

Of course her idea of a ‘robot’ is more the mid-50′s modern tin robot toy look.  So I ordered two R2 trays from thinkgeek.com and two ChillBots Ice Trays.  After a quick clean they were ready to fill and chill…

Filling them was a little tricky.  I used fridge-cold reverse-osmosis water and poured it slowly.  How much water to fill in is one question – and it’s difficult to control that.  Some of the pieces are smallish (the keys on the chillbots), and it’s a challenge to be precise on the pour amount.  As you can see in the photo, some of them were filled a little too much which sort of bulges the back.  I think less is probably better since water expands when freezing.

We checked them at an hour and they were still wet in the middle.  But by the next morning they were solid and frozen dry.

First the R2 ice…

The ‘cubes’ do have a good amount of detail.  In a drink I expect it will melt very quickly and lose the detail.  Especially the smaller ones.  But it’s still a decent novelty.  They also basically all have clouds in the middle.  There is a way to make ice without the cloud but it’s takes more attention.  I think since the ice freezes on the outside first air get trapped on the inside.  To eliminate the cloudiness you’d have to freeze it in layers by pouring only a little water in, freezing that, and then adding more, freezing that, etc.  I’m not sure that’s worth the effort – but maybe.

The big R2 was also cracked in the back.. probably from checking it too early.

And the smaller ones lose their detail quickly as they melt.  It’s hard to get them out of the trays fast enough to keep them frozen (without breaking them).

The Chillbots…

The chillbots are very cool looking ice pieces.  But the problem with them is that the hands and legs are very delicate and break off very easily.  To remove them from the trays in one piece I had to carefully press around the edges until I saw air get in behind the cube.  Still about half become amputees.  And of course they also have the cloud issue.

Overall I’m glad I bought them.  It will add a nice detail to the party.  My next challenge is how to store the ice.  I’ll be making ice all week and just dropping them in a bowl will probably result in a bunch of broken limbs and possibly having them all stick together. I’ll probably lay them out on a sheet in the freezer.  How to serve them will be another issue.  Leaving them out in an ice bucket won’t work.  It’s sort of an open house party. They’ll get melty and lose their detail and the novelty will get lost.  So it has to be something that is handled deliberately.

I’ll try making some chocolates or something with them later in the week and will update with those results.

Side note.  These arms are a lot smaller than I thought they would be.  But man that restraining bolt is way bigger than I thought it would be.  I have it on the fridge right now.

It’s no secret that Apple keeps a tight reign on what they allow developers to do with their hardware.  I don’t necessarily disagree with this.  But it can be annoying when you want to do something special – like use the computing power of an iPod Touch as the brain for a robot.  Because that requires some way to interface the iHardware with all the robot sensors and motor controllers.  But my understanding when iOS4 came out was that you could communicate with outside devices via the 30-pin dock connector.  Yeah, not exactly.

When I decided to jump in to this new hobby and custom build a full size replica R2D2, I immediately wanted to control the droid with hardware and software that is in my comfort zone.  The iDevice is small, fast, relatively cheap ($300), lightweight, has a great battery which should be able to last all day, can be quickly recharged or charged while in use by a conversion of the 24 volts on board, has built-in sensors like accelerometers, can play sounds, etc, and probably the most important reason to me, I already know how to code for it.  To me this was a slam dunk.

So I researched it and found that there is indeed an iOS framework to communicate to serial devices over the 30-pin.  But there was a major roadblock.  To do it you have to develop an “accessory”, and more importantly to gain access to the dock connector details, you have to join the MFi program and they’ll only allow REAL accessory companies to join.  How it works is that there is a chip in the cable that has a secret ID.  And in software you need that secret code, and the framework will only communicate with your device if you know that secret.  There are no generic cables.  And making a custom cable is expensive and the minimum order in like 20 or 30,000.  Oh and it requires an apple acceptance of your particular accessory and the application (use) for that connection.  Hobbyists were not going to be allowed to join.

There were two options.  First was that some resourceful people had made a custom cable that patches in to the serial port that is in every 30-pin dock connector.

http://resolvehax.blogspot.com/2010/…rial-port.html

Problem is that this solution requires a jailbreak – which I just don’t do.  So that was (mostly) out.  And another problem is that it might be dangerous, cause you could blow the device.  So again.. it’s out for me.

Second option was the SkyWire cable, which is an iPhone to serial adaptor/cable that is specifically made to control telescopes.  The cable is made by Redpark, and I contacted them to see if that cable would work for my application:

http://www.redpark.com/iphone.html
http://www.maclife.com/article/news/…es_ios_devices

The simple answer is no.  The problem is the secret code.  To use the SkyWire cable in my app I’d need to know the id.  I think I could figure it out with a little simple hacking.  There is a standard format for the id’s – so it would be easy to guess.  But if I’m going to share my solution with the rest of the club I’d have to disclose the code, and I can’t do that.  So that solution was out.

Back then I had a very nice talk with the makers of the SkyWire cable, Redpark.  They are the ones who explained all of these limitations to me.  But they were sympathetic to what I wanted to do.

