Apple Unveils Extra Leopard-isms To Developers

devilsecret writes to point out that some of the new Apple capabilities for developers on Leopard have been unveiled. The most interesting parts appear to be the opening of more of iLife to other programs, and the inclusion of Ruby on Rails.

Comments

No comments, but Slashdotted?

[yroJJory]

Wow.

Re:No comments, but Slashdotted?

[acaeti]

I felt a great disturbance in the Internets, as if millions of voices suddenly cried out in terror and were suddenly silenced.

Re:No comments, but Slashdotted?

[__aaclcg7560]

I was more impressed by the short summary and it somewhat made sense. I guess there's a first time for everything!

Re:No comments, but Slashdotted?

[gumbright]

It was on Digg (that brain damaged half child of Slashdot) too.

Re:No comments, but Slashdotted?

[Overly Critical Guy]

It should be noted that one of the reasons Apple is shipping Ruby on Rails is that Rails was developed on Macs, all the major Rails developers use Macs, and the preferred editor is a Mac application called Textmate.

Re:No comments, but Slashdotted?

[avronius]

Shouldn't that be, "I felt a great disturbance in the tube..."

RoR bandwagon?

[not already in use]

Anyone think Apple jumped on the RoR bandwagon a little too soon? The whole "movement" has lost a lot of steam and it doesn't appear to be the silver bullet everyone originally thought it was. Also, is this just part of the developer suite, or is RoR support somehow built in to the OEM OS?

Re:RoR bandwagon?

[nine-times]

Anyone think Apple jumped on the RoR bandwagon a little too soon? The whole "movement" has lost a lot of steam and it doesn't appear to be the silver bullet everyone originally thought it was.

Wouldn't that mean they jumped on the bandwagon a little too late?

Anyway, RoR isn't the solution to all programming problems, but it seems to have enough steam that it's going to stick around. OSX comes with Apache, and it's not hard to get PHP, MySQL, or whatever else installed. There's a ruby interpreter in the OS already, and a lot of the prominent people in the RoR community are OSX users.

I can't RTFA to know what they've actually done, but why wouldn't they support RoR? In spite of not finding the meaning of life, solving world hunger, or finding hot women for me, it's a pretty good tool. Something can be useful without solving every single problem, you know.

Re:RoR bandwagon?

[mblase]

Anyone think Apple jumped on the RoR bandwagon a little too soon? The whole "movement" has lost a lot of steam and it doesn't appear to be the silver bullet everyone originally thought it was.

So, did they jump on too soon, or too late?

Re:RoR bandwagon?

[Anonymous Coward]

I believe Perl, PHP and Python are included along with GCC and others...I just checked the terminal and without having installed the developer tools on my new(ish) MacBookPro I see ruby 1.8.2 (2004-12-25 build) is already installed.

Apache is installed by default.

It isn't an interface to anything required by the OS, but as Apple is a *nix and these scripting languages are pretty damn small, it doesn't take much to have them installed, so why not.

Re:RoR bandwagon?

[suv4x4]

Anyone think Apple jumped on the RoR bandwagon a little too soon? The whole "movement" has lost a lot of steam and it doesn't appear to be the silver bullet everyone originally thought it was. Also, is this just part of the developer suite, or is RoR support somehow built in to the OEM OS?

Do you have concrete links and facts to support your observation? I'm working (for a few months now) on a slightly RoR-like extension to PHP5 (and later), and I've also noticed some weaknesses of RoR which I'm trying to avoid in my designs.

It'll be interesting to point out what the RoR community is griping about.

Re:RoR bandwagon?

[mblase]

Do you have concrete links and facts to support your observation?

This is Slashdot. What do YOU think?

Re:RoR bandwagon?

[el_womble]

It might not be all things to all web developers and its far more of a paradigm shift than a casual glance at the syntax would have you believe. As a Java developer I expected to jump into Rails and be competent almost instantly. The reality was that there was a very steep learning curve when you wanted to stretch the framework away from a SOA portal or yet another blog, but IMHO it was well worth it. There are many things in Rails that make you start to wonder why J2EE wasn't designed in a similar vain. Perhapps the most gauling aspect of Rails is that it does impose a lot of 'good practice' on you: REST is pretty much compulsory; as is having a single primary key and it encourages you not to use session variables. The way I think of it is that Rails answers the questions that J2EE developers are asking 5 years on.

Re:RoR bandwagon?

[morgan_greywolf]

Do you have concrete links and facts to support your observation?

This is Slashdot. What do YOU think?

Sure. If he's anything like me, he probably just doesn't remember what they were. He could find them with Google, but doesn't feel like it and suggests you go Google it for yourself.

Re:RoR bandwagon?

[Anonymous Coward]

"I believe Perl, PHP and Python are included along with GCC and others..."

$ uname -srp
Darwin 8.8.0 powerpc
$ whereis python
/usr/bin/python
$ whereis perl
/usr/bin/perl
$ whereis ruby
/usr/bin/ruby
$ whereis php
/usr/bin/php
$ whereis apachectl
/usr/sbin/apachectl
$ whereis nfsd
/sbin/nfsd
$ whereis gcc
/usr/bin/gcc

OSX gives you many OSS tools out-of-the box which is more than can be said for the Redmond camp.

Re:RoR bandwagon?

[not already in use]

Too soon -- as in, got caught up in the hype before it became a time tested, proven and successful development platform. But that's really beside the point. From what I've seen with rails, read on forums, etc, there are a lot of ways to do something the wrong way. What you may initially learn in some tutorial or book doesn't necessarily transcend to the real world, with certain features and methods not scaling well when brought into production. I've seen this in other similar frameworks, like Prado for PHP. Yes, initial development time is cut down significantly, but then you spend a lot of your time working on caching techniques and scaling in order to improve performance, which is often gonna be the case when you have a framework doing so much of the work for you.

Re:RoR bandwagon?

[nuzak]

Sorry, but requiring integer surrogate keys and failing to support composite keys is not "good practice", it's a design limitation of ActiveRecord.

There are many things in Rails that make you start to wonder why J2EE wasn't designed in a similar vain.

Because J2EE was designed for people that demanded things like clustering and distributed transactions. I'm not saying J2EE's initial iterations weren't a complete clusterfuck, but they did have different design goals.

I have an app that requires submitting content snippets in about eight different languages, including Japanese and Chinese. It requires multi-step "wizard" workflows with their own transactions, as well as optimistic locking that catches stale edits. Ruby doesn't even support Unicode, and the hacks to make it do are grossly inadequate. The transactional features I would most likely have to laboriously recreate by hand. Rails and almost every other MVC "framework" out there were just complete non-starters.

Re:RoR bandwagon?

[shmlco]

Now if it only had a decent high-performance cross-platform engine beneath the hood...

Re:RoR bandwagon?

[RAMMS EIN]

``There are many things in Rails that make you start to wonder why J2EE wasn't designed in a similar vain.''

I think part of the answer may be that Java and Ruby are very different languages. Java is fairly static, rigid and verbose, with lots of redundancy in the code. Programs are typically designed top-down, along language features, so that the program is made to fit the language.

Ruby is almost the opposite: it's very dynamic, very succinct, and there's virtually no redundancy in the code. Programs can be built bottom-up, with the design evolving with the program, and the language being made to fit the program. Rails is a prime example of that.

Re:RoR bandwagon?

[Overly Critical Guy]

Anyone think Apple jumped on the RoR bandwagon a little too soon? The whole "movement" has lost a lot of steam and it doesn't appear to be the silver bullet everyone originally thought it was.

Normally, I demand facts and evidence to back up assertions, but because you wrote it so eloquently and described something shiny, I think I'm going to let it slide this time and simply believe everything you say.

Re:RoR bandwagon?

[jcr]

Apple's not on the RoR "Bandwagon". It's just another of the the widely-used UNIX apps that they include on the OS X disks, just like Python, Perl, Apache, BASH, PostFix, etc, etc.

-jcr

Re:RoR bandwagon?

[larkost]

A couple of notes:

RoR is simply going to be included. Nothing more at the moment on that count. Apple already has a easy-to-use database solution for Objective-C applications in CoreData (though I wish they would make it multi-user/computer capable).

