The Developer's Notebook

How fast was I passed?

2012-08-31T07:00:00.000-04:00

Woah! A car just flew by you on the highway and you've just about had it with the reckless driving! That's what happened to me. Being the big nerd in the car I thought, "I wonder if I could guess how fast the car passed me at?" As it turns out...there is a way for you to estimate this yourself...no technology required!

Off the cuff, I thought to myself...well I need to figure out how fast a car travels per second in feet. Using known quantities I did the following rough calculation:

There's 5280 feet in a mile. My car is probably 15 feet long. If someone passes me, how long will it take their car to pass me? Well, 1 mile per hour would be 5280 feet per hour. So if I do the math I get 1.4ft/sec [ (5280ft/hr) * (1hr/60min) * (1min/60sec) ]. For simplicity sake, I round that to 1ft/second = 1mile/hour.

So if someone passes you in the car, the simple equation to figure out how much faster they're going than you is to divide the length of your car by the time it took them to pass you, for instance:

A car that took 3 seconds to pass you would be [ 15ft / 3sec ] 5mph.

That's because our fancy equation boils down to speed = distance / time. Our distance is 15ft, our time is 3seconds. So we get 5 ft / second. Jumping back to our rough calculation from earlier, we made an estimate that 1ft/second equals 1mile/hour. Therefore 5ft / second is 5 mph.

If we normalize it to the actual value [ 5ft/1.4 ] we get 3.5mph.

So now, who needs a radar gun? Drive safe this Labor Day weekend!

Are hybrid cars worth the cost?

2012-08-09T22:49:00.000-04:00

I thought it'd be interesting to calculate whether or not purchasing a hybrid vehicle was cost effective yet. I'm personally not in the market, but if I was and the cost of ownership was the primary focus, this spreadsheet can give you a good indication of where you should put your money.

The Mountain Lion Memory Footprint

2012-07-30T08:30:00.000-04:00

So I was looking the other day for information regarding "How much memory does OSX Mountain Lion use compared to Lion?"

I was struggling to find this information regarding the memory footprint when I finally caved in to get this information myself. I threw down the $20 and here's the results:

OSX Lion after running the 'purge' command

OSX Mountain Lion after running the 'purge' command

The overall results concluded that the memory use is slightly higher on Mountain Lion, but not by much. From what I could tell the larger culprits included:

mds - Increase of ~40MB
dock - Increase of ~ 25MB

The kernel_task had a slight reduction of memory consumption. Also, if you compare the size of the process list, it seems that there's more processes running in OSX Mountain Lion than in Lion. This probably accounts for the remainder of the ~200MB increase.

I know this is hardly scientific as far as details go, but hopefully this gives some of you an idea of what to expect when upgrading your system if you're already capped as far as memory goes.

This was captured using a mid-2007 iMac with 4GB of ram.

What applications are compatible with Mountain Lion?

2012-07-26T06:47:00.000-04:00

You might be holding off with upgrading to Apple's new operating system "Mountain Lion" for fear that some of your applications won't be working when you upgrade, especially considering my recent post on Parallels 6 not working in Mountain Lion.

Have no fear, check out this massive list put together by RoaringApps with your compatibility matrix for "Lion" and "Mountain Lion" that's easy to search through for those apps in question.

Parallels 6 no longer works in Mountain Lion!

2012-07-25T22:18:00.003-04:00

If you're one of those Mac users like me who has OS X running Parallels 6 with a VM of Fedora or CentOS for linux development .... hold off!

Parallels 6 will no longer be supported by Mountain Lion. Instead, owners of the product will be required to purchase the upgrade to Parallels 7 for $50.

It doesn't appear that they plan to work on a patch or anything for it either. If you've been following Twitter or Facebook, there's a fair amount of Parallels folks a bit PO'ed about the incompatibility.

Screenshot from the Parallels Facebook Page

Changing the ring duration on your iPhone on AT&T

2012-07-25T08:00:00.000-04:00

I was looking for information on how to change the ring duration on my iPhone so that I could set up Google Voice to forward to my iPhone without doing Call Screening on incoming calls. For those that don't know, call screening is a feature on Google Voice that tells you who is calling and requires you to press "1" before you are connected to the incoming caller. For hands free operation, it's not the most optimal solution.

