Steve Jobs replies: iTunes Cocoa

[ipedro]

Wow... I guess Steve is comfortable delegating a lot of his previous work because he seems to be answering a lot of customer questions himself!

Reading into his response, it appears that he's suggesting that iTunes for iPad, iPod and iPhone are the future of the app and they'll be ported to Mac and PC.

 

[elppa]

There's one big problem:

[1] Cocoa iTunes would mean cocoa for Windows — unless Apple pull a Quicktime X and leave the Windows users with the old version. Not something I think they'd contemplate with a real flagship product like iTunes.

Also interesting is that Steve mentions the iPhone, iPod touch and iPad. On these devices there are separate apps for:
[1] Music and podcasts
[2] Video
[3] iTunes Store (Music)
[4] iTunes Store (Apps)

This could simply be because of RAM usage. But I think on the desktop Apple could do something similar, here's how I would envisage it working:

[1] iTunes sync agent (Background Daemon (Mac), Background Service (PC))
- Runs in the background all the time (much like the current iTunesHelper)
- Unlike the current iTunes helper if you plugin an iDevice, it will kick start the sync process using your default settings.
- In the future Apple can implement wifi sync to do this over the air
- So for basic syncing no app ever needs to be launched and run, the agent handles it all.
- It will have a basic UI to display the progress and let you know when it's done:
PC: Tray Icon to show status
Mac: HUD/menu bar/dock icon to show status (I kind of think the menu bar is full enough!)
- One of the options will be to launch the Device Manager.

[2] Apple Device Manager (64bit Cocoa Application)
- This is a standalone app which has a UI that manages your sync settings (much like when you click on a device in iTunes today). Music, Podcats, iTunes U, Books, Photos, Contacts etc. You get the idea.
- Folder monitoring, so you can sync without iTunes Music needed. However most users will be encouraged to download and install iTunes Music for album art, lyrics etc.

[3] iTunes Music (64bit Cocoa Application)
- Based on the iPad client, this will be the cut down Music player people want. It will just do Music, podcasts and Genius
- Integration with lala screaming subscription service
- Optional social network integration, to play catch up with services like Spotify

[4] iTunes Cinema (64bit Cocoa Application)
- TV Shows
- Movies
- Video Podcasts
- DVD
- BluRay (as rumoured) — cause let's face it, the BluRay players on PCs are clunky and rubbish.

[4] iBooks and iBookstore (64bit Cocoa Application)
- A port of the current iPad application, with some changes for the Desktop

[5] iTunes Store (64bit Cocoa Application)
- Same as today: Music, Movies, TV shows, App Store, Podcasts, Audio Books, iTunes U
- May want to think about either: splitting the Audio Books off to the iBooks app -or- brining the iBookstore into the iTunes Store. Unfortunately that second option would mean sacrificing the cool rotating bookshelf UI.

[5.1] iTunes Web Store
- HTML 5 application for users wishes to purchase content online. For users on a public terminal/corporate network with the iTunes Store app installed. Also could be used by Linux users who can't find the songs they want on Amazon/Ubuntu One Store.

So in summary:
Windows Users w/o an iDevice:
- The whole shebang, or just the components you'd like (Most will choose iTunes Cinema and iTunes Music because they will be best in class and free).

Windows Users with an iDevice:
- iTunes sync agent and Apple Device Manager (Again most will choose to combine with iTunes Cinema and iTunes Music). This will please some, but most regular users won't pick this route.

Mac Users w/o an iDevice:
- You get iTunes Cinema and iTunes Music with every Mac. You no longer get a separate DVD player. If your territory supports it you get iBooks + iBookstore and iTunes Store pre-installed

Mac Users with an iDevice:
- Plug an iDevice in and you will be prompted to download and install the Apple Device Manager (much like rosetta and Quicktime 7 today).

I'm not sure this would keep everyone happy, but if the execution of the native programs was good then it would keep most people happy for most of the time.

 

[ReggaeFire]

There's one big problem:

[1] Cocoa iTunes would mean cocoa for Windows — unless Apple pull a Quicktime X and leave the Windows users with the old version. Not something I think they'd contemplate with a real flagship product like iTunes.