And PHP is already included in the OS, you just have to turn it on. This is somewhat good from a security standpoint, but I wish they would put in a button to turn it on (next to the one to turn on Apache).

Re:RoR bandwagon?

[Fred_A]

But since they name all they releases after large cats, they had to go "RoR!"

Ehm. Ok, I know the way out...

Re:RoR bandwagon?

[Anonymous Coward]

I find it very odd that CoreData can't do multi-user stuff, they already speak SQL to interface with SQLite so I don't see why they couldn't throw in an option for ODBC or something. Here's hoping 10.5 has this.

Re:RoR bandwagon?

[grrrgrrr]

I think the inclusion of RoR is giving apple some nice publicity. As for developing web applications on a os-x server I think good old webobjects still beats RoR for easy and fast development..

Re:RoR bandwagon?

[MicrosoftRepresentit]

That is sort of what EOF/WebObjects is for, which is also free as of some time earlier this year. I think its still all Java though, heres hoping Leopard brings ObjC support

Re:RoR bandwagon?

[joshorion]

And PHP is already included in the OS, you just have to turn it on.

Minor correction, but, there's nothing you need to do in order to 'turn PHP on' with OS X. Pop open the command line and type 'php -v,' and, like any application, you'll notice it's always there for you to use.

It's also configured with Apache by default, so you just have to start apache and you're able to serve PHP documents. It's PHP4, though.

Re:RoR bandwagon?

[overunderunderdone]

but I wish they would put in a button to turn it on (next to the one to turn on Apache).

I'm glad they don't. If someone can't figure out how to turn it on without a simple button shouldn't be using it.

Re:RoR bandwagon?

[tacocat]

There are some interesting features of RoR that do represent a shift. But there are some disadvantages to RoR as well.

It is important that people start rethinking the codebase for web sites and web applications to split the backend and frontend a little cleaner. This makes it easier for teams to develop and easier to debug. RoR makes this pretty much mandatory.

When it comes to their database practices - primary keys and such -- they pretty much suck. The approach they have here is one of over simplified database structures which means you can definitely create bad data and violate data integrity rules all over the place without knowing it.

It's extremely inefficient in database use because they do dumb things like test for a record before inserting a record. This allows me to insert the same record twice but with different primary key instead of making an intelligent primary key and handling the insert errors. There is no statement handle caching so all that execution plan development that real databases do to improve performance is wasted. Referential integrity rules are enforced through lots of SQL lookups and not by making actual referential integrity rules.

What all this comes down to is RoR is a competitor to PHP but in no way is it capable of operating in a real environment where data is the business and mistakes are limited to a misplaced BLOG but money or worse.

But you'll really have to explain the idea of RoR not encouraging you to use Session variables. When by default the marshall then entire session object and all it's variables -- it's too easy to use Session variables and actually requires some real effort not to. If you refer to the notion of a session variable only referring to a session_id in a table then all I can say is "well, duh! you aren't supposed to use sessions as a data store."

I think RoR is a step in the right direction, but the cost of performance is significant. But the cost of data integrity, integration, and robustness is ultimately insurmountable.

Re:RoR bandwagon?

[Kichigai Mentat]

I know that Apache, NFSd, possibly PHP are installed by default, but GCC and other compiler toys are installed separately, but provided by Apple, usually on the OS disc. They're in the XCode Tools, but they're all there. Apple even gives you the option to download the latest versions of XCode for free, all you have to do is sign up through Apple's Developer Connection (free registration) and sit through the two gigabyte download. But given that this is FIRST PARTY support, that's damn good. I actually kind of like it how the developer tools are separately installed, because if I was an average luser, I'd be PISSED if I lost 2-3 GB on tools I never use.

Re:RoR bandwagon?

[larkost]

You are correct that the comand line version is there. But in the Apache config file the lines enabling the so's are commented out. So for Apache is it turned off.

slashdotted already?

[qw0ntum]

Here's the link to Apple's page describing the developer features: http://developer.apple.com/leopard/overview/index. html

Re:slashdotted already?

[Anonymous Coward]

The inclusion of Trusted Computing hardware, the new Mandatory Access Control and Trusted Binaries... all the key components of a DRM system.

I'm sorry -- what were you Mac fags saying about Apple not being hardcore about DRM? Face it, they are using this stuff to ensure that data can be "locked" to particular pieces of approved code (be it Music/Video locked to certain applications, or word processing documents etc etc). Perhaps now you'll start to realise just what a grubby little jail cell Apple has thrown you in... and kept the key themselves... and worst of all, charged you for the privilege.

Oh... what am I saying. Of course you won't. You'll just keep on kidding yourself that Apple has *your* best interests at heart... and not making money.

Re:slashdotted already?

[99BottlesOfBeerInMyF]

...the new Mandatory Access Control and Trusted Binaries... all the key components of a DRM system. ...I'm sorry -- what were you Mac fags saying about Apple not being hardcore about DRM?

Umm, the MAC is a direct port of TrustedBSD. It is not linked to TPM that I've ever seen and is a feature to help users stop malicious software, not to allow software more access than the user. You might as well complain about SELinux being a DRM system because it is the same architecture. As for signed applications (not binaries) you'll note it specifies them as a mechanism for determining trust levels, which is to say most likely the MAC, not some as of yet missing DRM component.

You'll just keep on kidding yourself that Apple has *your* best interests at heart... and not making money.

In a non-monopolized market, like Apple sells in, often the best interests of the customer are the best way to make money. It's called "capitalism." Maybe you want to stop your knee from jerking at least until you get some actual indication of such a DRM system, instead of what was listed in the develop notes. From what I see it looks like a great way for me to lock down applications and prevent them from behaving in ways I disapprove of, not of applications locking me out. Get a clue.

Re:slashdotted already?

[Anonymous Coward]

Umm, the MAC is a direct port of TrustedBSD.

What's your point... I mean other than demonstrating that you don't actually understand what you are talking about. The MAC is only to prevent you modifying any parts of the kernel without permission. It's the TPM that prevents you gaining the authority to override the MAC. Do some fucking research. TPM+MAC+trusted binaries==DRM.

And yes, SELinux *is* part of a DRM system -- when it the ultimate authority to override it is taken away from the owner. See Motorola's A85 mobile phone.

Maybe you want to stop your knee from jerking at least until you get some actual indication of such a DRM system, instead of what was listed in the develop notes. From what I see it looks like a great way for me to lock down applications and prevent them from behaving in ways I disapprove of, not of applications locking me out. Get a clue.

You are a fucking idiot. Apple has already demonstrated a DRM system built on this -- to the Bluray/HD-DVD consortia. Just the same as Microsoft.

???????? Link working?

[Anonymous Coward]

3 browsers and 2 OSes, no luck loading the link yet.. anyone else?

How utterly ridiculous

[Anonymous Coward]

Yes, let us link to a Mambo/Joomla powered site that's slashdotted without any comments, OR, link to Apple's high powered, easily accessible: Leopard Developer Overview that was recently posted.

Re: ????? Link working?

[gEvil (beta)]

3 browsers and 2 OSes, no luck loading the link yet.. anyone else?

The problem is clearly that you're using the wrong browser and OS combination. Keep on trying different ones. One of them will get the link to load.

Cool

[mjjw]

This is actually quite cool news. I use Macs to develop websites and I also use RoR so to see that Leopard (Server) will come with RoR pre-installed (from the APpl e Leopard overview) is really cool news for me.

Re:Cool

[Anonymous Coward]

May I ask why? It's not like it's superhumanly hard to install it yourself, and if you don't need RoR, it's just another thing that can go wrong.

Some working links:

[mblase]

VNUnet article
Apple Insider post
Apple's Developer Overview site

I wish MS would come out with something like this

[Salvance]

These development tools look great. Particularly Interface Builder and XRAY. I've never used Mac development tools, so it's possible that looks can be deceiving ... but seeing this really makes me wish that MS would start to push their IDE forward rather than adding minor enhancements with each .NET release.

Instant slashdotting!

[Xyde]

BOOM!

You're way behind the times, Mr Area-56

[frankie]

Ruby on Rails was announced back in August, and I (et al) added details about the rest of the dev overview to Wikipedia 5 days ago. Personally, I thought the most interesting new part was Mandatory Access Control.

