Write native iPhone applications using Eclipse CDT

Write native iPhone applications using Eclipse
CDT
How Windows and Linux developers can bypass the iPhone
SDK and write iPhone apps using open source tools

Skill Level: Intermediate

PJ Cabrera (pjcabrera@pobox.com)
Freelance Software Developer and Writer
Freelance

30 Sep 2008

Learn how to use the Eclipse C Development Toolkit (CDT) to program native
applications for the Apple iPhone, using open source tools to enable iPhone OS
development on any Eclipse-supported platform.

Section 1. Before you start

About this tutorial
In July 2008, with the introduction of the Apple iPhone 3G and the iPhone OS V2.0,
Apple Computer released production versions of an application development
environment for the creation of native iPhone and Apple iPod Touch applications by
third parties. But the application development environment is for Apple Mac OS X
only, leaving developers using the Microsoft® Windows® and Linux® operating
systems out of the loop with regard to developing native applications for this exciting
platform.

This tutorial explores the Eclipse C Development Tooling (CDT) project and how it
allows you to work with C- and C++-based projects. In this tutorial, you install open

Write native iPhone applications using Eclipse CDT
© Copyright IBM Corporation 1994, 2008. All rights reserved. Page 1 of 34

developerWorks® ibm.com/developerWorks

source tools to help develop native applications for the iPhone platform on Windows
and Linux, and learn about the source code of a basic iPhone application in
Objective-C.

Objectives
This tutorial shows how you can use Eclipse and the CDT plug-ins to develop for the
iPhone platform, using the open source GNU Compiler Collection (GCC). You use
these tools to compile an Objective-C example application for the iPhone.

Prerequisites
This tutorial is written for C and C++ programmers whose skills and experience are
at a beginning to intermediate level. You should have a general familiarity using a
UNIX® command-line shell and a working knowledge of the C language. The iPhone
Cocoa Touch native application development frameworks are written in Objective-C
and are based on the Cocoa frameworks used in Mac OS X. Familiarity with
Objective-C and Mac OS X Cocoa application development frameworks is not
assumed, but such skills are helpful in making sense of the iPhone platform. Further
study of Cocoa development is recommended after following this introductory
tutorial.

System requirements
To follow along with this tutorial, you need:

• A computer running Windows or Linux with at least 1.5 GB of free disk
space.
• An installation of Eclipse.
• Permission to write to the /usr/local directory in the file system
(C:/cygwin/usr/local for Windows users).
• Permission to modify your Eclipse configuration by installing plug-ins.

Section 2. The iPhone and iPod Touch platform
The iPhone platform and its cousin, the iPod Touch, are a WiFi-enabled mobile

Page 2 of 34 © Copyright IBM Corporation 1994, 2008. All rights reserved.

ibm.com/developerWorks developerWorks®

phone and a mobile Internet device, respectively, with a state-of-the-art graphical
user interface (GUI), a powerful integrated Web browser, and various other
applications, such as e-mail, calendaring, contacts, note-taking, basic Geographic
Information System (GIS) functions through its integration of Google Maps, as well
as entertainment functions, such as movie and music playing.

The iPhone platform is based on the same Darwin operating system and
Objective-C Cocoa frameworks as Mac OS X, modified to fit on an Advanced RISC
Machines (ARM)-based embedded device with 128 MB of RAM, several gigabytes
of flash storage, and a multi-touch touchscreen. The platform also features an
accelerometer for detecting device orientation and movement, and 3-D graphics that
rival the capabilities of the Sony PSP and PS2 consoles.

Darwin, Objective-C, and the Cocoa Touch frameworks
Objective-C is a superset of the C language that adds object-oriented features to C,
but with a syntax much different than that of C++. Objective-C was used by NeXT
Computer in the late-1980s and early 1990s as the development language of choice
for its NextStep and OpenStep operating systems. In the mid-1990s, Apple acquired
NeXT Computer's technology and assimilated much of NextStep and OpenStep into
the Mac OS operating system, producing Darwin and Mac OS X out of the mix.
Darwin is an open source UNIX-based operating system created by Apple as the
core of the Mac OS X platform.

As with NextStep and OpenStep, the development tools for Mac OS X and the
iPhone are based on Objective-C. Apple created a development framework based
on Objective-C for Mac OS X: Cocoa. The iPhone platform is based on the same
Darwin operating system core and Objective-C Cocoa framework as the Mac OS X
platform.Cocoa Touch is the development framework for the iPhone.

C++ and Objective-C are object-oriented extensions of the C language. But there
are key differences between C++ and Objective-C. Objective-C is a language and an
object-lifetime management runtime, whereas C++ is only a language, and the
programmer manages object lifetime entirely. The other main difference is the
syntax of object-oriented operations between the two. C++ extends the C struct
manipulation syntax, whereas Objective-C invents a new syntax with scoped square
brackets. Table 1 presents the main syntactic differences between the languages.