So on AT&T the default ring duration for the iPhone is about 25 seconds. When calls go unanswered, the shorter ring duration on my AT&T iPhone ends up answering the call first and routing the call to my iPhone's voicemail instead of Google Voicemail. Google recommends setting your iPhone to use Google Voice for voicemail, but then you lose the ability to use visual voicemail and distinguish which device the calls were received from.

The AT&T site says to call customer service to have this duration changed. However, if you're a power user like me and don't want to wait on the phone for customer service to do this, you can do it yourself.

To increase your ring duration on an AT&T iPhone, do the following (and I've tested this, because I'm using it now =) ):

On your phone, dial *#61# and click "Call"
A bunch of information will be displayed, one regarding your voicemail forwarding. Copy the number listed there on a piece of paper or to a notepad. Include the leading 1 as well. We'll represent this number as 1NNNNNNNNNN.
Now on your iPhone, dial **61*1NNNNNNNNNN*11*30#.
Hit Call (note some other posts reference the +. You don't actually dial that or substitute it with a 0. Just ignore it).
Step 4 should have given you a message saying it was successful. You can repeat step 1 to verify that the voicemail number is still correct but it will not show you the duration. To test this duration will now work for you in google voice, call from a number not associated with your google voice account. (Note: Calling from a number associated with the account will do a consultation call to your google voice mailbox).

Again, note that this only works for AT&T.

Google "askew" today!

2012-07-24T08:53:00.002-04:00

Amelia Earhart's 115th Birthday Google Logo

Google added a little "easter egg" today! Try googling the word "askew". I won't spoil it, but it may only be available today only.

Providing I remember, I'll update the post later today with a screen-shot of the results.

Likewise, today is Amelia Earhart's 115th birthday!

And if you're feeling adventurous, try the googling "do a barrel roll"!

Update:

Here's the screenshot of the askew google search!

Minecraft Sells 3 Million Copies on Xbox LIVE Arcade

2012-07-23T07:10:00.001-04:00

Minecraft Sells 3 Million Copies on Xbox LIVE Arcade via PlanetXbox360:

Developer Mojang is reporting that Minecraft for Xbox 360 has sold over 3,000,000 copies since the game debuted on Xbox LIVE Arcade in May as part of the Arcade Next promotion.

It's amazing that a game like this has sold so many copies, both on XBLA and originally in the PC/Mac Market (since it's java based). You can meet the team of developers behind the game here on the official Mojang site!

Is Diablo III a flop, or just a slow starter?

2012-07-21T08:00:00.000-04:00

Since the release of Diablo III, I'm not sure what to make of it. Part of me is immensely disappointed at the fact that:

After leveling my wizard to level 60, grinding for loot/gold and searching the auction house is the only way for me to really progress further in the franchise on inferno difficulty.
Carrying three other characters halfway to 60 while juggling a mixed stash with no way to automatically sort or organize it has become more of a chore.
The above two aspects are now completely boring me. Programming and going to work is actually way more fun!

The one good thing I can say is that crawling slowly through hardcore with two characters is actually somewhat rewarding, considering the fact that I can pretty much sell anything in the auction house for cheap that other people can use to restock the successors of their deceased characters!

On the flip side, I can't sell anything in the RMAH (Real Money Auction House). It's funny, but my goal wasn't even to make money when they finally opened the auction house. My goal was to make enough on my blizzard account so that I could buy a different blizzard game to play, like Starcraft II. As you can see below, that hasn't panned out so well. =)

A highlight of my RMAH Success

Now the gold auction house has been working pretty well in hardcore mode. Given that people are always looking for new items for cheap to bootstrap their reborn characters, it's often better to sell any kind of junk loot for cheap in the AH to obtain a marginally better profit than you would from the in-game vendors.

Hardcore Auction House

The time for these items to sell typically won't take long if you put a low buyout price, mine typically range anywhere from 200 gold to 500 gold for some of the garbage, but they're at least useful items for low level characters. If priced in this range, they can typically sell in less than an hour or two.

But I digress, even with some of these rewarding aspects, it still isn't enough maintain your attention span for that long. In my case, however, I've been Diablo III free for over a week now so we'll see if I maintain my distance and wait until the fall to start playing NHL 13, which looks pretty awesome!

I may turn around and start playing again if things get better. Mike Morhaime, a blizzard rep (blue), posted in the forums the other day regarding the current state of affairs. In it he states the following:

Dear Diablo Players,

Now that Diablo III has been out for two months, I would like to take a step back and discuss the launch, the feedback from the community, and talk about what the future holds for the game.