Boom!

[zoeblade]

BOOM!

Boom... Boom.

Re:Boom!

[Pollardito]

that clip gives me RainMan flashbacks "97X...BAM...the future of rock and roll"

Re:Boom!

[tbone1]

Boom ... and boom goes the dynamite.

Re:I wish MS would come out with something like th

[soft_guy]

I use both XCode and Visual Studio. I much prefer XCode and Interface Builder. There are also a lot of other very nice tools that Apple bundles for free. They are nicer to use than what Microsoft gives you, plus there are a lot of things that you get for free on Apple that you would have to buy third party on Windows such as the coverage tool (gcov) and the profiler (Shark). So, yes, Apple's tools ARE as nice as they appear to be.

Unfortunately, today I have to use Visual Studio and I'm trying to figure out how to get my program to run in a Release build. It runs OK in Debug, but for whatever reason I'm getting an error dialog about not having a manifest file to load the C++ runtime DLL (?). I wish I could use XCode to write Windows apps. Or alternatively that our Windows users would just all buy Macs.

Re: ????? Link working?

[waif69]

Yes, the site has been slashdotted!

Why do people pay for this stuff?

[Channard]

Maybe I'm missing the point here, but why would anyone pay the asking price of just under a hundred quid for a minor revision? No, this isn't intented to be flamebait - I'm a new Mac Mini owner myself and it's getting way more use than my PC. But I can't understand how Apple can charge for what is a pretty damn small upgrade. There were some major major differences between XP and 2000, and I can understand Microsoft paying for these. I can also understand Apple charging for the jump from 9.x to 10.x. But from 10.4 to 10.5? What am I missing here?

Re:Why do people pay for this stuff?

[alienzed]

First of all, Mac OS X.5 will be 64 bit. Also, there will be boot camp built in, a couple new apps, new features. Oh and it's a fraction of the price of Windows.

Re:Why do people pay for this stuff?

[Anonymous Coward]

It's not a minor revision, but a lot of the improvements are visible only to developers. The Tiger upgrade was the same way: from the user's point of view, the only improvements were Spotlight and Dashboard, but from the developer's point of view there were some substantial kernel architecture changes, the advent of Core Data, and the improvement of Core Image and Core Audio.

Re:Why do people pay for this stuff?

[MustardMan]

Uh, the changes in leopard are pretty huge. I'd pay a hundred bucks alone for something like Time Machine, not even counting the huge additions in developer tools and whatnot. I gladly forked out the cash for tiger, and the add-ons like spotlight, which made my life a whole lot easier. Plus, leopard will be truly 64-bit capable, from top to bottom, so that's a huge change. No, not all people will see enough value to upgrade, but then again Apple isn't a software company now is it? I personally will be right there to pick up a copy of Leopard as soon as it comes out.

Re:Why do people pay for this stuff?

[Anonymous Coward]

what you are missing is that this will NOT be a minor update, Apple's and MS's version number systems are different, MS changes the name with each major revision, Apple changes the number after the ., for instances 10.4.5 to 10.4.8 is a minor change, but 10.4.x to 10.5.x is a major change, when 10.5 comes up, you will be dealing with a different OS

Re:Why do people pay for this stuff?

[wiml]

There are some user-visible changes from rev to rev, like Spotlight (the desktop-grep thing) and Time Machine (versioned filesystem and backup) and fast user switching and iChat and so on. But there are also a ton of developer-visible changes; developers will rely on newer revisions because they have features that make writing programs a lot easier and/or allow writing much better programs; users then have to move to a newer rev in order to run programs which require them. And, of course, plenty of people don't pay for it. You can more or less get by buying every other OSX upgrade. I hear there are some people still running OS 9.

Re:Why do people pay for this stuff?

[2nd Post!]

In the jump from 10.4 to 10.5 you get:
You get built in backup and restore software
You get automatic backup functionality
You get virtual desktops
You get built in remote presentation and remote control software
You get new Widgets plus the ability to turn any webpage into a widget
You get a new mail program with increased planning functionality
New group management functionality in Mail and in iCal

Under the hood you get:
New animation libraries
New 64 bit CPU optimizations
New resolution independent ui

You pay for this stuff because you find it useful.

Re:Why do people pay for this stuff?

[theparag0n]

You're missing the fact that the jump from 10.4 - 10.5 is pretty much as big as the jump from 2000 to xp was, improved kernel, better support for new hardware, lots of new software, full native 64bit support, Time machine, virtual desktops, built in programming languages... i'll be upgrading :)

Re:Why do people pay for this stuff?

[Anonymous Coward]

This is not meant to insult you personally, but it's getting pretty old, in my anoncoward opinion, to keep hearing this question four to five years after Apple has set a precedent for performance and functionality gains in each "minor revision."

Each OS X revision Apple has released has genuinely improved upon its predecessor. 10.2 Jaguar and 10.3 Panther improved performance, 10.4 Tiger added features and major under-hood improvements, and presumably 10.5 Leopard, 10.6 Kittycat, and 10.7 Newton (Huey P) will add various mixes of features and under-hood improvements. As the underlying OS solidifes, future releases will likely bias toward user features and away from developer gimmes.

Like with all operating systems.

The problem with (relative) honesty is that people don't get excited by (nominally) point release; most people are gullible (or worse), and want COOL-N-AWESOME XP VISTA PRO DELUXE ULTIMATE MEDIA CENTER GODBOX IPOD EDITION UPGRADE 95! even if the "point release" is an entire OS rebuild and the ULTIMATE product is some kind of massively delayed retread of more of the same.

Re:Why do people pay for this stuff?

[System.out.println()]

Just because it's out, doesn't mean you have to buy it.

Re:Why do people pay for this stuff?

[Anonymous Coward]

Don't tell me you actually believe that those revision numbers mean something? The reason people pay for these is that there are performance improvements and new features in every release - if "Time Machine" in Leopard is as good as it looks its more than worth the dough by itself. not to have to dick around with some crappy backup software.

What you are missing is that these updates should not be characterized as "minor revisions". There are real user-visible features in each one that people want. If you don't want them - don't upgrade.

Re:Why do people pay for this stuff?

[Graff]

here were some major major differences between XP and 2000, and I can understand Microsoft paying for these. I can also understand Apple charging for the jump from 9.x to 10.x. But from 10.4 to 10.5? What am I missing here?

Apple uses a different numbering scheme than most software manufacturers, at least for the Mac OS X releases.

You see, the "X" in Mac OS X stands for the number 10. When Apple does a major release they don't want to have to go to Mac OS XI, Mac OS XII, etc. What they do instead is Mac OS X 10.0, Mac OS X 10.1, and so on. Each 0.1 release is actually a major release, not a minor one. When Apple goes from Tiger (Mac OS X 10.4) to Leopard (Mac OS X 10.5) it is similar to Microsoft jumping from XP to Vista, it's a major upgrade and you are charged for it. When Apple does a minor release they do it in the form of Mac OS X 10.x.y, those are free to download.

I'm sure that eventually Apple will eventually come up with some sort of new numbering scheme once they get to Mac OS X 10.10 because that will just be EXTREMELY confusing to the consumer. For now their numbering scheme is just a little different from what you'd expect but it makes sense once you understand it.

Re:Why do people pay for this stuff?

[t_c_gull]

I mostly agree with you. All of the new features are just bells and whistles except for Time Machine. Having backup and restore features built into the OS make it worth the price of the upgrade.

Re:Why do people pay for this stuff?

[ckelly5]

FWIW, I also have to mention that the upgrade from 2000 to XP was also technically a minor revision (2000 is Windows NT 5.0, XP is Windows NT 5.1).

Re:Why do people pay for this stuff?

[CrackedButter]

Skip it then if you got a problem and wait for Lion.

Re:Why do people pay for this stuff?

[zlogic]

Well, Vista's .NET approach to 32/64bit compability seems a more elegant approach to me than Apple's Universal Binaries. Firefox 2 for MacOS X s a 18Mb download (compared to 5 Mb Windows and something like 7 Mb Linux). Are we going to see universal binaries that contain the PPC, Intelx86 and Intelx64 versions?

