Well, finding the presents did not cost us a lot of time. A new bag for school, some stuff for her new room and the touchscreen of her Nintendo DS was broken, so replacing that would also be an awesome present. After looking around on the internet, I (suffering from this again) decided it would be better to just order a new Nintendo 3DS. Her being a girl and all we looked where we could find the pink version, which were hard to find this close to the holidays. After a search we found one at a webshop, that actually had one in stock, YAY!

Ordered it, paid iDEAL (instant payment from your bank account) and received a confirmation of my order within a couple of minutes. Success! The next day I received a confirmation that my order was packed and prepared for sending using the company GLS. I received a tracking code and everything seemed fine.

A day later, I was working till around 16:00 (4:00 PM) which meant I was home all day, but after that we had to go to one of nieces’ birthday party, but the package was not delivered yet. I checked the tracking code online and it showed me the package was ready to be delivered, but that the chauffeur hadn’t reached my house yet. No worries there, because most of the transport companies just deliver the package to one of my neighbors when nobody answers the door.

An hour later, at the party, I check the tracking code again, eager to see if the package has been delivered.

Yay! The package has been delivered! It says “Aflevering” at “Laatste afhandeling”, which means that it has been delivered. In the column “Naam ontvanger” it says “RENE”, which means that the person accepting the package is called Rene. That worried me a bit, because I know all of my neighbors by name and none of them is called Rene, but I thought it could be the weird text-recognition thing they use when signing for the package. I am glad that we can actually see the signature of the person accepting the package.

Ok, it still says that it was accepted by someone called Rene and the signature, well, it’s a signature. So after we got home from the party, it was about 19:30 (7:30 PM) we thought we would find a note which would say at which house it was delivered instead of ours, but there was no note (crap). So I decided to go to our nearest neighbors first, to ask if they have received a package. After ringing a couple of doorbells and asking my neighbors if maybe they received anything I got a bit more worried. Asking anyone in the street would be a lot of doors to ring, so I decided to do a couple more and ask if they knew anyone in our street called Rene. Nope, no package to be found and no Rene lives in our street.

So that evening my wife decides to send an email to transport company, asking where our package is. The next morning I decided a shorter route would be to just give them a call. A very friendly woman on the phone told me she already read our email and she would call the chauffeur to ask where our package went. While I was on hold, she called the chauffeur and asked where he did leave the package, his answer was: “I don’t know anymore, maybe I’ll remember when I get back to their house”. Of course this person delivers a lot of packages each day, so not remembering everything is OK, but it seems weird to not register it in your system when you deliver a package to a different house when the recipient is not at home. He told the woman on the phone he would drive by my house to clear everything up.

About an hour later, my doorbell rings. The chauffeur was at the door, with a package in his hands! He told me that it was delivered at my neighbors house, where he just picked it up and took it to my house. Sure I am glad that the package was found, but that story is just weird. I talked to that neighbor the evening before and they didn’t receive any packages, I know all my neighbors pretty well and consider them trustworthy. Was the package indeed delivered to them or did the chauffeur make stuff up?

After I received the package, I received an email from the office of GLS, asking me if everything was cleared up. Of course, I got my package, but it all seems very strange. I am not really sure what is going on, but with all the stories of packages disappearing people tend to draw their own conclusions.

Would my package ever show up again if we didn’t call/email the company immediately?
Where was the package, when they said it was delivered, but it wasn’t?
Should a tracking system no keep track of a package until it was actually delivered to the recipient?

I don’t know the answers to these questions, but you really don’t want to know the answers I think of.

I will extend this list when I get more tools and experiences with them. If you have any tools I NEED to have, just post a comment!

My post about using a NSURLConnection in a different thread has been read a lot. I made this post almost 3 years ago and I would like to show you how I would solve it nowadays. In the last 3 years I have learned a lot and the cocoa-touch framework evolved too.

Between that post and now I have been trying out frameworks that support async operations and after using ASIHTTPRequest for a long time, it was time to move on (the project was discontinued). After trying out several more, I came across LRResty, which is a nice framework designed with code-blocks in mind.