Table 1. Main syntactic differences between C++ and Objective-C
Object-oriented C++ Objective-C Explanation
concept
Method invocation object_or_class.method(2,
[object_or_class C++ extends the C
4); methodWithParam1: language syntax for
2 param2: 4]; accessing elements of
a struct and reuses it
for method invocations



on classes or objects.
Objective-C invented a
different syntax with
scoped square brackets
and named parameters
terminated by colons.
Object instantiation UIView UIView In C++ the new
contentView = new contentView = keyword allocates
UIView(x, y, [[UIView alloc] memory for the object
width, height); initWithFrame: and calls one of the
CGRectMake(x, y, declared constructors
width, height)]; that most closely
matches the
parameters you pass in
the code. In
Objective-C, you call
the alloc method of
the class, which returns
an uninstantiated object
of that class. Then, you
call an init method on
the object returned by
alloc to set initial
values for the instance.
Class definitions C++ extends the C
class MyClass @interface MyClass : NSObject { definition
struct
{ int myVariable; syntax by adding the
private: } ability to define
int myVariable; methods within a
void privateMethod(); @private struct or class. It
public: - (void)privateMethod; also adds the private
void setMyVariable(int param);
@public and public keywords
} - (void)setMyVariable: (int)param; as the
(as well
@end protected keyword
not used in this
example) to provide the
object-oriented feature
of encapsulation. In
Objective-C, the
@interface and @end
keywords are used to
define the attributes of
a class. Objective-C
uses the keywords
@private, @public,
and @protected with
equivalent meaning to
the similarly named
keywords in C++.
Classes in Objective-C
have to inherit from
another class. In this
example, the class
NSObject, whereas in



C++, a class can be the
head of its own class
hierarchy.
Class and instance C++ uses the static
methods class MyClass @interface MyClass : NSObject { within the
keyword
{ int instanceVariable; class body to mean the
private: } method or variable is a
static int classVariable; class method.
int instanceVariable; + (void)classMethod; Objective-C uses a
public: - (void)instanceMethod: (int)param; + in front of
prepended
static void classMethod();
@end a class method and a
void instanceMethod(int param); prepended - in front of
} an instance method.
Method declaration C++ adds the scoping
MyClass::privateMethod()@implementation MyClass
{ operator :: to declare
myVariable++; - (void)privateMethod { that a method belongs
} myVariable++; to a specific class.
} Objective-C uses the
MyClass::setMyVariable(int param) { @implementation
myVariable = param; - (void)setMyVariable: (int)param { keywords to
and @end
} myVariable = param; mark a group of
} methods as belonging
@end to a specific class.

Section 3. Removing application restrictions from your
iPhone
Unravel the background behind iPhone application development and discover how
to create your own iPhone applications.

Jailbreak
When Apple announced the iPhone, the company downplayed the lack of an
unfettered application development environment, citing the need to protect the
cell-provider data network from malicious applications running on hacked phones.
When introducing the iPhone software development kit (SDK) in March 2008, one of
the key features was that the iPhone would only be allowed to run applications
signed by Apple using unique keys handed out to iPhone developers with whom
Apple has a relationship.

Since September 2007, a group calling itself the iPhone Dev Team developed
methods for removing the restrictions Apple devised to limit what could be done on



the iPhone. The group helped create the open source tools used to compile
programs for the device.

The iPhone Dev Team also created tools that remove restrictions from the iPhone
and iPod Touch. These tools permit iPhone owners to use their devices on networks
other than AT&T in the United States or to use the iPhone in countries without an
iPhone carrier. This process is called unlocking.

This tutorial covers "jailbreaking" only, which consists of allowing applications not
blessed by Apple to run on the device. Note that the use of these tools is not
supported by Apple, and their use may void the device's warranty if Apple has
evidence of third-party software modification.

Before you begin
The information presented here is believed to be accurate as of this
writing. The steps needed to jailbreak a device may vary with later
versions of iPhone OS, QuickPwn, XPwn, Pwnage, and WinPwn.
Read the QuickPwn, XPwn, Pwnage, and WinPwn documentation
carefully before you begin.

As when using any other unfamiliar software, take your time before
you start. Read ahead to understand where you are going and be
certain you understand the possible outcomes to using these tools.

Jailbreak tools: QuickPwn, XPwn, Pwnage, and WinPwn
The iPhone Dev Team wrote a series of tools to jailbreak and unlock the iPhone and
iPod Touch, each with increasing user-friendliness and sophistication. Installation
details and pointers to online documentation are available in Resources.

If you own an iPhone, already use it with an Apple-approved service provider
account for your country, and only want to run applications of your own development
on the device, you need a quick and easy jailbreak solution. Similarly, if you have an
iPod Touch, a simple jailbreak tool is all you need.

Jailbreaking your device

QuickPwn performs the jailbreaking task quickly. All that's required is to download
the correct version of the iPhone OS for your device (only iPhone OS versions 2.0
through 2.0.2 are supported at the time of this writing). Then, connect it to your
computer with the USB charging/data cable and put your iPhone in what is called
DFU mode, as per the instructions in the software. In less than 10 minutes, your
device is jailbroken.

If you are using an iPhone in a country where there is currently no Apple-approved
cell service provider, or you want to use an iPhone with a provider other than the



designated iPhone service provider for your country, the iPhone Dev Team offers
Xpwn, Pwnage (see Figure 1), or WinPwn.

Figure 1. Pwnage is flagship tool for jailbreaking and unlocking iPhones

Installation details and pointers to online documentation are in the Resources
section. Basically, Xpwn, Pwnage, and WinPwn allow you to create a jailbroken
version of the iPhone OS on your computer. Then you use iTunes to download the
jailbroken OS into your iPhone. This sounds more complicated than it really is. The
tools do all the difficult work and use the same automated mechanisms Apple uses
in iTunes to upgrade the iPhone OS.

Preparing for development
After you have jailbroken your device and restarted it, you have a device with a full
UNIX command line and several utilities preinstalled to make development easier.