They've already done a port of a Cocoa application to Windows - Safari.

I think Steve was being subtle in his response. The iPhone OS requires Cocoa. The iPhone OS has iTunes (well its component parts). I read this as saying "the groundwork has already been done, Cocoa iTunes is coming".

 

[ipedro]

^ I agree. I think we'll see a brand new iTunes written from the ground up in Cocoa for iPad/iPhone that will be ported to Mac and PC.

 

[JackTheTripper]

Translation: The IMPORTANT products are Cocoa. Who cares about the desktop/laptop.

 

[ipedro]

How often does Steve actually answer customer questions? Was it just a fluke that he answered my very first question just a few hours after I sent it? He replied at 11:30pm, his time.

 

[Phil A.]

How often does Steve actually answer customer questions? Was it just a fluke that he answered my very first question just a few hours after I sent it? He replied at 11:30pm, his time.

I'd say he answered your e-mail but not your question - you asked about iTunes on OS X and he just told you that the iPod apps on the iDevices are written in Cocoa (not exactly a surprise seeing as there is no Carbon Framework on the iPhone OS).

 

[ipedro]

^ I'm inclined to agree with ReggaeFire that Steve answered my question in a subtle way, suggesting that the groundwork for iTunes in Cocoa is done and will make its way to Mac.

 

[elppa]

You could also read it as: we aren't being lazy, we have iTunes in cocoa on our priorities (iPhone and iPad) and there are only so many developers and so many hours in a day.

 

[gnasher729]

^ I agree. I think we'll see a brand new iTunes written from the ground up in Cocoa for iPad/iPhone that will be ported to Mac and PC.

Please explain to me: How would it benefit you if iTunes were written in Cocoa? And how would you benefit if a brand new iTunes application were written, which means that for a long long time until this job is done there will be no progress at all in the existing iTunes application?

Translation: The IMPORTANT products are Cocoa. Who cares about the desktop/laptop.

New code is written in Cocoa, whether it is important or not, because developing code in Cocoa is more efficient. But you know what is ten times more efficient than writing your code in Cocoa? Leaving existing Carbon code unchanged. Zero effort.

 

[Phil A.]

^ I'm inclined to agree with ReggaeFire that Steve answered my question in a subtle way, suggesting that the groundwork for iTunes in Cocoa is done and will make its way to Mac.

I'm not sure how much of what's in the iPod app would be re-usable in iTunes: They really are completely different apps.
I can't wait for the day when iTunes is re-written from the ground up as a Cocoa app (it certainly needs it), and when I can use coverflow, etc on connected libraries (the fact they haven't added front-row type access to shared libraries is a constant source of annoyance to me!), but I'm not sure this really gives us any more information (or hope!)

 

[ipedro]

Please explain to me: How would it benefit you if iTunes were written in Cocoa? And how would you benefit if a brand new iTunes application were written, which means that for a long long time until this job is done there will be no progress at all in the existing iTunes application?

It's not Cocoa itself but the fact that iTunes needs to be re-written from the ground up. Cocoa is the driving force behind that goal. iTunes has been added on to, layer after layer, for 9 generations already so there's a lot of junk and a lot of compromises and fixes put in.

There's no doubting that iTunes has become slow and inefficient. A Cocoa re-write in 64bit code will certainly enable it to deal with massive library sizes it was never originally intended to deal with.

 

[Policar]

Please explain to me: How would it benefit you if iTunes were written in Cocoa? And how would you benefit if a brand new iTunes application were written, which means that for a long long time until this job is done there will be no progress at all in the existing iTunes application?

iTunes in cocoa would be faster, more efficient in terms of system resource consumption, and more stable (not just because it's written in cocoa but also because it would be a total rewrite by virtue of that). It's not like iTunes is the biggest resource hog (although it's surprisingly bad given what it does) but it's a program people keep open in the background enough that a rewrite would benefit everyday performance-even improve battery life in some cases--for a huge audience. A lot of these legacy carbon apps (After Effects CS4 and FCP in particular) are just horribly slow and unstable and iTunes 9 suffers from similar problems on a much smaller scale.