We’ve had an enormous amount of feedback since the launch. I speak for everyone at Blizzard when I say that we appreciate the passion and affection many of you have for the Diablo series. We truly believe “Every Voice Matters,” so I want to thank you for sharing your experiences in an effort to help us improve the game. Your support makes it possible for us to continue doing what we do, and we never take that for granted.

The launch week of Diablo III was memorable for many reasons -- some positive, and some not so positive. We were thrilled that Diablo III had the biggest PC-game launch ever, surpassing the lifetime sales of Cataclysm (the previous record holder for biggest PC-game launch) in a matter of weeks. We’ve been floored by the response.

However, the launch had many challenges as well. It has always been difficult to forecast how many players we will have. With World of Warcraft, it was a challenge to handle the immediate demand when we launched back in 2004, and that was just in North America. We eventually expanded to other regions and reached nearly 5 million players by the end of the first year, and there were a lot of growing pains with that. However, we’ve never gone from 0 to more than 6 million players across multiple continents within a few days with a brand-new game. For Diablo III, we looked at historical sales for Blizzard games and other top-selling PC games and watched preorder numbers. We even upped our estimates to ensure we had additional capacity, or so we thought. In the end, it just wasn’t enough, and that is something we will work hard to conquer for future releases.

In response to the immediate and overwhelming demand for the game, the team worked around the clock to support all regions, increase capacity, ship additional hardware to our datacenters, and troubleshoot and fix bugs as they sprang up. While things have by and large been running smoothly for several weeks now, various game-related issues have come up that we have either already responded to or are continuing to investigate (such as the latency issue some of you are experiencing) and make adjustments for. Rather than address every subject individually, I’ll just say that even as we work to address or resolve current issues, it’s always possible that further issues will crop up. We hope that our actions in the past have demonstrated that above all else, we’re committed to delivering an awesome game experience, and we hope you’ll have faith that we will continue to keep that commitment and respond to any new or outstanding issues quickly.

We are not satisfied with breaking launch records; we want people to continue playing and enjoying Blizzard games for a very long time. The Diablo III team has made an epic, entertaining, and beautiful gaming experience. That being said, we know that it isn’t perfect. Our teams are working hard to improve the game balance, build on our design, and listen to what players are saying to make it the best game it can be.

You’ve seen some of that work already in patch 1.0.3, and you’ll see additional improvements with patch 1.0.4. On the game balance front, this update will contain changes designed to further deliver on the team’s goal of promoting “build diversity,” with buffs to many rarely used, underpowered class abilities. Another topic we’ve seen actively discussed is the fact that better, more distinct Legendary items are needed. We agree. Patch 1.0.4 will also include new and improved Legendary items that are more interesting, more powerful, and more epic in ways you probably won’t be expecting.

We’re also working on a number of interface updates, including social improvements that will allow players to more easily view their friends’ achievements, more quickly join games, and more efficiently communicate with each other. In addition, we’ll be making updates to the auction house in the future to provide players with better information through tooltips and notices, offer improved search functionality, and more.

Regarding the real-money auction house, our primary goal for including this in the game was to provide convenience and peace of mind for those players who might otherwise turn to third-party services to buy items. Black market trading sites can put accounts at risk and create many customer service challenges. We felt that the players themselves also deserved the opportunity to benefit from the extra loot they found, as opposed to having all of the benefit go to the black market/illegal trading organizations. We know the auction house isn’t perfect, but with your help and feedback, we’ll be able to continue making it a better experience for those who choose to use it. On the flipside, we are also committed to ensuring you have a great experience with Diablo III without feeling like the auction house is mandatory, which was never our intention. Thank you for all the feedback about that.

One other common topic we’ve seen in the forums is the always-connected experience, and the perception that the online requirement is nothing but an ineffective form of copy protection that has already been cracked. While we’ve never said that this requirement guarantees that there will be no cheating or game cracks, it does help us battle those problems (we have not found any fully functional cracks). More important to us is that the online requirement is critical for the long-term integrity of the game experience. I fully understand the desire to play Diablo III offline; however, Diablo III was designed from the beginning to be an online game that can be enjoyed with friends, and the always-online requirement is the best way for us to support that design. The effectiveness of the online elements -- including the friends list and cross-game communication; co-op matchmaking; persistent characters that you can use by yourself, with others, and in PvP; and some of our customer support, service, and security components -- is tied directly to the online nature of the game. These and other online-enabled features are essential to our design for Diablo III. That said, there are still improvements we believe we can make to expand the online experience and make co-op play even more rewarding, and this will remain one of our priorities moving forward. Overall, while there are some downsides to the online-only approach, I still believe this was the best long-term decision for the game.