Here is a small example of a piece of code, which is taken directly form the LRResty documentation page:

[[LRResty client] get:@"http://www.example.com"
            withBlock:^(LRRestyResponse *response) {

    if(response.status == 200) {
      NSLog(@"Successful response %@", [response asString]);
    }
}];

This piece of code fetches the content of an URL (www.example.com) and when the response comes back, it checks if the response is OK (status code 200). When it is, it does a simple NSLog.

Here you can see the power of code-blocks already. You define the way it handles a responses directly when you do the request, in my opinion this keeps code clean and simple.

Now we can take this a step further and see what happens if we launch this in a new thread. For this example I am taking a piece of code of an actual library I was writing some time ago, it involves getting some data through the GitHub API, waiting for the response and then executing a certain piece of code, while this can be a bit overwhelming at first, it sure beats writing dozens of extra methods for each response.

Because we are going the be launching the request from another thread, we have to take race conditions into account, this was a bug that hit me when I was first using code-blocks and multi-threading.

I created a method which can be called with a single NSDictionary as parameter, containing code-blocks and other stuff which it needs to perform the request. Because the GitHub API needs authentication also, I will also show how that works. The dictionary contains 3 objects with the keys ‘success’, ‘failed’ and ‘url’.

The objects linked to the ‘success’ and ‘failed’ keys are code-blocks, I will show the contents later.

// This method performs the actual request and can be called in a different thread.
- (void) startRequest:(NSDictionary*) parameters
{
// Take the URL out of the dictionary and store it in a NSString object.
	NSString* url = [parameters objectForKey:@"url"];

// Get both the 'success' and 'failed' code-blocks from the dictionary
	void(^success)(LRRestyResponse*) = [parameters objectForKey:@"success"];
	void(^failed)(LRRestyResponse*) = [parameters objectForKey:@"failed"];

// Check if there already is a client active in this class and if not, check if authentication is set.
// If not, just raise an exception for now.
	if (client == nil && self.username == nil && self.password == nil)
	{
		[NSException raise:@"Authentication missing" format:@"Username and/or password not set."];
	}

// No client is predefined and should be instantiated with the right credentials.
	if (client == nil)
		self.client = [LRResty authenticatedClientWithUsername:username password:password];

	NSLog(@"Requesting data from: %@", url);

	[client get:url withBlock:^(LRRestyResponse* response) {
// Grab the headers of the response
		NSDictionary* dict = response.headers;

// Check if the request was finished without errors, if there were no errors, execute our 'success' block.
// Otherwise execute our 'failed' block.
// The response is a parameter for our success and failed block.
		if ([[dict objectForKey:@"Status"] isEqualToString:@"200 OK"])
		{
			success(response);
		}
		else
		{
			failed(response);
		}
	}];
}

With this method I can perform any request I want using a different thread. I am not a 100% sure this covers it in your cases (it probably doesn’t).

Now, lets see the code which we use to actually call the startRequest method:

// This gets a page of gists using the GitHub API. When it is succesfully done it executes a supplied code-block which
// extracts the gists from the response and returns an array. When it fails it will execute a code-block
// which will show an error in the application.
- (void) getGistsForPage:(int) page success:(void(^)(NSArray*)) succesBlock failed:(void(^)(void)) failedBlock
{
// Prepare the URL for the request.
	NSString* gistsUrl = [NSString stringWithFormat:@"%@%@", self.githubUrl, @"/gists"];

// Here I define the code to be executed when the request is completed successfully. As you can see
// it grabs an array from the response and executes the supplied 'successBlock' with that array
// as a parameter.
	void (^success)(LRRestyResponse*) = ^(LRRestyResponse* response) {
		NSArray* array = [response.responseData arrayValue];
		succesBlock(array);
	};

// This is the code that will be executed when the request fails. It simply executes the
// supplied failedBlock code.
	void (^failed)(LRRestyResponse*) = ^(LRRestyResponse* response) {
		failedBlock();
	};

// Call the startRequest method, with a dictionary which contains our succesBlock, our failedBlock and the URL
// we created earlier in this method.
	[NSThread detachNewThreadSelector:@selector(startRequest:)
							 toTarget:self
						   withObject:[NSDictionary dictionaryWithObjectsAndKeys:
									   success, @"success",
									   failed, @"failed",
									   gistsUrl, @"url",
									   nil]];
}