The fix

Tonight I got an email from Redpark:

Hello Michael,

Apple policy has evolved — allowing us to release a new Redpark Serial Cable with SDK.   You may find this cable to be useful.Take a look at the Products section of our website (www.redpark.com).Mike

So now they sell an open cable and have an SDK to go with it: http://redpark.com/c2db9.html

Looking at the FAQ you learn that we still can’t develop apps that use this cable and sell them in the app store.  But this isn’t what I want to do anyway.  I just want to develop software for my own use and to share with a couple of other people in the builder’s club. And now this can be done without jailbreaking and without hacking the SkyWire cable. So I’m excited about this news.

The cable can be purchased here: http://redpark.myshopify.com/products/serial-cable

Question:

One of my friends asked me today (paraphrasing), “If you can’t sell the app in the store, how is it any use for anyone else in your club?”  Here was my response:

Personally, I’m not interested in selling these apps on the app store.  I actually couldn’t for several reasons, one being that I intend to use trademarked images in the app, like pictures of R2D2 and copywritten sounds.  Second that it wouldn’t be profitable – not to mention that I’m not allowed to make a profit.  Third is the reason to #2 which is that every implementation would be different and to make an app that works good for every r2d2 would mean too many settings, scripts, etc.  And while the club is like 7k people (signed up) the number of people who would ever use an ipod touch for that would be like single digits.

Also, just because I can’t sell them in the app store does not mean that I can’t distribute them.  I could compile and send out copies ad-hoc.  Apple doesn’t look at those apps or really even know or care what they do.  I can distribute 100 copies of each app, and I could just make a 2nd or 3rd app that does the same thing to get around that.  I can also publish code, which is what I’ll probably do for those who are interested.

It’s still good news that they lifted some of the restrictions.  It’s a first step.

The first thing to know about R2 building is that aluminum parts are built in runs. Some brave builder will find about 40 other people who need the same aluminum part he needs, and will organize a part run.  The idea is that no one makes any money doing the run, costs are kept as low as possible, and by doing 40 pieces we all get a lower price. I’ve looked in to doing one off CNC parts and it’s crazy expensive and shops don’t want to touch it.  But much of the cost is in the setup, testing, and then turning the machine on to do a bunch of pieces.  When doing only one the overhead is too much.  All the parts are manufactured to a club spec so they should all fit together.  And these days people are doing “diet” parts where they drill away parts of the metal that aren’t seen in order to lower the weight as compared to producing a solid piece of aluminum.

All of this means that parts are sometimes rare because they don’t run very often, maybe only once a year or two. Also parts are not kept on the shelf. They are run in ordered quantities as needed, and then not run again until there is enough demand to justify another run of 40 pieces.

When I first got the bug to build an R2D2 I panicked because I didn’t want to miss out on any current runs, and then not be able to get that part for a year or more.  So I started ordering all of these obscure parts for my R2 build.  But I quickly had to put the brakes on because I was racking up quite a bill and didn’t have any parts that would fit together.  So I decided to back up and think it through.  I decided that certain parts were critical to make R2 look like R2 and some of these parts were also the most rare, meaning that they hadn’t run for a long time.  These are ones I’d get in on at the first opportunity.  Here is the list:

• Dome
• Holo Projectors (HP’s)
• Radar Eye
• Utility Arms
• Skins
• And I need a frame to hang it all on

The legs can be made to order – so I’ll wait until I need them and can afford them.  Many of the body detail parts I’ll get when I get or I get in resin to hold me over.  And all of the internals, the electronics, the motors, servos, etc.. all are either readily available or I’ll have to custom design and build.

So far I’ve ordered the Dome, HP’s, and Utility Arms.  I’m waiting for Skins and Radar Eyes to run again, and I’m going to custom build a frame.  And to give you an idea of the waiting game, the Dome was ordered early in July and I’m slated to get it mid December.  That’s about 5 to 5 1/2 months.  But they’re hand spun, laser cut and are a piece of art when finished.  So waiting is the name of the game.

Well… my utility arms came in today….

The problem is that I got 2 lower arms and no upper arm.  The holes on the bottoms of them are the ‘diet’ part where they were drilled out to save weight.  Shouldn’t be easy to notice unless you’re laying underneath R2 while his arms are out.  But I can’t have those holes pointing up.  Had they not been diet maybe it wouldn’t matter.

I’ve messaged the part runner to see what he can do.  Maybe someone else was accidentally shipped two uppers and hasn’t noticed yet.  The problem is that I had him anodize mine and he was all sold out, so I hope there’s an upper arm waiting for me somewhere or I’ll be waiting a year for the next run!

One of the things that excites me about it is the engineering challenges. These things are not kits. Far from it. There are parts that are sort of available every now and then. There are plans and blueprints for different parts. There’s the builder’s club forum and a yahoo group to ask questions and co-op on part runs.  And there are 100′s of other blogs to follow fellow builders, see how they solved problems and learn from their mistakes. But depending on what you want the droid to be able to do, there will be an array of problems to solve.