I know many of you are also looking forward to patch 1.1, our PvP update, which will provide new experiences and give you a whole new way to apply the skills you’ve picked up while battling demons. This patch will also build on the social and auction house changes I mentioned above, and the team will continue to fix bugs and further tune game balance as well.

We’re also working on a gameplay system that will provide players who have max-level, high-powered characters new goals to strive for as an alternative to the “item hunt.” We’re not ready to get into specifics just yet, but I can say that we’re actively taking your feedback into account as we plan out the future of the game.

As always, we appreciate your candor and passion. Your constructive feedback and thoughts are valuable -- they will continue to help us be a better company. I just want to reiterate that while we can’t claim to have ever shipped a perfect game, we are committed to supporting our games relentlessly and making improvements where we can. Thank you for your support.

Sincerely,

Mike Morhaime

So at least from a developer perspective, I'm optimistic that things will change for the better and I really would like to see that endgame content extended and less of a dependency on the auction house to get through the game.

In fact, I can almost hear and picture the managers sitting down for their scrum meetings as they draw out their planning and the developers being a little bit dissatisfied that their creations are less than satisfactory to the vocal community at large. As with all development, it's important to stay focused and gear your changes and recommendations in a constructive and productive manner.

Now to reiterate on the topic, is Diablo III a flop? I'd venture to say not at all. The hype and excitement that built around this game was insane, which I think contributed to the disappointment to the community at large. People were expecting more and probably something along the lines of Diablo II with an updated engine.

Is it a slow starter? Probably. I think in some sense, this slow start has disenchanted a lot of people. Don't get me wrong, I think those people will likely keep monitoring the Diablo III forums for some time, but I don't think the investment of time will be even near close to where it was at launch.

Despite this, the willingness of Blizzard to continue to improve and finance the improvement of the game post launch is the one thing that over the years has kept the name untarnished. I suppose you could call it damage control, but likely the bulk of sales of people interested in this game has already been made. How Diablo III eventually pans out in terms of longevity and replay-ability will determine whether or not consumers will even consider purchasing another game or expansion from this franchise.

I can't speak for World of Warcraft, but at least back when I played in 2005/2006 it was a pretty good game. I only wish the price came down to like $6/month so I could have kept it in the budget to continue to play, but sometimes you just have to make the sacrifice!

ChargeCard: An iPhone charger in your wallet

2012-07-20T22:26:00.000-04:00

I came across this neat little device in my reader feed and felt it deserved a little extra publicity for the innovation factor! Enjoy!

ChargeCard: An iPhone charger in your wallet:

A lot of innovative ideas have popped up around the idea of always having an iPhone charging cable at your fingertips, but the ChargeCard -- a Kickstarter project that is a little over quarter-funded with 38 days to go -- is one of the most clever I've ever seen.

The inventors of the ChargeCard came up with the idea of a device charger cable that has a USB connector on one end and a standard 30-pin Dock connector on the other end. But the unique part is what's between those two ends -- a flat, credit-card structure with a fold-out flexible "cable".
The entire thing is designed to fold flat and take up no more space than a standard credit card. It's perfect for those situations where you forgot a charge cable, but you do have access to a USB port to plug one into. Open your wallet, take out the ChargeCard, flip open the "cable" and plug it in, then plug the other end into your iPhone. Done.

If you want a chance at getting a ChargeCard, you can pledge US$18 or more for various perks. Check out the video below to see how the inventors came up with the idea.

ChargeCard: An iPhone charger in your wallet originally appeared on TUAW - The Unofficial Apple Weblog on Thu, 19 Jul 2012 16:00:00 EST. Please see our terms for use of feeds.

Source | Permalink | Email this | Comments

Pinterest-ing Concept

2012-07-20T06:00:00.000-04:00

So lately I've been hearing a lot of information about this new application/site called pinterest. I honestly had no idea what it was, but after hearing about it on the radio and seeing posts from various friends on facebook, I decided to give it a whirl!