Re:Why do people pay for this stuff?

[AliasN]

Yeah.. but you sure WANT to..

Re:Why do people pay for this stuff?

[shmlco]

There are quite a few things like Time Machine, Spaces, significant upgrades to Mail, iCal, iChat, Dashboard, resolution-independent interfaces, plus a lot of under-the-hood enhancements like 2D Extreme, Core Animation, 64-bit, and so on.

Go to the OS X page on Apple's site to see most of the announced "user-side" improvements, and here to see the developer stuff. Just simple stuff like the iCal store can mean lot's of nifty little utilities being generated for that system.

And to reiterate, those are just the "announced" features. SJ implied that there are some other "surprise" features waiting...

Re:Why do people pay for this stuff?

[Trillan]

...and the biggest difference to developers, years of whining "Why won't your program work with 10.3?"

Re:Why do people pay for this stuff?

[RAMMS EIN]

Well, say what you like, but OS X is a very good system, and you get a lot of value for your hundred quid. Lots of applications and tools are bundled with the OS or available at the cost of a registration and a download, all of it made for and/or packaged for OS X. Try to get a GNU/Linux or Windows system up to the same level, and see how much that costs you in money and effort. Apple is also a major innovator in the desktop OS sphere, which means some features they offer may not even be available for Windows or GNU/Linux yet.

Of course, it's not all great for everyone. I prefer Debian over OS X, so I won't be paying for OS X. If all you want to do is run Windows apps, OS X may not be worth the price, either. But for some people, Leopard is well worth what Apple asks for it.

Re:Why do people pay for this stuff?

[Overly Critical Guy]

Maybe I'm missing the point here, but why would anyone pay the asking price of just under a hundred quid for a minor revision?

Because it's not a minor revision. Maybe you should take a look at some Arstechnica reviews to see how much changes in each OS X release.

You also have to remember that Apple doesn't reveal their products until just before release, and we've only been given a developer API peek at Leopard. MacWorld '07 will be the big Leopard reveal.

There were some major major differences between XP and 2000, and I can understand Microsoft paying for these.

And then you make this inane comment. There were major differences between XP and 2000?

Re:Why do people pay for this stuff?

[aristotle-dude]

Yeah, you are missing something. Release versions have nothing to do with how big a release is in terms of new functionality. Keeping the Major version number the same signifies a level of binary compatibility for the API. If Apple were to radically change the API (like they did between 9.x and 10.x) you would see virtually all binaries breaking. Apple did get around this by providing Carbon as a bridge API but the native Cocoa API and related frameworks would not be backwards compatible to Mac OS 9.x nor were non-Carbon apps compatible directly in OS X.

I really would suggest first RTFA, as it explains some of the updates being provided to the core of the OS for developers to harness. Each so-called point release of OS X has provided us with a full version kernel revision and although Apple tries to maintain some backwards compatibility, many drivers written against previous release fail to work in the new release. Sometimes all that is needed is a recompile while at other times, the drivers have to be modified.

I've written about this subject many times on Slashdot and elsewhere that I almost feel like I should write up a FAQ on what version number mean. Had MSFT marketed Windows 2000 Pro NT 5.0 and XP as NT 5.1 would you all be crying about paying for a point release? Open up the About Windows dialog on a Win2k Pro machine and on an XP machine and tell me what version you see.

In a nutshell, the 10.2, 10.3, 10.4 and 10.5 monikers are just names for marketing purposes and to denote compatibility.

Re:Why do people pay for this stuff?

[NivenHuH]

You forgot a biggie...

Now in Leopard, the Objective-C runtime has been updated to include a thoroughly modern and high performance garbage collection system, making memory management a thing of the past.

Garbage collection is included as part of the Obj-C 2.0 runtime... Say bye bye to most memory leaks.. :) I think this is turned on by default and is an opt-out option for your code.

Re:Why do people pay for this stuff?

[Graff]

That download DOES include the PPC and Intel versions of Firefox 2. I don't quite know why the Mac version is triple the size of the Windows or Linux version, it should be at most double the size and probably a lot less. The way Universal Binaries are supposed to work is that each version of the code shares the same resources (graphics, user interface files, etc) and only the actual compiled code is different.

For most applications compiled code is quite small when compared to the graphics and other resources, which means that a Universal Binary should only grow by a small fraction of its size for each new architecture it supports. If Firefox 2 for the Mac is double or triple the size of Firefox 2 for other platforms then I'd suspect that there was some other reason for it than Apple's Universal Binaries system.

Also, there is very little difference in actual code between 32 bit and 64 bit support on the Mac. For most applications you gain a lot of 64 bit support just by using the proper libraries. There are very few applications out there that would need both 32 bit and 64 bit versions of their code, so that probably wouldn't add to the size of the Mac version of Firefox 2.

Re:Why do people pay for this stuff?

[lurch_mojoff]

...No, this isn't intented to be flamebait - I'm a new Mac Mini owner myself and it's getting way more use than my PC. But I can't understand how Apple can charge for what is a pretty damn small upgrade...

How is this not a flamebait? First, you bring up a question that has been answered about a jazillion times already - Steve Jobs is in lovez with the number 10 (or in Apple lingo "X") and from now on for Mac OS the major version number is the one after the first decimal point. And second, even if you actually believed that the numbering scheme proved that this is a minor update you didn't even bother to check what is the new and improved stuff in Leopard, yet you felt the need to post a comment titled "Why do people pay for this stuff?". This is definitely a textbook example of a flamebait.

Re:Why do people pay for this stuff?

[kalidasa]

You do realize that the major change between 2000 and XP was in the user interface, right? And that 2000 was NT 5.0, and XP was NT 5.1? I'd argue that the real "major" OS change was from XP to XP SP2. Oh, and the charge for the 2000 to XP upgrade was $200; the charge for the OS X upgrades is usually $129.

Apple charged for all X.Y changes under OS X except for the leap (and it was a leap) from 10.0 to 10.1. These changes are significant version changes (heck, 10.4 includes a whole new architecture).

Re:Why do people pay for this stuff?

[Tom]

Maybe I'm missing the point here, but why would anyone pay the asking price of just under a hundred quid for a minor revision?

Because Apple's minor revisions contain more updates and changes than a completely new windos version, for example.

Check the list of changes, not the number that changes in the descriptor.

Re:Why do people pay for this stuff?

[Fahrenheit 450]

Does this mean Safari will finally stop eating memory like a Langolier?

Re:Why do people pay for this stuff?

[Fred_A]

Firefox 2 for MacOS X s a 18Mb download (compared to 5 Mb Windows and something like 7 Mb Linux). Are we going to see universal binaries that contain the PPC, Intelx86 and Intelx64 versions?

Let's just hope OS X isn't ported to Sparc, Alpha and MIPS...

Re:Why do people pay for this stuff?

[AhtirTano]

Never has your sig been more appropriate.

Re:Why do people pay for this stuff?

[Quila]

There were major differences between XP and 2000?

Upgraded UI, symmetric multithreading, better security (if you count SP2), CD writer support, Remote Desktop, firewall, fast user switching, system restore, DLL backup, etc. Plus if you count Server 2003 you get the very re-worked IIS 6 and some more toys.

IOW, Windows 2000 to Windows XP is about the same jump as a dot release in OS X, and both cost money.

Re:Why do people pay for this stuff?

[Llywelyn]

A couple of things that were missed that are more developer specific:

- Xcode 3.0, which has a lot of features I've been missing.
- Xray, built off of DTrace (this looks incredible).
- Dashcode, a widget development environment.
- Image Kit.
- OpenGL improvements.
- Code signing.

Then a few other things for end-users (and we don't even know all of what's in leopard yet) :
- Additional iChat features and integration.
- Resolution independence.
- Improved voice synthesis.
- Improved features for searching. ...and the list goes on.

Re:Why do people pay for this stuff?

[cerelib]

I am not trying to challenge your statements and I am not a Mac owner, but I find your omission of 10.0 and 10.1 to be rather convenient. If my memory is correct, most new applications for OS X usually put the requirement of 10.2 or greater, right? So it would seem that there is a bit of incompatibility between 10.x releases. What changed in 10.2 to leave 10.0 and 10.1 owners out in the cold?