But to tap that command-line and enable development, you must install on the
device the OpenSSH secure remote shell client and server, the Linker Identity
Editor, and the Respring utility to restart the dashboard. These enable you to
transfer your programs to the iPhone and log in to execute commands on the device,
sign your code with a bogus key to make the iPhone run it, and reset your
dashboard to show your newly copied applications.

Software installation is made simple on jailbroken iPhones through the use of Cydia,
a package manager front end for APT. Linux users are probably giddy right now as
the previous sentence sinks in: APT package management, the one used by
Debian-based Linux distributions including Ubuntu? Yes, indeed, my friends. Yes,
indeed.

Before running Cydia, make sure you're connected to either the cell data network or
WiFi — preferably the latter, as OpenSSH is a bit hefty. Run Cydia, wait for it to
download package manifests from the built-in sources, then click Search at the
bottom of the Cydia window. Search for openssh and install it. Then search for ldid
and install it. Finally, search for respring and install it.

Now that you have installed OpenSSH on your iPhone, you need to secure it. All
iPhones come with two user accounts: root and mobile. Both have the same
password: alpine. Log in to your device and change the password for these two
accounts. Make them different from each other — and memorable. Connect your
device to your WiFi network if you haven't done so already and find its IP address
using the Settings application.

Note: The first time you connect to the iPhone through OpenSSH, the device can
take a minute or so to appear to respond because it is generating unique system
keys. Subsequent connections should be nearly instantaneous.

You're ready to get started creating your own software for the device. The next
section shows how to install the toolchain that makes programming the iPhone
possible.

Section 4. Installing prerequisite tools
When developers refer to a toolchain, they mean the compilers, assemblers, linkers,
and other tools that work together to build an executable file for a particular platform
from nothing but sources and libraries. The iPhone toolchain used by Apple for its
iPhone SDK is actually based on the open source GCC — the same toolchain used
for Linux, Sun Solaris, and Berkeley Software Distribution (BSD) development.
Because of its open source origins, the base source code for the iPhone SDK is



actually available. Members of the iPhone Dev Team have produced instructions
and patches to enable developers in other platforms to build and use it. That's what
we do next.

There are a couple of requirements before we begin. You must already have a
toolchain installed for the platform you're running. You need tools installed that can
convert the GCC source code into executables for your operating system.
Depending on your platform, you can install tools for Windows or Linux.

Install prerequisite tools in Windows: Cygwin
Cygwin is a set of Windows programs that creates a UNIX-like environment for
Windows. Rather than use virtualization or emulation, the Cygwin team created a
Windows dynamic loading library that implements a good deal of the UNIX
application program interface (API), then proceeded to patch the source code of
many popular UNIX tools until they compiled and ran with sufficient stability on
Windows. They have been perfecting that environment for more than 13 years. See
Resources for links to the Cygwin site, where you can download the Cygwin setup
program. The Cygwin setup program is actually a rather small download because
you download the rest of the programs that make up the Cygwin distribution during
the last part of the Cygwin installation.

To install and configure Cygwin:

1. Downloaded the Cygwin setup program, run it and click Next at every
prompt until you come to the window shown below.
Figure 2. Cygwin root directory selection part of the installation



2. Make sure the options at the bottom of this window match those in Figure
2 or you'll have trouble working with the example code and toolchain
sources you download in the next sections of this tutorial. If you install
Cygwin to a different directory from the one in Figure 2, you must keep
that directory in mind and make modifications to the steps given in the
other sections of this tutorial to match your installation choice. I suggest
you keep things simple and follow the example.

3. Continue through the Cygwin installation windows, leaving the default
choices selected, until you come to the mirror selection screen, shown
below.
Figure 3. Cygwin mirror selection part of the installation



4. Select a mirror site close to you to distribute the bandwidth use
geographically. Try using the mirror at a university in your state or
province, for example. If you don't recognize any of the URLs as "local,"
you can't go wrong choosing either of cygwin.osuosl.org or
mirrors.kernels.org. Choosing URLs that begin with HTTP rather than
FTP might help if you're having problems with firewalls or proxies within
your network.

5. Continue through the installation until you come to the package-selection
window, shown below.
Figure 4. Cygwin package-selection part of the installation



6. Expand the Devel category, shown below.
Figure 5. Expand the Devel category to show the packages within
that category



7. For the autoconf, binutils, bison, flex, gcc-c++, and make packages, click
Skip to the left of the package you want to include in the installation, as
shown below. Clicking Skip once selects the Binary column check box
and the version number of the package to show where the "skip" was.
Note that the actual version numbers may differ from what is shown
because packages are updated continuously.
Figure 6. Select the packages to include in the installation



8. Scroll down, and expand Net, as shown below.
Figure 7. Expand the Net category to show its packages



9. Select the openssh and openssl packages, shown in Figure 8.
Figure 8. Select the openssh and openssl packages



10. Click Next. Cygwin setup downloads the packages from the mirror site
and installs itself at the directory specified earlier. This process can take
upwards of an hour, even on 2-megabit broadband and faster, because
those mirrors are in high demand, not only for Cygwin packages but for
other open source software they mirror.

When Cygwin finishes downloading and installing packages, you can choose
whether you want Cygwin to install shortcut icons on the desktop or in the Start
menu, and you're done. You have just installed the GCC toolchain and a complete



UNIX-like environment that runs natively on Windows. You can now go directly to
"installing the iPhone toolchain from source."