I begin the experiment on July 15th, 2012 by navigating myself to the Pinterest home page.

It's a little confusing what to make of it at first. At first glance it simply looks like a photo board where people post cool pictures that other people either "like", "comment", or "repin". At this point, I'm still not quite sure what to make of it, but I go ahead and click on the link at the top to "request an invite".

A few minutes later, I have my first e-mail from pinterest:

Well, that was easy! But why the delay? To explain it to non-developers, the system of implementing a delay to join the site is to "rate limit" the frequency at which people are allowed to join. This prevents the system from being brought to its knees by the "viral" nature new, and popular, sites tend to have with the flurry of activity generated thereafter.

It wasn't long before I received my invitation the following day -- July 16th, 2012:

Wow, that was quick! Must be because I'm late to the bandwagon. Apparently everyone else already got invited and I'm one of the 20 people in the "cron" job today that were sent an e-mail. There's nothing like being the last one invited to a party, let me tell ya! =)

So now that I'm in, I was presented with a menu asking me what my interests were? Naturally I started selecting various things: Photography, Humor, Technology, Food & Drink, etc.

Of course I didn't know what I was doing, but in reality what pinterest was actually doing for me was selecting people to "follow" based on the category of their boards. Why, you ask?

When you follow someone on pinterest, the images displayed on the main page are those of the people you follow. Pinterest gives you the capability of searching deeper into categories, but by following friends that have similer "pinterests", the results displayed are more relevant to things you care about. Each image/pin can be clicked on and can, and typically does, link to the originating site from which that image was displayed from.

The stuff you pin is organized onto boards. Friends and strangers can follow your whole feed or just certain boards that you have, making it easy to gear your pinterest feed with the things you want to see more of, allowing you to ignore the rest.

So yah, that's pretty much pinterest in a nutshell. Just a way to share images, recipes, and links to websites in a creative way.

Verizon vs AT&T Shared Data Plans Comparison

2012-07-18T21:24:00.000-04:00

As a developer, you can't help but reading through technology articles, sometimes wondering, "What were they thinking?" This month, AT&T and Verizon released information regarding their shared data plans and yet, there's no real significant difference between the two. In fact, for the consumer you are worse off for it when you actually do the math behind it. Especially considering those of you who opt not to have a text messaging bundle currently and don't need the extra minutes, going with the shared data plan will cost you an additional $30.

I can only guess the motivation behind these plans, but I think somewhere along the lines marketing and sales came together and said, "Hey, people want to share their data. People are canceling their SMS plans left and right. So lets bundle unlimited SMS and unlimited Voice (which cost us hardly anything) and increase the cost of data to compensate for the cannibalism of our loss of SMS sales!" Fundamentally though, the market wants to say, "Hey, I'm paying $30 a month for 2GB of data. Let me share that with my other devices so I can pay $70 for 700 minutes, 3 smartphones sharing a 2GB plan for $30 and pay a connection fee of $10 each additional line bringing my total to $130 instead of paying currently $190." Unfortunately that would make too much sense...

Instead we have three problems to deal with:

Carriers are trying to increase their profit margins to combat increased costs and work to build infrastructure to capitalize on their ROI (Return on investment) in a timely fashion.
Consumers are already paying $100 a month for TV & Internet service, tack on another $100-$200 for family mobile plans. It eats into the budget. Consumers want reasonable costs they deem fair.
Points one and two are a conflict of interest.

Boil it all down, and what consumers probably want is a refund for the 1.5GB they don't use on their 2GB a month plan! Sound hard to believe? One company is actually doing this.

Update(7/20/2012): Endgadget also did a writeup of this showing even more of a comparison than I did. It's pretty good and worth a read for more information if you're investigating the comparison of these new plans.

Finding a Mercurial Editor - Part 2

2010-04-26T07:00:00.001-04:00

Oh the trials and tribulations! So I left off in Part 1 with "Documentation that looks promising"...but wasn't. Call me a fool, but I've spent way, way, WAY to much time looking into this issue. So what did I start doing? Command line!