I’m actually not one of these people who gets in to costuming.  Last Halloween I considered getting a real Stormtrooper costume.  But at about $1200 I decided against. But at some point before that I really wanted life sized R2 and C3PO statues. I found some really good licensed ones for about 7 to 10 grand each, and they were all sold out. Well for just slightly more you can build your own and it’s not a statue.  It’s a fully operational metal droid.

But with a price tag in the neighborhood of 10 grand, how do you justify the new hobby to your wife? Well there are worse things a husband could be doing… much worse things. And I suppose you just have to bring in enough bacon that it doesn’t matter. And then there’s the fact that it can take years to complete and therefore years to pay for. Although I am a ‘want it now’ sort of person – so I’ll be trying to move things along as fast as I can.

So I’ll be sharing some of my journey in this blog, starting with the first part that I just got in today.  And as I get time I’ll add resources, etc.  Note that in general this will not be a how to blog.  Others have done that quite well, but I don’t have the patience to take all the necessary process photos and discuss each step.  This is an experience blog and a whatever-I-feel-like-typing-today blog.

But today with multi-code processors and multi-threaded code, I have to admit that half the time I don’t really know much about what’s really happening in the compiled code from a program flow standpoint.  Does it sometimes run down two threads of code at the same time?  I sort of assume that it does at times.  But I don’t really know.

For example, early on I found that I needed to ‘put off’ initializing some of my game variables rather than doing it inside the loadView because otherwise the OS would think my app was locked up and crash it out.  The loadView needed to end as quickly as possible so the game title screen would flip off.  So I used:

[self performSelector:@selector(initGame) withObject:nil
      afterDelay:.1];

This moved the initialization out of the loadView so loadView could return quicker.  But almost immediately (.1 seconds later) run my initialization method.

But again.. I really don’t know what’s happening here. I assume that there is a main control loop that starts a timer and then executes the initGame selector when the timer runs out.  But at the same time I do have an animation timer that triggers a drawView every 30th of a second.  So.. what happens when one timer triggers before another one is completed and has returned?  Does one method cease to run until the other one returns, does one timer have to wait in queue until the other one returns, or do both methods run along side each other? I know that I’ve seen my animations stop mid stream when something else happens, and then continue after.  So maybe there is no multi-threading going on here.  Maybe some know it all can leave a comment and tell me the answer.

But narrowing in on the point of THIS post.. in my table view I need to place certain functionality inside the iphoneViewController so that it will have access to properties (variables) of my game method… but I need to call it from inside one of the tableView controllers when someone taps the appropriate tableview cell.  The way I’ve been handling this so far has been via messages and notifications.  Once I find something that works I exploit it as much as possible.  But I want to know what’s really going on so I don’t have some accidental conflict.

So to my experiment… Using NSLog is an interesting way to track the program flow and see what’s going on. I’m about to do that now and publish the results.

In my iPhoneViewController I set up the following notification:

[[NSNotificationCenter defaultCenter] addObserver:self
   selector:@selector(testControl) name:@"testControl" object:nil];

Then:

-(void)testControl {
    NSLog(@"inside test control");
}

Then in my table view, inside didSelectRowAtIndexPath I have the following:

NSLog(@"test - sending a message to testControl");
NSNotification* notification = [NSNotification
     notificationWithName:@"testControl" object:self];
[[NSNotificationCenter defaultCenter] postNotification:notification];
NSLog(@"test - after the message is sent");

The point is to see if the ‘after message’ prints out before or after the ‘inside test control’. Is it a linear path, just like I called a subroutine (in old coding terms), or is it a sequential series of events?  Here are the NSLog results:

2011-05-25 04:40:37.809 X[15647:207] test – sending a message to testControl
2011-05-25 04:40:37.809 X[15647:207] inside test control
2011-05-25 04:40:37.810 X[15647:207] test – after the message is sent

In this test it appears to be a sequential series of events in the flow.  At least it looks like the message was picked up immediately, even before the clock had time to tick a 100th of a second, and the testControl method ran, and then after that the program flow returned. But with only that one NSLog in testControl maybe the timing just worked out that way. To do a better test I added a long delay in testControl via counting to 10000 in a for loop.

-(void)testControl {
   NSLog(@"inside test control");
   for (int x=0; x<10000; x++) {
     // do nothing.. blow some time here freeze the app a bit
   }
   NSLog(@"exiting test control");
}

Ok so counting to 10000 happens in a blink.  I had to up that to x<100000000 in order to have a barely noticeable delay.. and here are those results:

2011-05-25 04:47:54.642 X[15782:207] test – sending a message to testControl
2011-05-25 04:47:54.642 X[15782:207] inside test control
2011-05-25 04:47:54.898 X[15782:207] exiting test control
2011-05-25 04:47:54.898 X[15782:207] test – after the message is sent

I think this is a more reliable test.  Sending the message did stop the program flow in that method, run all of the code in testControl and then return and continue.  This is a successful test for me because it helped clear the fog a little bit.

Another note is that the [15782:207] does apparently indicate the process and thread. Above those lines in the output was:

[Switching to process 15782 thread 0x207]

Therefore we know that everything happening in those NSLogs was all sequential code in one program flow.  The postNotification message sender had to wait to see if anyone received the message and run that code before returning and continuing the flow.