Install prerequisite tools in Linux
Good news for Gentoo Linux users: You more than likely have everything you need
already installed and can move on to "installing the iPhone toolchain from source." If
you're using Debian, Ubuntu, or any distributions based on those two, you can install
the software needed to compile the iPhone toolchain by installing the bison, flex,
build-essentials, and autoconf packages using the apt-get command:

sudo apt-get install build-essential bison flex autoconf

If you're using SuSE or Red Hat Linux or any derivative of these, such as CentOS,
you need to install autoconf, binutils, bison, gcc, g++, flex, and make with yum.

sudo yum install bison flex autoconf binutils gcc gcc-c++ make

If you're using any other Linux distribution, consult your documentation for
instructions on installing packages. You want to install the packages listed for
yum-based systems. Be aware that gcc-c++ may be called something else on your
system or may be included with GCC. When the requisite packages are installed,
you can move on to "installing the iPhone toolchain from source."

Section 5. Installing the iPhone toolchain from source

Download the toolchain source code
Why build the toolchain from source rather than downloading a binary? In a word:
licensing. The source code contains few distribution limitations, being open source.
The iPhone Dev Team based the toolchain on source code found in public
repositories and modified the code so it would compile on Linux and Cygwin.

I packaged the source code needed to compile the toolchain and provide a link in
the Resources section. Download that file and copy it to the home folder of your user
account. If you're using Cygwin, copy it to C:/cygwin/home/yourusername if you
installed Cygwin to the recommended directory. Extract the package with your
favorite gzipped tar-compatible tool, and you'll find a folder named



iphone-2.0-toolchain in your home directory.

Download the iPhone firmware from Apple
Next, obtain the iPhone firmware and add it to the toolchain source code. The Apple
code you need is the iPhone V2.0 firmware. It matters which version you obtain: You
want the iPhone OS V2.0 firmware for the first-generation iPhone, with the file name
iPhone1,1_2.0_5A347_Restore.ipsw. Refer to Resources for any of the Pwnage or
WinPwn how-to articles for the link from which to download it.

The firmware restore file is nothing more than a ZIP file. After you have the firmware,
change the extension to .zip and open it with any extraction utility for your platform.
Inside the ZIP file, you'll find a file called 018-3785-2.dmg. Copy this DMG file to the
home folder of your user account, right next to the iphone-2.0-toolchain folder from
earlier. (Windows users: We refer to the Cygwin user home folder, not your Windows
user home folder.)

Mount the operating system image
This DMG file is an actual encrypted image of the operating system partition on the
device. You must decrypt this file before you can mount the image and extract the
files you need. Luckily, the iPhone Dev Team wrote a tool to decrypt the partition
image and have obtained and published the encryption key. The next step is to
compile the image decryptor and use it to obtain a decrypted image you can mount.
Open a terminal session (in Windows, use the Cygwin shortcut, not the Windows
command line), and type the following :

gcc ~/iphone-2.0-toolchain/src/vfdecrypt.c -o ~/vfdecrypt -lcrypto

This command compiles the DMG decryptor code to an executable file named
vfdecrypt on your home folder. Now you can decrypt the DMG file.

~/vfdecrypt -i ~/018-3785-2.dmg -o ~/decrypted.dmg -k
2cfca55aabb22fde7746e6a034f738b7795458be9902726002a8341995558990f41e3755

Notice that this is two separate lines, and don't miss that backslash () at the end of
the first line. It is really a single command, but we split it into two lines with the
backslash to avoid ambiguity caused by the sheer length of the command. You may
want to paste the two lines in rather than manually enter the monstrous decryption
key.

Now that you've decrypted the DMG, mount it, and extract the contents. Linux users



can install HFS drivers and mount the DMG. It's OK if your distro only has read-only
HFS drivers, as you just want to copy some files. Windows users can use either
PowerISO or Transmac (see Resources) to open the DMG file and copy the files you
want.

Copy the required folders
Now that you've mounted or opened the decrypted DMG, you must copy three
folders from there. Assume that decrypted.dmg refers to the mounted volume,
regardless of whether you've mounted it on Linux or opened the file with Transmac
or PowerISO. The folders are decrypted.dmg/usr/lib,
decrypted.dmg/System/Library/Frameworks, and
decrypted.dmg/System/Library/PrivateFrameworks. Inside the iphone-2.0-toolchain
folder are two folders: iphone-fs/System/Library/ and iphone-fs/usr/. Copy the lib
folder to iphone-fs/usr/. Copy Frameworks and PrivateFrameworks to
iphone-fs/System/Library/. The steps below help clarify which folders you need to
copy and where to paste them:

1. Copy the lib folder to your desktop temporarily, as shown in Figure 9.
(You'll move it where it belongs shortly.)
Figure 9. Copy the lib folder to your desktop

2. Copy the Frameworks and PrivateFrameworks folders to your desktop
temporarily, as shown below.



Figure 10. Copy the Frameworks and PrivateFrameworks folders

4. Move the lib folder from your desktop into the iphone-fs/usr folder.

5. Move Frameworks and PrivateFrameworks together into the
iphone-fs//System/Library folder, as shown below.

Figure 11. lib folder goes into lib folder and the Frameworks and
PrivateFrameworks folders go into /System/Library



Compile the toolchain
After you have copied the folders to their destinations, you can unmount the DMG or
close Transmac or PowerISO. Now, open a terminal and type the following
commands:

cd ~/iphone-2.0-toolchain
./make-toolchain

These commands finally start the process of compiling the iPhone toolchain and
installing it to /usr/local on your system. The process takes about two hours on a
2.0-GHz Core 2 Duo MacBook running Microsoft Windows XP inside VMware with
640 MB of RAM allocated to the virtual machine (VM). After a wait of a few minutes,
you'll see a lot of gibberish scroll by very quickly. Don't worry: That means it's
working! The only time you should be concerned is if it stops and returns to the shell
prompt after displaying several error lines, such as those below.