It wasn't too bad just to get a viewer up and running. All I did was create a file ~/.hgrc and put the following inside it:

[hooks]
preoutgoing.verify = hg verify

[extensions]
hgk=
rebase =
transplant=
hgext.hbisect=

[hgk]
path=~/Mercurial//contrib/hgk

Now lets take a look at the parameters. In the hooks section, I added an outgoing hook to do an hg verify on the repository before I push it. Why? If you don't and you push your change sets to another repository, you'll effectively ~~infect~~ corrupt the repository for everyone else...and let's face it. You don't want your name on the commit tree when that happens.

In the extensions I added hgk so I could at least view the repository. It looks like this:

It wasn't hard to add. At least once you have tcl/tk installed, all you have to do is download the source/binary from the download page, extract it, and point to hgk file located in the contrib directory. You'll see the path where I extracted it to above under [hgk].

I had a comment in Part 1 from hluz, and I'm planning to re-investigate that route and post more detail when it becomes available. It'd me nice for this to be an easy process for Mac users...which might resort to doing it the old fashion way as opposed to good ole' Macports!

...and the Android is being renamed to Cyborg

2010-04-01T08:44:00.000-04:00

....well not really. But it's April 1st, and that means Google's got another gag!

Official Google Blog: A different kind of company name

The Power of @protocol

2010-03-25T07:00:00.000-04:00

It's nothing new, but @protocol is at the crux of satisfying the Interface Segregation Principle (ISP).

Now if you've never heard of this principle, it's a great fundamental design methodology. Here's how it works:

Say you have an application with an object called "Bank" and an object called "Client". Something the "Client" may want to do is withdraw money. Now you could just give the "Client" a pointer to the "Bank" object and have it call the method withdrawMoney(amount). That's what most programmers do right? So you did. Now I want you to test your "Client" object.

What's the problem? Nothing you say? Well let's play around with a few scenarios.

When you try to plug in the real "Bank" object, you realize that calling that method causes your test to fail. And worse off, it's the "Bank" object that's generating the error, not your client application. How do you prove this?
Ok, so you've thought about it a little more. You're going to replace the "Bank" object with a fake Bank. Coding....coding....gha!!! Ok, we've got a problem. "Client" takes a pointer to the real "Bank" object. Unless I modify my code for testing or do something really really ugly, I can't plug in my test object. Hmm...what now?
Your project manager comes in and says that we'll be handling different clients for the bank. Some only let you look at what money you have...so no withdrawMoney(amount) function! How are you going to handle this mess? It almost seems like everyone has to know about everyone else's business! But do they?

The Interface Segregation Principle can solve all these problems. I won't bother re-explaining how you do it in the iPhone, but the documentation on Apple's site is very good. If you want to see it in action, there's some code available in devkit-bb for the iPhone relating to the viewManagerProtocol.

When it rains, it pours...and when it floods?

2010-03-22T07:00:00.005-04:00

...it takes you longer to get to work... or does it? New Jersey tends to flood in the north, and unlike school, work does not close unless the state declares an emergency.

So that got me thinking again, how nice would it be to be able to telecommute to work? I mean seriously, especially on a Mac, if you have to talk to someone or work together on something, iChat makes that collaboration very possible...and easy! Multiple developers working on a project? Not a problem either! Just pile everyone all in an iChat room and it's like you're set up for agile development! With easy communication, open collaboration, and a short commute.... you have yourself software developer's dreamland!

Now herein lies the problem. Where are these jobs? For the most part, they don't exist much. But for smaller development firms, having an agile setup to do this kind of development could provide a competitive advantage that allows smaller companies to trim costs by not having to maintain extra space and overhead for an office, thus maximizing their ROI, not to mention, working from home is a definite boost in job satisfaction!

Playing devil's advocate for a second, the downside to these types of work environments is that you must ensure that employees working are fully committed to their trade and have a strong work ethic. You need to be able to rely on your developers to do what they say and be comfortable with communicating frequently and effectively.

So what do I say?

Set up a VPN and sign me up!

Finding a Mercurial Editor - Part 1

2010-03-18T07:00:00.004-04:00

One of the problems I've been running into on the Mac is finding a GUI on the Mac for Mercurial that provides a visual way to see the current tree structure, branch, commit, and push and pull repository changes. I really liked using TortoiseHg. But after running into problems with it corrupting my repository on my Mac using a shared directory in parallels, I quickly abandoned its use.