Re:Why do people pay for this stuff?

[akincisor]

I'm sure that eventually Apple will eventually come up with some sort of new numbering scheme once they get to Mac OS X 10.10 because that will just be EXTREMELY confusing to the consumer.

I'm sure they will run out of big cat names before that :)

Re:Why do people pay for this stuff?

[AnotherDaveB]

You get new Widgets plus the ability to turn any webpage into a widget

I have a little web-app, so I didn't much care for the sound of this. A trip to Apple tells me more:

Clips and flicks

Create your own Web site widget using Web Clip in Safari for Leopard. Just visit your favourite site and click the "Open in Dashboard" button in Safari. Dashboard launches a new clip of the site in a customisable widget. From there, you can resize your Web Clip and choose from a handful of window themes. And since your Web Clip is always live, it acts just like the Web site it was clipped from. Dashboard for Leopard also introduces a movie widget that finds cinema schedules in less time than it takes to make popcorn.

Is there a "meta" tag I can use to ask OSX not to do this?

Re:Why do people pay for this stuff?

[jafac]

Bah - I'm just waiting for Apple to screw me, because I have a G5, and haven't upgraded to the intel mac yet.

Re:Why do people pay for this stuff?

[Dog-Cow]

It's not any different than the user leaving Safari open with a window/tab open to the page. Why in the world would you possibly care? Are you just an anti-Apple /-hole?

Re:Why do people pay for this stuff?

[Anonymous Coward]

safari is written in C++

Re:Why do people pay for this stuff?

[aristotle-dude]

When I was referring to backwards compatibility, I was referring to the ability of the new release to run binaries compiled against a previous release of OS X (ie applications compiled against 10.3 should run on 10.4). If a developer chooses to compile in and use features of an existing API or to use an API new to 10.4, it will obviously not run in 10.3 on earlier. When you see software requiring 10.2 or greater, that means they are using features not present in previous releases. I cannot think of what changed exactly in 10.2 but I think it might have had something to do with changes to Core Audio. I know that 10.2 brought OS X hardware accelerated compositing in the form of Quartz Extreme. IIRC, 10.0 was a paid release and 10.1 was a free upgrade for anyone who had bought 10.0 since 10.0 was still basically beta.

Re:Why do people pay for this stuff?

[Overly Critical Guy]

Better security doesn't count because SP2 which came years later. DLL backup was in 2000, as was symmetric multithreading. CD writer support was in 2000 (I assume you mean built-in CD writer support, which is a minor feature). The firewall was in 2000, and System Restore came from Windows ME and is hardly a major feature by itself.

Basically, aside from the goofy Luna theme that was hacked in after Apple revealed Aqua, everything you listed could have been in a Microsoft Plus! pack.

IOW, Windows 2000 to Windows XP is about the same jump as a dot release in OS X, and both cost money.

So far, you haven't offered much to support such a claim. OS X Tiger, for instance, introduced entire developer APIs like Core Image/Core Video/Core Data, ACLs, updated the kernel with a new locking scheme, introduced a 64-bit UNIX addressing space, introduced a metadata indexing service with kernel notification hooks, introduced a widget layer, and much more.

Re:Why do people pay for this stuff?

[2nd Post!]

Yeah, you don't click the "Open in Dashboard" button and then OS X will NOT do this.

It's a tool, and if you don't want to use it, then don't use it.

Re:Why do people pay for this stuff?

[SeaFox]

But I can't understand how Apple can charge for what is a pretty damn small upgrade.

Because people will pay for it? As long as the majority of consumers are willing to pay the price then the price is not too high. That's how the free market works.

And not to nitpick, but there's no such thing as an OSX "upgrade". There's only two boxed verions (client and server). The disks you buy contain the whole OS, you don't have to give the license key from a previous version or have a previous version installed on your hardware already. Upgrades give a finiancial incentive for staying up to date, so the older the version you have, the more you pay. But with Apple, you could jump straight from OS9 to 10.5 and pay the same as someone going from 10.4 to 10.5.

Re:Why do people pay for this stuff?

[Quila]

Better security doesn't count because SP2 which came years later. DLL backup was in 2000, as was symmetric multithreading. CD writer support was in 2000 (I assume you mean built-in CD writer support, which is a minor feature). The firewall was in 2000, and System Restore came from Windows ME and is hardly a major feature by itself.

I count through SP2 since it was free (like saying advances with 10.4.8 don't count).

W2K hyperthreading didn't really work, it just saw two processors, meaning upgrading your 2xSMP box to hyperthreading chips put you beyond 2K's two-processor limit. Some 2K boxes wouldn't even work right with hyperthreading, and it could even hurt performance. XP understood hyperthreading and could really take advantage of it, and would see that system as two processors with two logical processors each.

For CD writers, you had to have a CD writing program for 2K, not so for XP.

By firewall I didn't mean the advanced TCP/IP settings, I meant the ZoneAlarm copy.

And system restore works better in XP than ME.

And don't forget remote assistance and included ZIP (finally got something I had with Norton Desktop in the early 90s).

And you dismissed all the improvements in Server 2003, and they should be included since there will be an OS X 10.5 Server.

I think the comparison is apt, and gets the point across to MS-only users who can't understand paying for a dot release just because it has a dot in it.

Re:Why do people pay for this stuff?

[PygmySurfer]

APIs did often change in early releases of OS X. 10.0 and 10.1 were basically the foundations future releases built upon. It took Apple some time to get things stabilized. It wasn't until 10.4 that Apple promised "no API disruption for the foreseeable future." OS 9 to OS X was a huge change, it's really not surprising it took Apple a few releases to nail down API stability.

Re:Why do people pay for this stuff?

[ChrisA90278]

So you think 10.4 to 10.5 will be a minor upgrade? How do you know? Apple is not telling us what will be included in 10.5 except for a few features they are previewing to developers. The best features ar being kept secret. You may be right but we won't know untill spring 2007.

Why pay? you don't have to. If you like Tiger (10.4) then stick with it abd if you buy a new computer in 2007 the newer OSX will be included for "free".

Re:Why do people pay for this stuff?

[Overly Critical Guy]

I count through SP2 since it was free (like saying advances with 10.4.8 don't count).

SP2 came out years after XP, and all it really did was update the firewall, add a Welcome Center, and recompile some DLLs using the latest Visual Studio. I'm not counting advances in 10.4.8 because those kinds of free minor updates from Apple are mostly bug fixes.

W2K hyperthreading didn't really work, it just saw two processors, meaning upgrading your 2xSMP box to hyperthreading chips put you beyond 2K's two-processor limit.

You said SMP before, not Hyperthreading. Of course Windows 2000 didn't support Hyperthreading because it wasn't around when Windows 2000 was released! Intel doesn't even use Hyperthreading anymore.

For CD writers, you had to have a CD writing program for 2K, not so for XP.

Again, bundling in a CD writing wizard is hardly a major feature.

By firewall I didn't mean the advanced TCP/IP settings, I meant the ZoneAlarm copy.

Bundling a crippled firewall isn't a major feature, either.

And system restore works better in XP than ME.

System Restore is simply a background process that takes snapshots of system files.

And don't forget remote assistance and included ZIP (finally got something I had with Norton Desktop in the early 90s).

Wow! Zip support! That's definitely worth $250 for Windows XP Professional!

As I said before, all of this could just have easily been in a Windows 2000 Plus! pack.

And you dismissed all the improvements in Server 2003, and they should be included since there will be an OS X 10.5 Server.

I dismissed Server 2003 because we were talking about the difference between 2000 and XP.

I think the comparison is apt, and gets the point across to MS-only users who can't understand paying for a dot release just because it has a dot in it.

The comparison is hardly apt, and you ignored the Arstechnica articles I linked as well as the list of new APIs and fundamental architectural updates I mentioned in OS X Tiger that far surpass the level of changes introduced in XP, which barely had any new APIs and technologies at all compared to 2000.

Re:Why do people pay for this stuff?

[tim1724]

Safari is written in a combination of both C++ (for the KHTML bits) and Objective-C (for the GUI stuff). And switching it to use garbage-collected Obj-C would involve a bit of work, so I doubt it'll happen soon.

