WWDC 2014

Very few comments on the developer features in this thread. Swift, Metal etc. I'm guessing most here are end users and not OSX or iOS devs ?
 
Very few comments on the developer features in this thread. Swift, Metal etc. I'm guessing most here are end users and not OSX or iOS devs ?
Nope not everyone...
Overall WWDC (especially this year) has become far more developer focused, as it should rightly be.

Biggest takeaways for me this year are: Swift (finally) and adaptive UIs (universal storyboards): because that's probably just what iOS would need to more easily support other form factors... So key launch of apps for Apple TV, larger iPhones, other devices, etc...
 
Last edited:
[)roi(];12814325 said:
Nope not everyone...
Overall WWDC (especially this year) has become far more developer focused, as it should rightly be.

Biggest takeaways for me this year are: Swift (finally) and adaptive UIs (universal storyboards): because that's probably just what iOS would need to more easily support other form factors... So key launch of apps for Apple TV, larger iPhones, other devices, etc...

Have you been trying out Swift ?

I like the way they have implemented closures. Turns a for loop into a function. Pretty cool, I thought.
 
Have you been trying out Swift ?

I like the way they have implemented closures. Turns a for loop into a function. Pretty cool, I thought.

Yip aspects like closures generics, tuples are definite trumps, and they've even been able to make switches fun again. Btw I take it you've been watching the WWDC session videos?

The overall language so far seems to thankfully take most of the best elements from the popular pool;

It's probably not too early to say that I'm going to love this coding in swift (over Objc, C# and Java); really happy they didn't just adopt Java and thankfully avoided garbage collection.

Really intrigued to see how fast this language evolves, including native swift support for many of the popular Objc frameworks, cocapods, etc...
 
[)roi(];12817257 said:
Yip aspects like closures generics, tuples are definite trumps, and they've even been able to make switches fun again. Btw I take it you've been watching the WWDC session videos?

The overall language so far seems to thankfully take most of the best elements from the popular pool;

It's probably not too early to say that I'm going to love this coding in swift (over Objc, C# and Java); really happy they didn't just adopt Java and thankfully avoided garbage collection.

Really intrigued to see how fast this language evolves, including native swift support for many of the popular Objc frameworks, cocapods, etc...

Watched the 3 Swift ones so far. Making it integrate seamlessly with Objc was a smart move.

Too many videos... I would like to look more at Metal. I'm intrigued how they could get it to perform so much better than OpenGL/ES on the ipad. Cant help thinking there has to be a trade-off made somewhere. Was OpenGL/ES really that bad ? Or are they able to make better use of the processor by dropping OpenGL/ES. Looks like Apple have decided to focus on proprietary solutions as they possibly believe that attempts at cross-platform compatibility has led to bloatware ?
 
Watched the 3 Swift ones so far. Making it integrate seamlessly with Objc was a smart move.

Too many videos... I would like to look more at Metal. I'm intrigued how they could get it to perform so much better than OpenGL/ES on the ipad. Cant help thinking there has to be a trade-off made somewhere. Was OpenGL/ES really that bad ? Or are they able to make better use of the processor by dropping OpenGL/ES. Looks like Apple have decided to focus on proprietary solutions as they possibly believe that attempts at cross-platform compatibility has led to bloatware ?

Yip; it's seamless Objc integration is paramount; without that it would have been a failure from the start. Quite rightly the approach let's the developers decide on the pace of transition; swift in that sense becomes both the carrot and stick.

As to Open GL ES; Apple long supported the standard so you can't fault them for trying (guess the overheads where just too much to justify keeping it).

Plus Metal was inevitable if you consider Microsoft's Directx, AMD's mantle, ...

Wonder how long before Google follows suit.
 
I'm a web developer, I've never programmed before (not even JS), but since swift is totally new I might give it a shot and see how bad or good a programmer I can potentially be. :D
 
I'm a web developer, I've never programmed before (not even JS), but since swift is totally new I might give it a shot and see how bad or good a programmer I can potentially be. :D

It's had PHP-esque similarities so you might just be at home; however the never having programming before might be your first stumbling block (unrelated to swift of course).

Either way I'd suggest you try out the swift playgrounds; so much easier to learn when you can immediately see the outcome of what you've typed (as opposed to full compilation).
 
[)roi(];12824507 said:
It's had PHP-esque similarities so you might just be at home; however the never having programming before might be your first stumbling block (unrelated to swift of course).

Either way I'd suggest you try out the swift playgrounds; so much easier to learn when you can immediately see the outcome of what you've typed (as opposed to full compilation).

Actually from the first glance it seems borrow quite a bit from Go.
 
In effect it's borrowed many of the best attributes from all the popular bunch; Go probably included, but certainly not limited to.

As I said thankfully they avoided Java (& C#, etc...), and most of all the worst aspects of these languages: garbage collection.