Thus, I'm back to square one...looking for an editor on the Mac that supports what I want and need in an editor. Now I tried Murky, and that was pretty good, but it still doesn't support branching. It also continues to plague me with this message:

So that's no good. Now I've begun to move onto different editors, all with semi-complex setups that I can't seem to overcome. First we'll start with the TortoiseHg OS X port.

Now I didn't have pygtk installed, so I cut over to Macports and let the installer run overnight...yes, parts of it took long enough that I simply decided closing my eyes was more appropriate than watching the installer finish. Now whether or not I installed the correct package is up for debate. I'm thinking not because after completing the operation and still trying to set environment variables and call macport python_select operations to switch to python25 or python26, I continued to receive the following error:

But wait! As I continue to jot down in my notebook things that I have or haven't tried, I've come across some documentation that looks promising! So here I go, off to follow another set of directions....

*chug* *chug* *chug* *chug*

(to be continued...)

Mercurial - Corrupt Repository

2010-03-15T23:07:00.001-04:00

There's nothing like getting deep into your development cycle, only to realize that your repository has corrupted itself and has replicated its corruption to the clone's that you've pushed your changes to. That's what I'm finding right now in Mercurial.

Luckily, I've figured out the cause of the problem. Currently I've been using TortoiseHg in Parallels. I'm not exactly sure why, but operating on a shared directory through parallels causes the index files to be corrupted on the first commit. You can try and recreate the repository all you want through Tortoise (1.0) and Mercurial (1.5) but you won't have any luck.

For this issue, you'll need to go back to your Mac, create it, and don't mess with your repository from parallels. As nice as TortoiseHg is, it's not worth corrupting your repository and pushing those changes to everyone else in your development environment.

I guess the issue now is, how do you go about repairing a repository with this corruption while still maintaing the history?

Google Down for a few minutes

2010-02-18T08:02:00.000-05:00

It may have just been me, but moments ago, all access to google services ceased as I was unable to access mail, search, as well as this blog. I was able to access facebook. If this happened to you, let me know. I'm curious why, security fix with Buzz maybe?

iPhone Tips & Hidden Gems!

2010-02-17T07:13:00.002-05:00

A friend of mine sent me a link yesterday to this site, which initially I thought had to do with iPhone development, but it's really hidden gems of using your iPhone. It's worth a read, as I found a few interesting features that I didn't really know existed before!

And for anyone reading, yes I'm aware the article is dated December 2008. So what if I like digging up old relics! :P

Devkit-BB

2010-02-10T22:33:00.001-05:00

I finally got around to creating a reuse repository for commonly used code. Actually it had existed, but I got sidetracked. It's called devkit-bb, the "bb" stands for "Building Blocks". Initially at the time of this writing, it contains simple code for iPhone developers to use to support the enabling and disabling of logging in their code where they can selectively enable or disable certain types of logs they wish to view. This is often a problem, and something that people reinvent over and over. No more! If you wish to join the project for better code, check out the code repository! Mind you, I'm using Mercurial in google code, so you'll need to be familiar with git style repositories.

Understanding Model-View-Control (MVC)

2009-08-31T07:00:00.002-04:00

If you're new to iPhone development or new to programming in general, I encourage you to take a look iTunes-U for a wide variety of classes and courses. In particular, look for the Stanford course for iPhone Development. It is complete with a wide array of resources, lecture slides, materials, and will definitely save you time in your quest to develop the next iPhone application.

Ok, enough with the free plug. What is MVC? The Model-View-Control design pattern is used extensively throughout the development of an iPhone application. It is used to separate the presentation of your graphical layout from the data that populates that view. The connection between these two components is the controller and is often referred to as "the glue". Understanding it properly will save you from my mistake, which incidentally led to "Class explosion". Here's the problem:

"I have way too many UIViews! It's getting hard to keep track of all my classes. It seems like my UIView is composed of nothing more than my UIText, UILabel, etc. Seems like a waste."

In fact...it is a waste. Hopping on the MVC train by simply explicitly saying...view elements go in the view, control logic goes in the control, and my model goes in its model area will saturate some of the power that the controller will provide for your view.

You might be saying, "Hey! That violates the design! The Controller shouldn't be managing elements contained in a UIView!" While from a purist standpoint, it tends to make sense.... the deficiency arises in the fact that you now need an entire class just to house the @property and the IBOutlets for each of your objects. It's somewhat overkill to have this many views doing really nothing in regards to the view.