Re:Why do people pay for this stuff?

[aJester]

That is it. You sir are a troll...

Nevertheless.....
I am tired of seeing this question in EVERY Mac OS discussion.
For the millionth time, 10.4 --> 10.5 is NOT a minor update.

Mac OS X is a brand name.
Read Tiger as "Mac OS X. Release 4."
Read Leopard as "Mac OS X. Release 5".

(If you are still having problem, tell ur brain it is New Mac OS 4 and New Mac OS 5. They are the Next MAJOR release of Mac OS X. Since Apple has spent a lot of money marketting the Mac OS X brand, they want to leverage it for as long as possible.]

To give a windows parallel.
Windows NT--> Windows 2000 --> Windows XP -----> Vista

Similarly for the Mac OS X, the sequence is as follows.
Jaguar------> Panther --------> Tiger ----------> Leopard.

They are NOT minor updates or point release despite the naming convention,which has some historical reasons.

Jes

Re:Why do people pay for this stuff?

[sofla]

Yep. Someone from Sun marketing must have gone over to Apple. Each "minor" rev of 10.x has been a major rev. And each one has been different cat (Puma, Jaguar, Panther, Tiger, Leopard).

I think maybe their marketing guys are afraid to go to 11 because then they'd have to call it OS XI or something (keep in mind many consumers, myself included, call it "oh ess exx" not "oh ess ten"). We should see something different after 10.9, sooner if they run out of cat names.

Re:Why do people pay for this stuff?

[The One and Only]

No. Safari's memory leaks are in the C++ code (which is, in turn, a fork of the KHTML code). The only Objective-C in Safari is the Cocoa wrapper.

Re:Why do people pay for this stuff?

[doh123]

you just look at the size of the download and assume that means the program is 3X the size? Also its about compression of the downloaded file. Need to look at the actual program size... basicaly look at the full install directory. Mine appears to be 50.2 MB on OSX. Havent installed it on Windows yet...

Re:Why do people pay for this stuff?

[Shados]

Then they can use reptile names! Mac OS X Cobra has a nice ring to it.

Re:Why do people pay for this stuff?

[ccmay]

I hear there are some people still running OS 9.

I'm aware of at least one site still running System 7.5.5 on 68040 hardware, because it works fine for what they need.

-ccm

Re:Why do people pay for this stuff?

[zeno_2]

Amazingly enough its 25.1mb on Windows, exactly half the size. Id still rather be using firefox on a mac though..

Re:Why do people pay for this stuff?

[endofcell]

Now that's what I call a decent Slashdot comment. Informative and insightful. thanks!

Re:Why do people pay for this stuff?

[IntergalacticWalrus]

Those aren't small updates, they're just as significant as the 2000->XP update. There's lots of new apps added, and some pretty important "under the hood" changes.

Besides: Windows 2000 was NT 5.0, Windows XP was NT 5.1. Notice a pattern?

Re:Why do people pay for this stuff?

[Kichigai Mentat]

I decided to have some fun with Monolingual (http://monolingual.sourceforge.net/), after reading your comment. After stripping out the unneccessary localizations, FF2 for OS X is 50.2 MB. After stripping out PowerPC G5, G5 64-bit, Intel and Intel 64-bit architectures (keeping previous generations because I have some apps that were compiled for G3s and earlier that I'd like to keep functioning) and retaining PowerPC, PPC G3, G4, G4+, was 30.6 MB. I expect that if I were to remove PPC/G3/G4 (I use a G4+) I would find my FireFox to be even smaller.

Thing to remember is that these are precompiled binaries, built to function anywhere. Binaries compiled on-system (from source) are most likely significantly smaller (depended on how efficiently they're coded that is)

Other notes: On the whole, I recovered about 1.1 GB by removing un-utilized localizations, and around 4.1 GB in unused architectures. There's definitely some bloat in the software, but that's what you get with a multi-architecture system. I expect if more Windows apps had 64-bit binaries included, we'd see (more) bloat there. Big thing to remember: Apple is in the trailing edge of the PowerPC-x86 transition, we can't expect them to abandon PPC all of a sudden.

Re:Why do people pay for this stuff?

[Quila]

all it really did was update the firewall, add a Welcome Center, and recompile some DLLs

This list of fixes in SP2 is HUGE, in addition to improvements to IE, Outlook Express, and automatic updates (not that I use any of them). It also did a lot with wi-fi and bluetooth.

You said SMP before, not Hyperthreading.

No, I said "symmetric multithreading," and "Hyperthreading" is just Intel's brand of it.

Again, bundling in a CD writing wizard is hardly a major feature. Bundling a crippled firewall isn't a major feature, either. Wow! Zip support!

They all add up.

I dismissed Server 2003 because we were talking about the difference between 2000 and XP.

Leopard will be available in both client and server versions. It is only fair to count Windows client and server versions.

Re:Why do people pay for this stuff?

[Altus]

are you kidding me? Mac OS X 10.10... thats Mac OS XXX the hottest OS release EVER!

*cue porno music here*

Re:Why do people pay for this stuff?

[Achromatic1978]

Oh and it's a fraction of the price of Windows.

Spin spin spin. To buy OS X.5, as fanboys are telling me constantly, it's technically an upgrade, not a full version. After all, you have to own the existing OS.

Apple Store sells OS X 10.4 for $129
Amazon sells Windows XP Home Upgrade edition for $89

So, I guess it's a fraction of the price of Windows. If by fraction you mean 13/8ths.

Re:Why do people pay for this stuff?

[Anonymous Coward]

Mac OS X 10.n (spitting one-eyed trouser snake)?

Re:Why do people pay for this stuff?

[Achromatic1978]

And to reiterate, those are just the "announced" features. SJ implied that there are some other "surprise" features waiting...

Gasp. Really? And here I was thinking he'd say "Nothing else. That's it. If you're not blown away by what I've shown you, don't go thinking there's more waiting. You'll be disappointed"...

Re:Why do people pay for this stuff?

[demallien2]

Yup, you're missing the point. A change from 10.4 to 10.5 is a major version release, hence the name change from Tiger to Leopard. It's just that Apple wan't to keep the 10, so that they can justify the X at the and of MacOSX. And why keep the X? Because it's traditional that Unixy operating systems have it in the name (think Unix, Linux, NeXt etc etc etc) So what you are thinking of as minor point releases are actually the changes from say 10.4.7 to 10.4.8. And as you would expect, they are free.... See other posts for a list of the new features that have been added in this next release - suffice it to say that some of these new features, such as TimeMachine, are groundbreaking, and well justify the price of an upgrade... Of course, if you don't agree, don't upgrade. It's a free market.

Only in the SERVER version.

[Anonymous Coward]

Ruby on Rails is only going to be installed on OS X server. Not the regular user version. Pretty important to note considering the install base of SERVER vs. regular.

Re:Only in the SERVER version.

[horati0]

Wrong.

Re:Only in the SERVER version.

[rollthelosindice]

There is a difference between being included on the developer DVD( something users can optionally install ) and being included in the install that is already on their hard drive.

So does this have a performance impact?

[Channard]

If this is as big a jump as from 2000 to XP, is there likely to be a performance hit as there was with XP? Did OSX 10.4 run slower on Macs than 10.3 did?

Re:So does this have a performance impact?

[avalys]

New versions of Mac OS X have always run faster than their predecessors, because Apple is continually including optimizations. I believe that was also the case with the 10.3 to 10.4 transition, although it was less noticeable speedup than from 10.2 to 10.3.

I doubt Apple will be able to keep it up forever. I haven't heard anything on what the early builds of 10.5 are like, but (being early builds) it's probably too soon to draw any conclusions from them. Still, eventually they'll run out of things to optimize, and all the new features will catch up to them.

Re:So does this have a performance impact?

[Anonymous Coward]

"If this is as big a jump as from 2000 to XP, is there likely to be a performance hit as there was with XP? Did OSX 10.4 run slower on Macs than 10.3 did?"

No, in fact from 10.1 to 10.2 and 10.3, OS X has been running faster, which is true on older machines too. It appears that Apple is continually optimizing the code for performance, which doesn't appear to be what Microsoft does.