This is kind of unrelated but I recently saw Final Cut Studio 3 crash due to an issue with the "KGCore Plugin." What's KG? "Key Grip"--the codename for Macromedia's video editor before Apple bought it and turned it into Final Cut, which means the code that crashed dates back to 1996-1997. Apple needs to convert their flagship apps to cocoa and the complaints they've leveled at Adobe are hypocritical to the extent that they still haven't.

 

[gnasher729]

iTunes in cocoa would be faster

Says who? Lots of basic classes in Cocoa are toll-free bridged from CoreFoundation, and the actual implementation is in CoreFoundation - using them from Cocoa just adds another level if indirection.

more efficient in terms of system resource consumption

Says who? Have you ever compared storage requirements for an NSString and a CFString? Any difference?

and more stable (not just because it's written in cocoa but also because it would be a total rewrite by virtue of that).

You are joking, right? A total rewrite is more stable? Please tell me where Apple hires these super human programmers who write completely bug-free new code.

It's not like iTunes is the biggest resource hog (although it's surprisingly bad given what it does) but it's a program people keep open in the background enough that a rewrite would benefit everyday performance-even improve battery life in some cases--for a huge audience.

If your assumptions were right, then yes. But they're not.

It's not Cocoa itself but the fact that iTunes needs to be re-written from the ground up.

That is a fact? I take it you have seen the iTunes source code, or what would you base your judgement on?

 

[Policar]

Okay, whatever. No one knows for sure that a cocoa rewrite would be better because it doesn't exist; point taken.

But iTunes is a pretty serious resource hog given what it does, and, from what I understand, Apple's other cocoa rewrites (Safari, Logic, Finder, Quicktime, etc.) improved, sometimes very substantially, upon the earlier code in terms of speed and resource consumption.

The core issue here isn't even speed but Apple being hypocritical in calling out Adobe, but all the same I suppose you're right: there's no way to know for sure a complete rewrite would be worth it, let alone objectively better, and there's no pressing need for it. But just because you seem to know more about programming than other people in this thread (you certainly know much more than I do) doesn't change that a rewrite of iTunes would most likely be faster and more efficient and that Steve's open letter is at least mildly hypocritical.

 

[Brien]

Bumping this because 10.0 has come, and it's still carbon.

So, what do people think now? iTunes 11? 12? 49?

 

[rorschach]

Bumping this because 10.0 has come, and it's still carbon.

So, what do people think now? iTunes 11? 12? 49?

Honestly, I think Apple doesn't see iTunes as the main app anymore, but rather just a storage bin for all your media that you use to sync content to your iDevices.

I'm guessing that Apple thinks (or more likely, they've done actual research on this) that most people just use iTunes to get their music/movies/TV shows onto their iPod/iPhone/iPad/Apple TV and then listen or watch from there...even if you connect your iPod to a stereo.

IMO, iTunes is a dog and is really slow with a LOT of content (which I have). Beachballs, jerky scrolling, etc. It could definitely use improvements. Sadly I don't see it happening for the reason I mentioned above.

 

[Brien]

Yeah, but if that's the case, I could see a case being made for moving the media management/syncing into the Finder itself.

 

[elppa]

Yeah, but if that's the case, I could see a case being made for moving the media management/syncing into the Finder itself.

And what should they do on Windows?

 

[rorschach]

Yeah, but if that's the case, I could see a case being made for moving the media management/syncing into the Finder itself.

Or iSync. And then they could just release it for Windows.

 

[Tom Cat]

That is a fact? I take it you have seen the iTunes source code, or what would you base your judgement on?

Calm down. I agree that iTunes is a good piece of software, but after 9 years, a rewrite is definitely due.

Remember that many fans of Apple (and, specifically, Mac OS X) scorned Windows because Windows was considered 'bloat.' Wouldn't Apple want to keep their flagship applications 'un-bloated'?

Furthermore, here's a bit of proof to show you why we awfully need iTunes to be rewritten. Try doing this in your own library: select all songs from an album, delete their artworks, and add them again - using the lower left panel to speed things up if you like. Notice this?

It constantly gets locked up when you try to perform several tasks. Restoring your iPod, for example, locks up iTunes. (While, of course, one should not mess with the restoring process, I sometimes just want to select a different song.)

That's a sure sign that an app is in need of a rewrite.