It’s pretty straightforward. Now just to be complete, I will post the code I use to call the getGistsForPage method too, so you get a complete picture:

// Instantiate our API class, containing the methods I described above
GitHubAPI* api = [[GitHubAPI alloc] initWithURL:@"https://api.github.com"];

// Make sure the username, password and delegate are set
api.password = @"your_password";
api.username = @"your_username";
api.delegate = self;

// Call the method, which uses the multi-threaded part.
[api getGistsForPage:0 success:^(NSArray* array) {
	NSLog(@"Array count: %lu", [array count]);
} failed:^(void) {
}];

That’s it, there is nothing more to it. For me this works like a charm!

PS. You can still use a delegate with LRResty, instead of the code-blocks. You need to implement this method when using it this way:

- (void)restClient:(LRRestyClient *)client receivedResponse:(LRRestyResponse *)res;
{
  // do something with the response
}

You can also use Illume to send morse code messages. Just switch to the ‘morse mode’ and type in your message and hit the play button.

Illume is available in the AppStore, click on the button below to download it!

What will be in v1.1:

Receiving morse messages using the camera of the iPhone

Like you can see above in the image, it just looks awesome!

To achieve this, I followed a couple of simple steps:

• Create a cool button image, where you want the circle progressbar to appear in
• Find out the center and radius of the circle progressbar, you can do this either by using an image editor or just guess some values and tweak them when they are not completely perfect.
• Create a class subclassing a UIView, which has 2 properties : MaxValue (the value that is the maximum you want to use) and Progress (what is the current progress of the progressbar)
• Put it all on another View and change the Progress property when the progress of your process changes

//
//  SBCircleProgressBar.h
//
//  Created by Wim Haanstra on 17-11-11.
//  Copyright (c) 2011 Sorted Bits. All rights reserved.
//

@interface SBCircleProgressBar : UIView

// The maximum value of the progressbar
@property (nonatomic) float MaxValue;

// The current value of the progressbar
@property (nonatomic) float Progress;

@end

//
//  SBCircleProgressBar.m
//
//  Created by Wim Haanstra on 17-11-11.
//  Copyright (c) 2011 Sorted Bits. All rights reserved.
//

#import "SBCircleProgressBar.h"

@implementation SBCircleProgressBar

@synthesize MaxValue, Progress;

// SOME VALUES YOU CAN CHANGE TO CHANGE THE BEHAVIOUR OF THE PROGRESSBAR
#define MAX_DEGREES	360
#define START_DEGREES	270

// Define the radius that the circle needs
#define CIRCLE_RADIUS	27

- (id)initWithFrame:(CGRect)frame
{
    self = [super initWithFrame:frame];
    if (self) {

		self.backgroundColor = [UIColor clearColor];
		self.userInteractionEnabled = NO;

		// Default values for the progress bar
		self.MaxValue = 100.0f;
		self.Progress = 0.0f;

    }
    return self;
}

- (void) setMaxValue:(float) _MaxValue
{
	MaxValue = _MaxValue;
	[self setNeedsDisplay];
}

- (void) setProgress:(float) _Progress
{
	Progress = _Progress;
	[self setNeedsDisplay];
}

- (void)drawRect:(CGRect)rect
{
	CGContextRef context = UIGraphicsGetCurrentContext(); 

	// Set the color of the circle to appear for the progressbar
    CGContextSetRGBStrokeColor(context, 0.341, 0.635, 0.961, 0.6); 

	// Set the line width of the circle to appear
    CGContextSetLineWidth(context, 10.0); 

	// Calculate the middle of the circle
	CGPoint circleCenter = CGPointMake(self.frame.size.width / 2, self.frame.size.height / 2);

	// Calculate the amount of degrees the circle needs to have filled
	float currentDegrees = (MAX_DEGREES / self.MaxValue) * self.Progress;

	// Draw the ARC (part of the circle
	CGContextAddArc(context, circleCenter.x , circleCenter.y, CIRCLE_RADIUS, radians(START_DEGREES), radians(START_DEGREES + currentDegrees), 0); 
	CGContextStrokePath(context);
}


@end

You need to put this view OVER the image of your button. Because we do this, we also set userInteractionEnabled otherwise you wouldn’t be able to push your button anymore, because the progressbar would catch all the touches.