(It could also mean that earlier versions of OS X were sufficiently inefficient that it wasn't hard to make later versions run a little faster!)

I've been very happy with the continually improving performance of OS X on my older Macs which are at the edge of the compatibility list.

Re:So does this have a performance impact?

[Moofie]

With the conspicuous exception of 10.3 to 10.4 (which added Spotlight file indexing), each release of OS X has been faster than the previous one.

Turn off Spotlight, and 10.4 continues the trend. It's up to the user to determine if the utility is worth the slight performance hit. I find it very, very useful.

If you don't percieve the utility, I recommend not buying the upgrade. However, my guess is that you don't own a Mac, which makes me wonder why you care.

Re:So does this have a performance impact?

[Graff]

Believe it or not, generally each revision to Mac OS X GAINS speed rather than slows down.

The engineers at Apple have been tightening up their code with each release, finding new and better ways to do things. With each new release I've been finding my systems gaining reliability and speed. Memory requirements have slowly been climbing but overall CPU usage is steady or even a bit lower.

On older systems you might not be able to use some of the new technologies in the newer releases but it shouldn't affect your ability to run the latest version of Mac OS X. The newer graphics and such that won't run on older systems will just fall back into a simpler mode and will continue to work just fine as they did in past revisions.

Re:So does this have a performance impact?

[mstroeck]

Probably not, though it's hard to say without having seen the current builds of Leopard. So far, every new incarnation of OS X has been faster and more usable on the same hardware than its predecessor. 10.4 for example, is just as fast as 10.3, while having stuff like Spotlight and Dashboard going on in the background all the time. I know people who happily run 10.4 on G3 iBooks(!). They don't get all the 3D-accelerated eye-candy, but all the rest works.

Keep in mind though: Even if there is no performance increase (or a slight decrease), 10.5 will be a different beast altogether from 10.4 usability and feature-wise: Time machine. Resolution-independent user interface. Cocoa will be fully garbage-collected(!). 32 bit and 64 bit and applications on the same machine, running right next to each other. Insane screen-sharing technology via iChat. All that in addition to the stuff that's already there in OS X and all the power of most UNIX applications ever written...

Re:So does this have a performance impact?

[Anonymous Coward]

Actually, Spotlight and Dashboard have been known to slow down Tiger considerably on older systems with low amount of RAM. Tiger would pretty much kill my Pismo/400mhz/192mb, but it showed great improvements from 10.3 to 10.4 on my mac mini (I had the early model where I got a "Free 10.4 when it's out" coupon). Also.. Many mac users would agree 10.2 really sucked.

Re:So does this have a performance impact?

[644bd346996]

I expect 10.5 to be much faster on Intel, particularly the core 2 systems that support 64bit processing. However, those systems are already so responsive that it won't feel much faster. Optimizations just give Apple the headroom to squeeze in more features without slowing things down.

Re:So does this have a performance impact?

[gravesb]

You are right, but I wish they would stop offering the 512MB option on products. I upgraded to 1 gig, and still find it limiting. And that's with all compiling done on a linux box.

Re:So does this have a performance impact?

[Quila]

In addition to what was said in the other comments, Apple has been specifically tweaking the kernel to better handle all the threads flying through it, and that has been giving better performance every time. In addition, Apple has been offloading more and more of the UI to the video card. As of Leopard, not only will almost all UI heavy lifting be on the video card via OpenGL, but it will automatically spin the little work that the CPU must do onto a different thread that can be run on another core (all Macs are now at least dual core).

Job-specific performance increases also exist. Anyone doing video filters and page transitions with an application that uses Apple's Core Image library will see the effects in real-time due to their being offloaded onto the video card. This works in all Intel Macs except the MacBook and Mac mini, but Core Image does degrade gracefully by switching to the CPU for processing when the GPU isn't capable (no need for the developer to worry about whether GPU acceleration exists on the user's computer).

Re:So does this have a performance impact?

[bnenning]

Anyone doing video filters and page transitions with an application that uses Apple's Core Image library will see the effects in real-time due to their being offloaded onto the video card. This works in all Intel Macs except the MacBook and Mac mini

The integrated graphics in the MacBook and mini do support hardware CoreImage acceleration, although obviously performance won't be as good as with a discrete GPU.

Re:So does this have a performance impact?

[Quila]

The integrated graphics in the MacBook and mini do support hardware CoreImage acceleration, although obviously performance won't be as good as with a discrete GPU.

I'd like to know the source, because everything I've read says they don't. They do support the GPU-accelerated UI though.

RoR was developed on Macs

[Anonymous Coward]

Need I say more?

Re:RoR was developed on Macs

[BrianPan]

Need I say more?

yes

Re:I wish MS would come out with something like th

[zlogic]

You can call me a Microsoft fanboy, but I think Visual Studio 2005 is one of th best development environments out there. It has Intellisense, debugging (done right!), great project management (if you use it correctly). Not to mention a lot of .NET stuff like an API for saving application settings in an XML file (no registry keys and just as easy as adding a component to a form).
The only IDE I think is just as good is NetBeans.
Oh, and if you develop non-C# applications you probably won't notice half the features, because C++ doesn't seem to be updated a lot since VS 6.0

What major differences between XP and 2000?

[argent]

why would anyone pay the asking price of just under a hundred quid for a minor revision?

I don't know... I don't normally bother upgrading unless I need to, I'm still using Panther at home and the only reason I upgraded from Jaguar was that I bought new hardware.

There were some major major differences between XP and 2000

I'm still using 2000 because I've yet to find anything in XP that I actually need. The only thing that I've even missed is Bluetooth support. If they'd put a full Citrix server in there instead of ripping so much out that it's basically a low-bandwidth VNC plus virtual consoles... maybe.

And the icing on the cake is: XP Pro *upgrade* costs as much as OS X, and is still bundled with a remote destruct switch in Microsoft's hands.

I can also understand Apple charging for the jump from 9.x to 10.x. But from 10.4 to 10.5?

The jump from 9.x to 10.x was like the jump from Windows 9x/Me to Windows NT/2000. Whole new OS, from the ground up.

The jump from 10.x to 10.y is much more like the jumps from NT4 to 2000 to XP to Vista. New frosting on the same cake. If you need the new features, get them. If you don't, don't. You don't need to jump if you don't want to.

Quartz 2D Extreme

[X_Caffeine]

Has anyone figured out if Quartz 2D Extreme is enabled by default on the Leopard edition? Does it work any better than the Tiger version?

Re:Quartz 2D Extreme

[Drizzt Do'Urden]

It is enabled if your video card supports it.

Re:Quartz 2D Extreme

[Quila]

Ars Technica said Quartz 2D Extreme was there and possible to use, just not enabled because it probably hadn't been completely worked out by Apple yet.

But given fact #1, that Ars said that Q2DE is basically like running your whole desktop as an OpenGL scene, and fact #2, that Leopard will have "resolution-independent interfaces," I'm betting that Q2DE is fully running and implemented in 10.5.

Re:I wish MS would come out with something like th

[Anonymous Coward]

You should try cocoa bindings and interface builder... makes .Net look outdated!

ichat improvements

[ubrgeek]

The iChat improvements (being able to do presentations, etc.) look really neat. But I have to assume there's a typo in the description: "Another new feature in the Instant Messaging framework is the ability to manage presence. This means that your application can query the iChat status of the user, indicating whether they are available or busy and what their status message is, as well as manipulate it. " That can't be right, can it?

Re:ichat improvements

[quoll]

I don't see the problem. If an application is presenting in full screen mode, I would find it useful for it to mark me as "Occupied" or something like that. As for the status message, there is already a link to iTunes to get the song that is currently playing.

Re:ichat improvements

[bjackson]

I think that it is - similar to what iTunes does now - for example, changing your status message to the currently playing song / movie / tv show. New functionality - maybe WoW could display the server you're playing on, etc...

Re:ichat improvements

[Rob_Warwick]

It's not entirely new functionality, though as far as I know you previously had to rely upon Applescript calls to iChat to do this.

A way to do it that doesn't feel kludgy? Excellent.

