iOS 5.0 MKMapView (MapKit MapView) Edge Cases

The following are notes from my adventures implementing complex custom callouts in an MKMapView on iOS 5.0.

• The selected MKAnnotationView is brought to the front of the group.
• When the selected MKAnnotationView is deselected, other undefined annotations may appear in front.
• Calling -[MKMapView selectAnnotation:animated:] on an annotation with a disabled MKAnnotationView will not select the MKAnnotationView. 
• The MKMapView will not call -[MKMapViewDelegate mapView:didSelectAnnotationView:].
• If MKAnnotationViews overlap and one is selected, touching the selected one will select an unselected one.  Since custom callouts must be implemented as ordinary MKAnnotationViews, you must disable all MKAnnotationViews behind a callout MKAnnotationView if you want the user to be able to interact with UIControls inside your callout MKAnnotationView
• Any UIControls in your MKAnnotationView must exist within your MKAnnotationView's frame. Otherwise, when the user attempts to touch them, the map will think the user touched outside the MKAnnotationView and deselect your MKAnnotationView.
• You may change the frame of your MKAnnotationView after the MKMapView calls -[MKMapViewDelegate mapView:viewForAnnotation:] and -[MKMapViewDelegate mapView:didAddAnnotationViews:].  However, you can't change the -[MKAnnotationView centerOffset].  So, if you change the frame size, it needs to grow equally in all directions.
• When you don't grow the MKAnnotationView's frame evenly, it will appear to look fine at first, but when you zoom the map, it will snap to its old centerOffset.

Consolidated VMware VMDKs

I just consolidated a 25GB VMware virtual machine into 10GB by consolidating unused snapshot virtual disks. This vmware faq explains how to do it extremely well.

Edited value not of required format!

I'm using the xcode predicate editor to create a core data fetched property, and Apple's developer tools have proved inherently unusable yet again.

For a little back-story, I have some reference data, and my user can mark certain pieces of the reference data as favorites. I cannot store my favorites data in the same NSPersistentStore as my reference data. If I did, I couldn't updated my reference data file without migrating my favorites data out. Thus, I'm splitting my favorites data and reference data into separate NSPersistentStores. According to Apple's CoreData Programming Guide Cross-Store Relationships section I cannot have a relationship that spans two persistent stores. I need to use a fetched property.

Now you're caught up. When you create a fetched property, you have to create a predicate to use to find the entity the fetched property represents. This seems straightforward enough, but I kept getting this cryptic message "Edited value not of required format!" After some googling, I found the answer.

Basically, the editor defaults to expecting a constant value rather than a variable value. To switch to a variable, right-click in the empty space between the text field and the plus/minus buttons. Of course, a drop-down menu would be much better, but I never accused apple of having good tools.

Bootstrapping Buckminster & Subclipse with P2 Director

On my current project at work, we are changing from using PDE-Build to create our Eclipse RCP product to using Buckminster. I found a very valuable tutorial to get me started, but I kept getting exceptions like these:


ERROR: Could not instantiate provider org.tigris.subversion.subclipse.core.svnnature for project com.foo.
org.eclipse.team.core.TeamException: Could not instantiate provider org.tigris.subversion.subclipse.core.svnnature for project com.asolutions.soaf.validation.ws.documentation.test.
at org.eclipse.team.core.RepositoryProvider.mapNewProvider(RepositoryProvider.java:165)
at org.eclipse.team.core.RepositoryProvider.mapExistingProvider(RepositoryProvider.java:235)
at org.eclipse.team.core.RepositoryProvider.getProvider(RepositoryProvider.java:507)
at org.eclipse.team.internal.core.TeamHookDispatcher.getProvider(TeamHookDispatcher.java:97)
at org.eclipse.team.internal.core.TeamHookDispatcher.getRuleFactory(TeamHookDispatcher.java:105)
at org.eclipse.core.internal.resources.Rules.factoryFor(Rules.java:92)
at org.eclipse.core.internal.resources.Rules.modifyRule(Rules.java:144)
at org.eclipse.core.internal.resources.Project.touch(Project.java:1191)
at org.eclipse.jdt.internal.core.SetContainerOperation.executeOperation(SetContainerOperation.java:115)
at org.eclipse.jdt.internal.core.JavaModelOperation.run(JavaModelOperation.java:728)
at org.eclipse.core.internal.resources.Workspace.run(Workspace.java:1975)
at org.eclipse.jdt.internal.core.JavaModelOperation.runOperation(JavaModelOperation.java:793)
at org.eclipse.jdt.core.JavaCore.setClasspathContainer(JavaCore.java:4842)
at org.eclipse.pde.internal.core.PluginModelManager$UpdateClasspathsJob.run(PluginModelManager.java:57)
at org.eclipse.core.internal.jobs.Worker.run(Worker.java:54)



The solution is to install subclipse, but doing so within Buckminster is painful because you can only specify a single repository location from the command line. Subclipse has a lot of dependencies that already exist in a default eclipse install, but are missing from buckminster-headless, so I needed to specify multiple repositories.

p2 director to the rescue!