Figure 12. Example error lines during the make process



If the make process stops but you don't see any error lines like those in Figure 12,
you have successfully completed the compilation process. Now that you have built
the toolchain, you're ready to get started creating your own software for the iPhone.
Next, you must install Eclipse CDT and you can explore how it lets you program C
and C++ within Eclipse.

Section 6. C and C++ development with Eclipse CDT

Install Eclipse CDT and other plug-ins
To work with Objective-C in Eclipse, you need to install two things: Eclipse CDT (to
manage makefiles and your project files) and an Eclipse editor plug-in (to give you
Objective-C syntax highlighting). The Color Editor plug-in offers syntax highlighting
for more than 140 languages.

If you don't already have Eclipse or if you don't feel like integrating all these plug-ins
into your installation of Eclipse, check out the EasyEclipse project (see Resources)
and download the EasyEclipse for the C and C++ distribution. EasyEclipse includes
the Eclipse, Eclipse CDT, Subclipse, and CVS plug-ins for managing your source
revisions, as well as several plug-ins for syntax highlighting and editor default
handing of different languages, including Objective-C. Find all of this in one simple,
single download for Windows, Linux, and Mac OS X.

If you already have Eclipse installed and have no problems adding plug-ins
manually, you can add both to your copy of Eclipse easily. To install the Color Editor,
download the plug-in and drop the JAR into your Eclipse plugins directory. To install
the Eclipse CDT, go to the Eclipse CDT download page and find the correct URL for
your version of Eclipse. In Eclipse, go to Help > Software Update > Find and
Install... > Search for new features to install > New Remote Site... and enter the
correct CDT URL for your version of Eclipse. Proceed with the installation and
restart Eclipse (see Resources for the CDT and Color Editor).

Create an Eclipse CDT project
After installing Eclipse CDT and the Color Editor, either by installing EasyEclipse or
manually adding the plug-ins, you can begin to program in C or C++.

Begin by creating a simple project for Eclipse CDT:



1. Run Eclipse. From the workbench, choose File > New > Project to bring
up the New Project window.

2. Choose General > Project, and then click Next, as shown below.
Figure 13. The New Project window

3. Name the project HelloWorld.

4. If you're using Windows, change the default location to
C:/cygwin/home/yourusername/HelloWorld, then click Finish.
The new project appears in your Project Explorer on the workbench. Click
Yes if Eclipse asks you to change to the C/C++ Perspective.

5. Create two files inside the project: Makefile and HelloWorld.c.

6. Copy the code in Listing 1 into Makefile.



Listing 1. Contents of Makefile

CC=gcc
CXX=g++
LD=$(CC)
all: HelloWorld
HelloWorld: HelloWorld.o
$(LD) $(LDFLAGS) -o $@ $^
%.o: %.c
$(CC) -c $(CFLAGS) $(CPPFLAGS) $< -o $@
clean:
rm *.o HelloWorld

7. Copy the code in Listing 2 into HelloWorld.c.
Listing 2. Contents of HelloWorld.c

#include <stdio.h>
int main(void) {
printf("Hello, world!n");
return 0;
}

Run the project
Now you can run this project. Open a terminal, and change directories to where your
project was created. From there, type the make command and check for make
errors. If you created the Makefile and HelloWorld.c files correctly, there should be
no errors, and the output produced in the console should just be two lines.

gcc -c HelloWorld.c -o HelloWorld.o
gcc -o HelloWorld HelloWorld.o

Typing make at the command line causes a program called make to interpret the
makefile, which effectively tells make that it should compile HelloWorld.c into an
object file called HelloWorld.o, link HelloWorld.o implicitly with the standard C
libraries, and produce an executable file called HelloWorld. That is the meaning of
the two commands produced as console output by make.

If you have programmed in C on UNIX before, this is probably not exciting news.
However, it's wise to verify that the simple C code and makefile work correctly by
executing the compiled and linked program with the following command:



./HelloWorld

This should produce the output "Hello, world!" in the console. Now, modify the
makefile and produce an executable file that runs on the iPhone using your new
toolchain.

Create an executable file for the iPhone
To create such an executable file, you need to overwrite the first two lines of the
makefile with the following code:

CC=/usr/local/bin/arm-apple-darwin9-gcc
CXX=/usr/local/bin/arm-apple-darwin9-g++

This command tells make to use your new iPhone toolchain, rather than your
system's toolchain you used earlier. Type make clean at the command line, and
the make program runs the clean rule at the bottom, which deletes the executable
file and object file produced earlier by the previous version of the makefile.

Now you can type make, and the new toolchain is invoked, rather than the system
toolchain. This produces a HelloWorld.o object file and HelloWorld executable file for
the iPhone. Copy it to your device to see whether your toolchain really works.
Connect your device to your WiFi network, find out its IP address in the Settings
application and send the HelloWorld executable file to it with a command like the
following (substitute IPHONE_IP with the IP address for your device).

scp HelloWorld root@IPHONE_IP:~/

Execute the program
Now, log in through ssh to execute the program. Substitute IPHONE_IP again with
your device's IP address.

ssh root@IPHONE_IP

When you've logged in, you're at the root user prompt on the iPhone.