The method “radians” you see in there, is a small inline method used to calculate degrees to radians:

static inline float radians(double degrees) { return degrees * M_PI / 180; }

This site already won an innovation award given to us by SIDN. The site was also mentioned in several other media.

I am always looking for blogs to exchange links. I will put it up on my blogroll and maybe link to interesting articles.

You can post them here or just email them to me @ wim@wim.me.

For example my TV. When I bought it I knew exactly what it looked like. The front, sides AND the back. I had skipped through the device manual to see if certain stuff was possible and it had everything I was looking for.

The whole process of comparing and buying it took me over 4 days. Buying it was a breeze though, because all I had to do was to find the web-shop with a good prize/delivery-time.

Only sometimes this process really haunts me.

Last week my oldest daughter (of 6 years old) had to have one of her back tooth pulled and because this is a very painful process we promised her a gift once it was all over. My wife and I already talked it over, because she wanted a radio/cd player on her room for a long time now, so this was the perfect moment to get it for her.

Well this weekend we have been hunting the internet for a radio that she liked. First of all, it had to be pink (according to her), have a FM radio and a CD player. Finding something like that, for an affordable price (let’s say for around 50 euros) was easy.

Then it struck me again, I was not going to burn all her MP3′s to CD anymore. She already has a MP3 player (an old iPhone 3G, 16GB) with all of her favorite songs and it would make more sense to search for a radio that supported external sound sources. I also didn’t want her to have plastic speakers with it, because they sound like ****. Before I knew it I started looking for well-known brands that offer simple radio’s and HIFI sets for that kind of money and I was wiping that “it has to be pink” feature of the table.

Long story short. Why do I give so much about this stuff? Why does the radio of my daughter have to fit in my ideas of good equipment? Sure, a better product lasts longer, but we are not talking about a radio of hundreds of euros here. By the time she turns 10, she wants a new one anyway, because of how out-dated it is.

I just hope my wife takes care of this, before we have to install a 7.1 DTS certified receiver in her room, which has an iPhone integration unit and a touch-screen remote control.

PS. I do love it though, spec-comparing is really the best there is. After the purchase that all fades away and it’s just a normal TV/radio/phone/etc.

There are some methods I use regularly and most of them apply to the string class. So without further ado, here are some of my favorites.

Validating

public static bool IsInt(this string text)
{
    int integer = 0;
    return Int32.TryParse(text, autoload =  out integer);
}

Of course you can create methods like this for all kind of types to support. This one just shows an example for the Int32 type. This way I can easily check if variables are correct. It is too bad that you can’t just parse the variable into the void so an extra variable is needed.

One other is an extension to make a valid filename from a string, like this:

static public string ToValidFileName(this string name)
{
    string invalidChars = Regex.Escape(new string(Path.GetInvalidFileNameChars()));
    string invalidReStr = string.Format(@"[{0}]+", invalidChars);
    return Regex.Replace(name, invalidReStr, "_");
} 

Hashing
There are several reasons to hash a string, I personally use it to store passwords. Nowadays I use SHA1 hashing because MD5 is not that secure.

public static string ToSHA1(this string text, Encoding enc)
{
    byte[] buffer = enc.GetBytes(text);
    SHA1CryptoServiceProvider cryptoTransformSHA1 =
    new SHA1CryptoServiceProvider();
    string hash = BitConverter.ToString(
    cryptoTransformSHA1.ComputeHash(buffer)).Replace("-", "");
    return hash;
}

public static string ToMD5(this string input)
{
    MD5 md5 = System.Security.Cryptography.MD5.Create();
    byte[] inputBytes = System.Text.Encoding.ASCII.GetBytes(input);
    byte[] hash = md5.ComputeHash(inputBytes);
    StringBuilder sb = new StringBuilder();

    for (int i = 0; i < hash.Length; i++)
    {
        sb.Append(hash[i].ToString("X2"));
    }
    return sb.ToString();
} 

Encrypting/Decrypting
Encrypting and decrypting data is very important in one of my latest projects. For that reason I created these 2 extension methods. They encrypt and decrypt a string with a supplied passphrase. If there was no passphrase supplied, it will try to fetch one from the .config file.