In the tutorial, there is a section on bootstrapping buckminster with p2 director. The tutorial loads just the buckminster command-line application from p2 and then uses the buckminster command-line application to load more features into itself. This is convenient when all features and dependencies exist within a single repository, but as I said above, this is not my situation.

Assuming you are using bash and have p2 director installed, here is the command:


director/director\
-r http://download.eclipse.org/tools/buckminster/headless-3.6,http://download.eclipse.org/releases/helios/,http://subclipse.tigris.org/update_1.6.x\
-d `pwd`/buckminster-headless-3.6/\
-p Buckminster\
-i org.eclipse.buckminster.cmdline.product\
-i org.eclipse.buckminster.core.headless.feature.feature.group\
-i org.eclipse.buckminster.pde.headless.feature.feature.group\
-i org.tigris.subversion.subclipse.feature.group\
-i org.tigris.subversion.clientadapter.svnkit.feature.feature.group


5 Questions Response

In response to Ben Lee's Five Questions:

What are you currently hacking on?

What are you currently getting better at?

What do you do when your computer is asleep?

Describe that ‘big fish’ project that’s been stewing in your brain.

What are you gonna post about this coming week?

1. JAXB, JAX-RS, Google App Engine for a real-time rating app I'm building with Mike Gaffney.
2. C/Objective-C. At work, I'm building a locator iPhone app for a fortune 500 company, and I'm running into all kinds of low-level issues that Java abstracts away from me. I can't help but feel like a badass when I see registers in my debugger.
   
3. Hurling, play with my wife and son.
4. In conjunction with #1, I'm building a real-time rating web service in preparation for a hackfest for my local mobile development group.
5. Probably nothing. I usually just use this blog as a public place to store answers to annoying technical problems. Entries should probably be cross-posted on stackoverflow as questions/answers.
   

Disabling Firefox Preview in Aptana

Eclipse uses xulrunner to embed mozilla firefox within SWT applications. Unfortunately, xulrunner does not work with 64 bit cocoa (cocoa-x64). I'm trying to do Palm Pre development, so I don't need to preview my html in firefox. In fact, previewing in Safari is probably better anyway since, like the pre browser, it is webkit-based.

To disable Firefox preview, and stop errors like this:

(Build 1.2.7.024747) [ERROR] Unable to create Firefox browser control
No more handles (java.lang.UnsatisfiedLinkError: Could not load SWT library. Reasons:
 no swt-xulrunner-cocoa-3624 in java.library.path
 no swt-xulrunner-cocoa in java.library.path
 Can't load library: /var/folders/Az/Az+0r2ubHtm3wMGMatRuQE+++TI/-Tmp-/swtlib-64/libswt-xulrunner-cocoa-3624.jnilib
 Can't load library: /var/folders/Az/Az+0r2ubHtm3wMGMatRuQE+++TI/-Tmp-/swtlib-64/libswt-xulrunner-cocoa.jnilib
)

and this:

(Build 1.2.7.024747) [ERROR] (Build 1.2.7.024747) [ERROR] Unable to create Firefox browser control
No more handles [Could not detect registered XULRunner to use]

Go to the Window->Preferences->Aptana->Editors->HTML->Preview preference page and uncheck Firefox.

If you want to use xulrunner with eclipse out of the box, you must use eclipse carbon 32-bit. If you want to be a little more adventurous, an aptana bug report claims that xulrunner is a universal binary and runs on carbon and cocoa 32-bit. I haven't tried to verify this.

XCode Beach Ball Lock Up While Running Unit Tests

I downloaded the iphone unit test calculator example project, iPhoneUnitTests. I followed the included directions, and my XCode locked up with a beach ball. I'm running OS X 10.6 and XCode 3.2. According to this article, there is a problem with XCode's regex parsing of the test completion dates for GMT-5:00. Since I'm in Central Daylight time, I am affected. The simple solution is to switch your time zone to Pacific.