./HelloWorld

And I bet you're disappointed by the "Killed" output in the console. What happened is



that the iPhone executable loader expects all executables to be signed. The iPhone
Dev Team did manage to remove nearly all restrictions from the iPhone OS V2.0,
but they didn't want to mess too much with the executable loader. All is not lost: You
can use the ldid command from the Linker Identity Editor utility you installed with
Cydia earlier. The ldid command produces a bogus executable signature to fool
the executable loader.

ldid -S HelloWorld
./HelloWorld

This time, it works, and the crowd roars its approval. You've successfully run code
on your device that you wrote and compiled yourself just minutes ago. You reach the
only logical conclusion a budding iPhone programmer can reach at this point: Let's
put a GUI on it!

Section 7. Compiling and building a Cocoa Touch
iPhone native application bundle
Assuming that your earlier experiments have worked, this is now where the rubber
meets the road. You will create a simple GUI application that can be launched by
touching its icon from the iPhone's touchscreen.

Modify your project makefile
You start creating the GUI application by modifying the makefile for your project and
writing some Objective-C code. To keep things simple, there's sample program code
in the iphone-2.0-toolchain folder in examples/GUI/HelloWorldiPhone. Grab the
makefile and all the other files in that folder and overwrite your Eclipse project files
with it. Delete HelloWorld.c, which you replaced with HelloWorld.m, in Objective-C.

Open the new makefile in Eclipse, shown in Listing 3, and examine it.

Listing 3. The new make file

CC=/usr/local/bin/arm-apple-darwin9-gcc
CXX=/usr/local/bin/arm-apple-darwin9-g++
LD=$(CC)
CFLAGS=-I/usr/local/lib/gcc/arm-apple-darwin9/4.2.1/include
-isysroot /usr/local/iphone-sysroot



LDFLAGS=-framework CoreFoundation -framework Foundation -framework UIKit
-lobjc -bind_at_load -isysroot /usr/local/iphone-sysroot
all: HelloWorld.app
HelloWorld.app: HelloWorld Info.plist
mkdir -p HelloWorld.app
cp Info.plist HelloWorld Default.png icon.png HelloWorld.app/
HelloWorld: HelloWorld.o HelloWorldApp.o
$(LD) $(LDFLAGS) -o $@ $^
%.o: %.m
$(CC) -c $(CFLAGS) $(CPPFLAGS) $< -o $@
clean:
rm -rf *.o HelloWorld HelloWorld.app

This makefile version adds CFLAGS and LDFLAGS, which tell the compiler and linker
where to find its include and library files. Notice the -framework linker parameters.
These tell the linker to use the frameworks you grabbed from the iPhone firmware
when preparing your iPhone toolchain.

The CC compilation rule has also changed. You're no longer compiling .c files to .o
files; rather, you now concern yourself with .m files. As you probably inferred when I
said that HelloWorld.m substitutes HelloWorld.c, .m (which stands for module) is the
source file for Objective-C programs.

There is an extra rule for something called HelloWorld.app. From looking at the
commands below the rule, it seems to be a folder in which you pile the executable
and some other files, like Info.plist. Listing 4 shows that file.

Listing 4. Info.plist

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple Computer//DTD PLIST 1.0//EN"
"http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
<key>CFBundleDevelopmentRegion</key>
<string>en</string>
<key>CFBundleExecutable</key>
<string>HelloWorld</string>
<key>CFBundleIdentifier</key>
<string>com.pjtrix.helloworld</string>
<key>CFBundleInfoDictionaryVersion</key>
<string>0.1</string>
<key>CFBundleName</key>
<string>HelloWorld</string>
<key>CFBundlePackageType</key>
<string>APPL</string>
<key>CFBundleShortVersionString</key>
<string>1.0.0</string>
<key>CFBundleSignature</key>
<string>????</string>
<key>CFBundleVersion</key>
<string>1.0</string>
</dict>
</plist>



The Info.plist file is used by the iPhone dashboard UI to figure out what the name of
the executable is and what to call the icon on the touchscreen. The Info.plist is what
turns your HelloWorld.app from just a folder with some stuff in it into an actual
iPhone application bundle. Note that Info.plist has an uppercase I in its name: The
iPhone dashboard is picky and expects the files in the bundle to be named precisely.

The bundle also contains two graphics files: Default.png and icon.png. The iPhone
dashboard uses icon.png as the icon for your program on the dashboard.
Default.png is used to display a dummy graphic on the screen while the user waits
for your executable to load and draw its UI.

Note: The icon.png file name must be all lowercase or it won't be displayed.
Likewise, Default.png must have an uppercase D or it won't be displayed at
application start-up.

Now, move on to the real meat of the application: the Objective-C code that makes it
all happen.

The Objective-C code
Begin with HelloWorld.m, which Listing 5 shows.

Listing 5. HelloWorld.m

//
// HelloWorldApp.h
// HelloWorld
//
// Created by PJ Cabrera on 08/18/2008.
// Copyright PJ Cabrera 2008. All rights reserved.
//
#import <Foundation/Foundation.h>
#import <UIKit/UIKit.h>
#import "HelloWorldApp.h"
int main(int argc, char *argv[]) {
// This autorelease pool is managed by the UI's event dispatcher. It autoreleases
// any allocated objects that fall out of scope. The iPhone's implementation of
// Objective-C 2.0 does not have garbage collection yet, but autorelease pools and
// proper release of allocated objects is still a good practice.
NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];
// This is where UI execution gets started. This instantiates the UIApplication
// subclass and UIApplicationDelegate subclass specified as string parameters.
// The name of an UIApplication subclass can be passed as both UIApplication and
// UIApplicationDelegate.
UIApplicationMain(argc, argv, @"HelloWorldApp", @"HelloWorldApp");
// Force release of the autorelease pool and all objects still allocated.
[pool release];
return 0;



}

It actually doesn't look terribly different from C. Foundation and UIKit are the two
Objective-C frameworks you will use the most in the Cocoa Touch platform.
Foundation is where strings, lists, and base object-oriented hierarchies are defined
in Cocoa and Cocoa Touch. UIKit defines quite a lot of the UI elements in an iPhone
application.