If you're like me, you need to rationalize it somehow. I rationalize it by saying that each element populated in the UIView, is simply a type of view. For instance, UILabel has a property "setText", and that gets relayed to the actual view where it is displayed. UIView is as much of a view as UILabel is. The only difference is that UILabel doesn't have the word "View" in the class name. By placing the UILabel as an IBOutlet in the controller, you are providing easier access for the view's components. Additionally, the UIView then maintains a single responsibility, which is to control the placement of the view elements. The controller maintains the single responsibility of propagating model elements to the view.

So what have I found out works well in using the XCode IDE? Take advantage of Interface builder. Unless you need to do something with the actual UIView, you can pretty much get away without having to create a separate UIView class. Put your UILabel and objects in your viewController as IBOutlets. That way you can operate directly on the elements. If you follow the pattern, one controller per view, so you shouldn't need to worry about who's UILabel it is. It's the view that the controller owns. You'll greatly reduce the number of classes you need to keep track of, and accessing those views will be a much less tedious task.

Side Note - C++ reinterpret_cast

2009-08-25T01:00:00.000-04:00

If you find yourself handed code that has memory corruption issues and uses reinterpret_cast, take a closer look. Reinterpret cast in C++ is not meant to be used to downcast a derived class to a more basic class. Doing so will give you the incorrect address for the base type.

Instead, do a dynamic_cast. It's much safer and is the correct way to make a cast of this type.

Walking Backwards - Introductory Tutorials - The AppDelegate

2009-07-31T16:10:00.000-04:00

When I started coding for the iPhone, I spent the first few coding sessions reading and watching tutorials and books. They're priceless when it comes to learning at a rapid pace. In fact, there's a good deal of freely available video tutorials that show you first hand how other developers go about coding on the iPhone.

One example in particular I would like to highlight is the Hello World App. Like any program, it's great to start out with. However, you need to read it with a bit of caution...because if you put it into practice in the actual program, it doesn't scale rather well and things quickly grow out of hand.

For instance, the Hello World application has you create your delegate methods which provide functionality in your AppDelegate class. If you have one window and one view, it's perfectly fine. Now what if you have multiple views? Do you really want all your functions in the app delegate? If you said yes, you'll quickly find yourself swamped and lost in your own program.

So what does the Application Delegate do? The delegate provides a functionality to allow global references to do stuff. Now...how would you go about looking for a pay stub from last year? One way would be to put everything in a filing cabinet. Perhaps you might sort it by year?

However, maybe you're not organized. All your papers are in a big heap on the floor in a spare room that nobody uses. All the junk mail with coupons that you thought you were going to use is in there too. Now go find that pay stub! This is exactly what you don't want to do with your Application Delegate, you'll never find things and debugging and tracing the flow of your program will be a complete nightmare...both for you and anyone that works with you.

So what do you do? What the architecture eventually forces you to realize is that you'll want to house your view control logic in your ViewController. If you're using a XIB, this will be your "File Owner". References you make from the controller will be made simply to retrieve some global data, like a model representing the data you're trying to present and be accessible to multiple views.

By moving this control logic to your control, you compartmentalize a great majority of the logic in places where it is important and relates to the task at hand. If you want to to switch views, the task of switching is fairly easy, as you can call your delegate to swap out views from the root view.

Mysterious issues creating plist files

2009-06-26T21:36:00.000-04:00

So I've been using plists as a way to internally store data to my application. My data set has gotten so huge that I needed to beef up my unit tests to uncover some test algorithm bugs in order to produce a better histogram as to how well my application is performing.

Unfortunately, I ran into a little snafu. When I created my test plist file, I neglected to pick the correct type of file. I simply right clicked in XCode to add a new file. I didn't select text file, nor did I select plist. I selected a non-type kind of file (That's not the official name...but I'll update this blog once I get back home).

Suffice it to say when you don't, you miss out on a key piece of information in the DTD part of the XML telling XCode that you're writing a property list. It will proceed to populate your list with enumeration type data sets rather than XML plist when editing it in XML view in the XCode editor. If you view the file as a text file, you'll see what I mean. This came to bite me in the...well you know, and I proceeded to get exceptions in my application telling me that a string isn't a BOOL.

Doh!

Anyway, I'll clean this blog up later with some screenshots and the like, but in case you have a coding craving this weekend, here's something that may have gone under your radar.