public static string Encrypt(this string Message, string passphrase = "")
{
	if (passphrase == "")
		passphrase = ConfigurationManager.AppSettings["encrpytionPassphrase"];

	byte[] Results;
	System.Text.UTF8Encoding UTF8 = new System.Text.UTF8Encoding();

	MD5CryptoServiceProvider HashProvider = new MD5CryptoServiceProvider();
	byte[] TDESKey = HashProvider.ComputeHash(UTF8.GetBytes(passphrase));

	TripleDESCryptoServiceProvider TDESAlgorithm = new TripleDESCryptoServiceProvider();

	TDESAlgorithm.Key = TDESKey;
	TDESAlgorithm.Mode = CipherMode.ECB;
	TDESAlgorithm.Padding = PaddingMode.PKCS7;

	byte[] DataToEncrypt = UTF8.GetBytes(Message);

	try
	{
		ICryptoTransform Encryptor = TDESAlgorithm.CreateEncryptor();
		Results = Encryptor.TransformFinalBlock(DataToEncrypt, 0, DataToEncrypt.Length);
	}
	finally
	{
		TDESAlgorithm.Clear();
		HashProvider.Clear();
	}

	return Convert.ToBase64String(Results);
}

public static string Decrypt(this string Message, string passphrase = "")
{
	if (passphrase == "")
		passphrase = ConfigurationManager.AppSettings["encrpytionPassphrase"];
	if (Message == string.Empty)
		return string.Empty;

	byte[] Results;
	System.Text.UTF8Encoding UTF8 = new System.Text.UTF8Encoding();

	MD5CryptoServiceProvider HashProvider = new MD5CryptoServiceProvider();
	byte[] TDESKey = HashProvider.ComputeHash(UTF8.GetBytes(passphrase));

	TripleDESCryptoServiceProvider TDESAlgorithm = new TripleDESCryptoServiceProvider();

	TDESAlgorithm.Key = TDESKey;
	TDESAlgorithm.Mode = CipherMode.ECB;
	TDESAlgorithm.Padding = PaddingMode.PKCS7;

	byte[] DataToDecrypt = Convert.FromBase64String(Message);

	try
	{
		ICryptoTransform Decryptor = TDESAlgorithm.CreateDecryptor();
		Results = Decryptor.TransformFinalBlock(DataToDecrypt, 0, DataToDecrypt.Length);
	}
	catch
	{
		Results = new byte[0];
	}
	finally
	{
		TDESAlgorithm.Clear();
		HashProvider.Clear();
	}

	return UTF8.GetString(Results);
}

I am aware that there are other companies doing the same, but my experience with them hasn’t been that good. Most of them lack features, act nervously when trying to scroll or other defects.

So we thought it was time to create something ourselves and make it available. What can our control do, I hear you ask (well not really, it’s the internet, we don’t really hear you), well here are the features:

• Add ViewControllers to the control without much changes to them. All you need are about 3 lines of code for the base functionality to work.
• Lots of configurable options, don’t like shadows? Turn the off. Want to make them animate faster/slower? Just change the duration, etc etc.
• Notifications: whatever happens, a notification can be send for it. If a view is moved, you can receive a notification for that. A view barely visible? Notification. Etc etc.
• Disable the sliding gesture on a viewcontroller: When you view expects a sliding gesture, but our control is stealing it, you can turn it off for that view controller

Here is a short movie showing the default functionality of the control:

So what makes our controls better then the other ones?

• When you buy something from us, we take support seriously. Depending on the type of license you buy, you get either forum support, email support or within 24 hour email support. This because we know that you can’t wait 2 weeks to get an answer (or in some of our experiences more than 2 months) to your questions.
• The cheapest option is to just buy the compiled control, but you can also buy the source code, which makes it easier for you to customize anything you want.
• Try before you buy, you can just download a free version of the component to see if it works like you want to.
• We take coding seriously, so we document our code.
• Full support for upcoming iOS releases, we test our controls on the latest iOS releases available too.

Currently the control is already available for purchase through the links below, but is also going through some test phases. When purchases are made, we will notify you when a new version is available for download and upgrades for the 1.x releases are free for all users.