Btw If Google really was committed with Go, then surely they'd kick Java to the curb and center their products behind this single language. Without that level of commitment Go will sadly be always relegated to the non starter lane.
 
[)roi(];12825729 said:
In effect it's borrowed many of the best attributes from all the popular bunch; Go probably included, but certainly not limited to.

As I said thankfully they avoided Java (& C#, etc...), and most of all the worst aspects of these languages: garbage collection.

Btw If Google really was committed with Go, then surely they'd kick Java to the curb and center their products behind this single language. Without that level of commitment Go will sadly be always relegated to the non starter lane.

Not sure what Googles commitment is, but there is a rather large open source following in it. I've written a few backend services with it, but its not the most elegant language i'll admit. Was more used to learn and play than to actually get full steam in to.

Never had an issue with the C# garbage collector, so you would need to elaborate on your dislike towards it.
 
Not sure what Googles commitment is, but there is a rather large open source following in it. I've written a few backend services with it, but its not the most elegant language i'll admit. Was more used to learn and play than to actually get full steam in to.

Never had an issue with the C# garbage collector, so you would need to elaborate on your dislike towards it.

That's unfortunately the case with many Google projects; after the fanfare, most end up being relegated to Google's Graveyard.

My disdain for garbage collection is two fold: memory footprint and cyclic / non deterministic performance hits (occasionally referred to as pauses in the mobile space) -- as example: there's a reason why Android top end devices require more RAM and larger CPUs.

ARC (which they've again adopted) provides a far more deterministic control of the Apps memory footprint and the destruction of objects than GC; if any, the only downside of ARC is that it cannot automatically handle retain cycles, but for that they've again included the weak type attributes.

Btw I don't dislike either Java or C#, as language constructs I think they're superb; however the lack of control afforded from GC perturbs me to no end; I often find it's difficult to tweak App performance around the non deterministic behavior of GC.
 
Last edited:
[)roi(];12826399 said:
Btw I don't dislike either Java or C#, as language constructs I think they're superb; however the lack of control afforded from GC perturbs me to no end; I often find it's difficult to tweak App performance around the non deterministic behavior of GC.

Not sure about the Java GC, but its highly doubtful that you can tweak the app better than the C# GC can. Also tampering with C# GC causes the algorithms to go a bit off the path. That being said, there are ways to change the way it collects and the way it handles finalization, but I've only ever had to chance it once on a massive project. We had to import a CSV file 2gb in size we handle to tell the GC to hold back on collection,otherwise the gen2 became massive.
 
Not sure about the Java GC, but its highly doubtful that you can tweak the app better than the C# GC can. Also tampering with C# GC causes the algorithms to go a bit off the path. That being said, there are ways to change the way it collects and the way it handles finalization, but I've only ever had to chance it once on a massive project. We had to import a CSV file 2gb in size we handle to tell the GC to hold back on collection,otherwise the gen2 became massive.

No value in debating Microsoft's GC benefits over Java.

Re .NET's GC, AFAIK the design for the GC was originally steered much by the need to accommodate semantic equivalence with VB6 and hence a number of compromises had to be made; ARC although originally considered was abandoned primary because of the limitations to adequately accommodate VB6 and to provide for a common language independent COM interop; http://blogs.msdn.com/b/brada/archive/2005/02/11/371015.aspx

Yet since the introduction of Windows Mobile: MS has had to come grips with the limitations of many of these past decisions, so whilst they might have made more advances with reducing GC's footprint and resource reclamation performance impact; GC still unfortunately comes with an overhead that does not plague a well built ARC App.

Your example btw appears to refer to desktop or server apps, and if that assumption is correct then that could explain why you are not bothered by GC. Basically the challenges with the GC model are only really evident with limited resource devices (CPU, GPU, RAM, ...); I.e. building apps for mobile apps.
 
Last edited:
Just ended up updating via the App Store, no issues so far.
 
Updating now.

Hopefully sorts out the few issues I've had with battery related random kill shots.
 
So far noticed wifi icon I can now tell far better when it's not connected.
Photo Booth has appeared
Safari lot more responsive
Boot time is quicker
Battery shows correct usage
Preview and TextEdit folders showing up in iCloud folder
Spotlight opens every time now
 
Last edited:
xcode beta 2 is a bit buggy. Running the iPhone 5 simulator failed with a null alert. Loaded the same code in the first beta and the compiler caught the error.

loaded this tutorial http://www.raywenderlich.com/74904/swift-tutorial-part-2-simple-ios-app and the compiler crashed..

Some betas are always better than others. I looked at the release note and decided to avoid this beta as it did not address enough of the known swift bugs.

I suggest you monitor both the release note and the dev forums to see if a new beta is worth downloading.
 
Top
Sign up to the MyBroadband newsletter
X