Objective-C being a superset of C, application start-up is defined pretty much the
same way as in C: It all begins at the main function. The first line of actual code
inside the main function is the first bit of Objective-C in this part of the code. There,
you're allocating and initializing an instance of an Objective-C class,
NSAutoreleasePool, used in managing memory. In Mac OS X and Objective-C
V2, it's rather like garbage collection, but it is not fully implemented in the iPhone
Objective-C environment in this version of the iPhone OS.

The second line of actual code is where the UI framework begins
executing.UIApplicationMain instantiates the classes named in the last two
string parameters and calls specific methods in these two class instances that get
the application rolling. The last bit of code gets called when the UI quits. It releases
the auto-release pool and returns to the calling environment.

Now, look at HelloWorldApp.m. This is where UIApplicationMain has moved the
action. Load the following file in Eclipse.

Listing 6. HelloWorldApp.m

//
// HelloWorldApp.m
// HelloWorld
//
// Created by PJ Cabrera on 08/18/2008.
// Copyright PJ Cabrera 2008. All rights reserved.
//
#import "HelloWorldApp.h"
@implementation HelloWorldApp
- (void)applicationDidFinishLaunching:(UIApplication *)application {
// Create a "full-screen" window to hold the UI.
window = [[UIWindow alloc] initWithContentRect:
[UIHardware fullScreenApplicationContentRect] ];
// Create a view to hold the window contents.
contentView = [[UIView alloc] initWithFrame: CGRectMake(0.0f, 0.0f, 320.0f, 480.0f)];
// Create a navigation bar for the top of the view, with two buttons.
// The buttons do not do anything in this example.
nav = [[UINavigationBar alloc] initWithFrame: CGRectMake(0.0f, 0.0f, 320.0f, 48.0f)];
[nav pushNavigationItem:[[UINavigationItem alloc] initWithTitle:@"Hello World"]];
[nav showButtonsWithLeftTitle: @"Left" rightTitle: @"Right"];
[nav setBarStyle: 0];
[contentView addSubview: nav];



// Create a text view, this is a basic text editor, with incorporated keyboard.
text = [[UITextView alloc] initWithFrame: CGRectMake(0.0f, 48.0f, 320.0f, 480.0f)];
[text setText: [[NSString alloc]
initWithString: @"Hello WorldnCocoa Touch Application"]];
[contentView addSubview: text];
// UIWindow con only hold one view, the contentView. The contentView can hold many
// navigation controllers, which control the different views in your application.
window.contentView = contentView;
// These three instructions effectively show the window.
[window orderFront: self];
[window makeKey: self];
[window _setHidden: NO];
}
- (void)dealloc {
// Release the UI elements as they were allocated
[text release];
[nav release];
[contentView release];
[window release];
// And don't forget to call the parent class dealloc
[super dealloc];
}
@end

When UIApplicationMain instantiates the HelloWorldApp class, application
execution resumes with the method called applicationDidFinishLaunching.
This is where you create the initial UI elements that get the application started from
the user's point of view. First, the function allocates a window instance and initializes
it with the screen size by calling the UIHardware class method
fullScreenApplicationContentRect.

Next, it allocates and initializes a UIView instance, which holds all the other screen
elements. There is a navigation bar with two buttons called Left and Right that in
this example are just for show. There is also a big UITextView instance with the
text "Hello, World" and "Cocoa Touch Application" on two separate lines. Finally, set
this view as the main view for your window, and set the window to receive input and
be visible.

If you run make and everything goes on without a hitch, make will have created a
HelloWorld.app folder and copied into it your Info.plist file, the icon.png, and
Default.png images, and your HelloWorld executable file. Copy that folder to your
device. As before, replace your iPhone's IP address for IPHONE_IP in the
commands below.

scp -r HelloWorld.app root@IPHONE_IP:/Applications/
ssh root@IPHONE_IP

When you log in to the device, sign the executable to make it runnable.



ldid -S /Applications/HelloWorld.app/HelloWorld

Make the application accessible
To make your application accessible from the iPhone dashboard, copy the whole
application bundle to a folder called Applications on the root of the iPhone file
system. But for it to show on the dashboard without having to reboot the device, next
run the Respring application you installed at the beginning of the tutorial, and
"respring" the dashboard.

After you've resprung the dashboard, your HelloWorld application shows among the
other icons. Touch it to launch it and behold: a navigation bar at the top with two
buttons named Left and Right, and a large text view with the phrases, "Hello World"
and "Cocoa Touch Application," on two separate lines. And the big surprise: Touch
the text area and a keyboard pops up from the bottom of the screen to let you type.
The keyboard comes "for free" when you create a text area. You didn't have to write
code to enable it.

Section 8. Summary
Using Eclipse to program applications for the iPhone is not without issues.
Objective-C is not supported by the refactoring tools, it is not easy to set up remote
debugging at the moment (even though GDB supports Objective-C debugging), and
third-party plug-ins like Color Editor have to be installed to obtain syntax highlighting
of Objective-C code. I could have spent more time setting up Eclipse CDT to allow
compilation within Eclipse rather than using make from the command line. But then,
this would have turned into an Eclipse CDT tutorial instead of a Cocoa Touch
tutorial.

With time, I hope to gather a group of developers to help me extend Objective-C
support for Eclipse CDT. If you want to learn more about Cocoa Touch development
without the iPhone SDK, using Eclipse on Windows or Linux, you can reach me
through my e-mail, and I can direct you to more information. Likewise, contact me if
you get stuck on any point of this tutorial, such as problems getting the toolchain to
build or getting the examples to run.

I hope this tutorial gets you started. Programming the iPhone can be lots of fun.



Resources
Learn
• Stop by the iPhone DevCenter to learn about the Apple iPhone SDK.
• The official iPhone Dev Team information portal, containing "accurate, useful
and concise information portal about Dev Team progress."
• For more information about jailbreaking your iPhone using WinPwn, read "How
to Unlock/Jailbreak Your 2.0 EDGE iPhone with WinPwn on Windows" from
iClarified.
• For information about jailbreaking your iPhone 3G using WinPwn, read "How to
Jailbreak Your 2.0 3G iPhone with WinPwn on Windows" from iClarified.
• For information about jailbreaking your iPhone using Pwnage, read "How to
Unlock/Jailbreak Your 2.0 EDGE iPhone with Pwnage Tool on Mac OS X" from
iClarified.
• Read the developerworks article "Develop iPhone Web applications with
Eclipse" to learn how to create iPhone Web sites using Eclipse, Aptana's
iPhone Development plug-in, and the iUi framework.
• Read the developerworks tutorial "Debug iPhone Web applications with Eclipse"
to learn how to debug Asynchronous JavaScript + XML (Ajax) Web applications
using Eclipse, Aptana's iPhone Development plug-in, Aptana's Firefox
JavaScript debugger, and Firebug.
• For information on jailbreaking your iPhone 3G using Pwnage, read "How to
Jailbreak Your 2.0 3G iPhone with Pwnage Tool on Mac OS X" from iClarified.
• Check out CNet's video explanation of jailbreaking the iPhone.
• For information about jailbreaking your iPhone using QuickPwn, read "How to
Jailbreak Your 2.0 iPhone with QuickPwn on Windows" from iClarified.
• See the official information page for Xpwn.
• Check out the "Recommended Eclipse reading list."
• Browse all the Eclipse content on developerWorks.
• New to Eclipse? Read the developerWorks article "Get started with Eclipse
Platform" to learn its origin and architecture, and how to extend Eclipse with
plug-ins.
• Expand your Eclipse skills by checking out IBM developerWorks' Eclipse project
resources.
• To listen to interesting interviews and discussions for software developers,
check out developerWorks podcasts.



• Stay current with developerWorks' Technical events and webcasts.
• Watch and learn about IBM and open source technologies and product
functions with the no-cost developerWorks On demand demos.
• Check out upcoming conferences, trade shows, webcasts, and other Events
around the world that are of interest to IBM open source developers.
• Visit the developerWorks Open source zone for extensive how-to information,
tools, and project updates to help you develop with open source technologies
and use them with IBM's products.
Get products and technologies
• Download the author's source code to iPhone development tools from his Web
site.
• Eclipse Ganymede is at the Eclipse Foundation.
• EasyEclipse is a bare-bones Eclipse distribution for experienced C and C++
developers who are new to Eclipse. EasyEclipse for C and C++ includes
Eclipse, the C Development Toolkit, Subclipse, and editor plug-ins for syntax
highlighting of several languages, including Objective-C.
• Find out more about Eclipse CDT, and the CDT URLs for each version of
Eclipse.
• Learn more about Color Editor and download the syntax coloring tool for Eclipse
for more than 140 languages.
• Find out more about Cygwin and download the precompiled UNIX tools for
Windows.
• Check out Transmac, commercial software for opening DMG images in
Windows.
• Check out the latest Eclipse technology downloads at IBM alphaWorks.
• Download Eclipse Platform and other projects from the Eclipse Foundation.
• Download IBM product evaluation versions, and get your hands on application
development tools and middleware products from DB2®, Lotus®, Rational®,
Tivoli®, and WebSphere®.
• Innovate your next open source development project with IBM trial software,
available for download or on DVD.
Discuss
• The Eclipse Platform newsgroups should be your first stop to discuss questions
regarding Eclipse. (Selecting this will launch your default Usenet news reader
application and open eclipse.platform.)



• The Eclipse newsgroups has many resources for people interested in using and
extending Eclipse.
• Participate in developerWorks blogs and get involved in the developerWorks
community.

About the author
PJ Cabrera
PJ Cabrera is a freelance software developer specializing in Ruby on Rails
e-commerce and content-management systems. His interests include Ruby on Rails
and open source scripting languages, and frameworks, Agile development practices,
mesh networks, compute clouds, XML parsing, and processing technologies,
microformats for more semantic Web content, and research into innovative uses of
Bayesian filtering and symbolic processing for improved information retrieval,
question answering, text categorization, and extraction. Check out his blog.


Write native iPhone applications using Eclipse CDT

Recommended

More Related Content

What's hot (8)

Similar to Write native iPhone applications using Eclipse CDT (20)

More from Atit Patumvan (20)

Recently uploaded (20)

Write native iPhone applications using Eclipse CDT