Re:ichat improvements

[ubrgeek]

But then my wife would know where to find me ... ;)

"Only for the developers" is an oxymoron

[mstroeck]

People in these discussions keep moaning about how various changes "are only really any good for the developers"... That's kind of dumb. Developers, duh, develop the applications end-users rely on. The under-the-hood changes in Leopard are bound to benefit everyone. Garbage-collection, better debugging, 64 bit, a crazy animation toolkit, etc. (if you don't know how that impacts the user, try this for example: http://www.discoapp.com/ . It's a new disc-burning application with one of the most interesting work-flows I've ever seen).

Re:"Only for the developers" is an oxymoron

[pkulak]

No it's not, it's a phrase.

Re: Link working?

[exp(pi*sqrt(163))]

If you just stuck with one OS and one browser maybe it'd reduce the load on the servers and you'd give people a chance to read the article.

Re:I wish MS would come out with something like th

[tconkling]

I'm a die-hard Mac user at home, but I write games for Windows (using MS's dev tools) for a living. As a Mac development hobbyist, I spent years using Metrowerks' CodeWarrior IDE, and -- more recently, and to a lesser extent -- Xcode.

Although I prefer the look and feel of Apple's dev tools to Microsoft's, I find that I get work done more quickly with Visual Studio than with Xcode. More accurately, I get work done more quickly with Visual Studio and the excellent third-party plugin Visual Assist, which provides a number of important code navigation shortcuts including code completion that completely eclipses VS's IntelliSense.

Visual Assist is one of those tools that's painful to be without once you start using it. It sounds like the company isn't planning on a Mac version, which is a shame... do any Mac devs out there have tips on how to make the Xcode development experience less painful -- specifically, how to deal with its poor code completion facilities and slow text editor?

Re:I wish MS would come out with something like th

[soft_guy]

It sounds like the company isn't planning on a Mac version, which is a shame...

If you are talking about my case (the GPP), then yes, actually I am doing a Qt application that ships on Mac, WIndows, and Linux. And Windows is a fscking pain in the ass. What I meant was that if the WIndows users would just switch to mac, I wouldn't have to screw with Windows and could just do Mac :-)

Today (and I literally mean today), I am working on a specific issue in Visual Studio. But we DO have a Mac version.

Re:I wish MS would come out with something like th

[tconkling]

No, I was talking about the Visual Assist plugin for Visual Studio. The company that makes Visual Assist, Whole Tomato Software, apparently has no plans to produce a Mac version of the plugin.

Best news for me is Ruby and Python...

[jocknerd]

become 1st class languages on OS X. Apple is adding the ability to write cocoa apps in Ruby and Python. Sure, you could do it in the past, but Apple wasn't very supportive to the PyObjC and the RubyCocoa projects. But from what I saw on the demo's, Apple has been working hard to make Ruby and Python a legitimate choice for cocoa development. So now you'll have 3 choices: Objective C, Ruby, and Python. By the way cocoa development with Java is canned.

Re:Best news for me is Ruby and Python...

[I'm Don Giovanni]

Well, Apple used to support writing Cocoa apps in Java. They pulled that support as of 10.4. Do you really want to rely on Apple continuing Cocoa support in any language except Obj C?

Re:Best news for me is Ruby and Python...

[bnenning]

Well, Apple used to support writing Cocoa apps in Java.

Which they realized was a lousy idea because of the large impedance mismatch between Java and ObjC. Java has to know everything at compile time, which makes bridging to other languages a pain; for example you have to explicitly create stubs for every method that could possibly be called. More dynamic languages like Python and Ruby don't have those limitations, and are much easier to support.

Re:Best news for me is Ruby and Python...

[LizardKing]

Apple haven't pulled support for writing Cocoa apps in Java with 10.4, however they do discourage it. If you work through the tutorial with XCode 2.4 on Tiger, it still works.

Re:I wish MS would come out with something like th

[soft_guy]

But there are lots of similar things for the Mac, editors, etc. Have you looked at TextMate?

Mandatory Access Control

[alxtoth]

Mandatory Access Control is another compelling reason, at least for me. Currently there are no viruses in the wild, but few trojans and some proof-of-concept Bluetooth exploits. Anyway, the risk is there, and this particular implementation (derived from TrustedBSD) looks promising.

Re:Mandatory Access Control

[Anonymous Coward]

Incidentally, MAC is an absolute requirement for a DRM system that will be built with the Trusted Computing TPM chip built into all Intel Apple Macintosh.

Re:Mandatory Access Control

[99BottlesOfBeerInMyF]

Incidentally, MAC is an absolute requirement for a DRM system that will be built with the Trusted Computing TPM chip built into all Intel Apple Macintosh.

Do you have any actual information here? My reading is this is a straight port of the MAC from TrustedBSD, which is to say the user can edit any ACL. It is possible that it will be modified to use the TPM to lock the user out of changing some ACL, but why? OS X already uses encrypted binaries for that, which are more secure (from a DRM perspective) than what you propose.

Re:Mandatory Access Control

[Anonymous Coward]

Mandatory access control is an absolute requirement for trusted binaries (and DRM). You can't make them work without it. Once you lock down the kernel with the TPM, mac prevents anything changing it further.

There are plenty of security papers on the web documenting this. Educate yourself...

Re:Mandatory Access Control

[99BottlesOfBeerInMyF]

Mandatory access control is an absolute requirement for trusted binaries (and DRM).

Umm, OS X already has encrypted binaries in 10.4 that work for DRM purposes, so MAC is not a requirement for that feature. I fail to see how MAC will make it any more restrictive for them.

There are plenty of security papers on the web documenting this. Educate yourself...

I have a reasonable handle on both OS X's current DRM and the MAC in TrustedBSD that Apple copied. I think you're talking out of your ass. If there are plenty of papers on Apple's MAC integration with DRM, or even using a MAC to implement DRM, I'd like to see them. I've seen absolutely nothing to indicate Apple's implementation has anything to do with DRM. My suspicion is that you don't either. Put up or shut up.

Re:I wish MS would come out with something like th

[drakken33]

I'm a Linux and Mac user at home with almost no MS software (I have played with Vista RC1 recently and have Win2k in a VM for web site testing). I'm not a fan of MS or their software but I have to use XP at work. I think it's important that I say this up front because of what is to follow.

We have VS 2005 at work and I recently got permission to install it (no one else was using it as our resident Windows dev has gone back to Delphi) to create some tools to make my life easier. I have to say I was very pleasantly surprised. I can make my tools very quickly indeed with C# and .NET and the IDE is pretty good.

The downside is that .NET can make you lazy because it does so much for you. It does 90% of what I need but the last 10% has me spending too long looking through the docs to see if there are properties, methods or events that let me do what I want easily. For example, I'm using a TabControl but I couldn't find a way to detect right-clicks on a TabPage's tab so I could pop up a context menu so I wrote some code to loop through all the TabPages, see if the right-click was on each TabPage in turn and if so pop up the context menu. I'm still learning the framework but that seems like a round about way to do things and something that should be there already.

Next to VS 2005, Xcode/Objective-C/Cocoa feels quite "old school" but I like that. It's more like the way I was taught. It can be hard work but it's worth it. I think Apple's approach may be less RAD but it maybe a more flexible approach. It's too early in my VS 2005 usage to be sure. At least Apple provide dev tools with their OS. If you want to attract the home coder you need good free dev tools so I'm glad MS offer the Express versions of VS and I plan to look at at least Visual C# Express to see how that compares to Xcode.

Re:I wish MS would come out with something like th

[tconkling]

TextMate's code completion facilities are not even close to what Visual Assist provides for C++.

From the TextMate FAQ:

Q: Does TextMate have code completion, e.g. type a variable name and see all member data for that variable (object)?

A: No.

What it does have is word completion based on current buffer and insertion of snippets or commands using tab triggers. There is also a PHP code completion bundle that adds completion and help for native PHP functions.

Better code completion is really what I'm after. TextMate does look like an interesting tool -- just not necessarily an improvement over Xcode for C++ development.

64 bit applications in Leopard

[iljitschvanbeijnum]

For most applications you gain a lot of 64 bit support just by using the proper libraries. There are very few applications out there that would need both 32 bit and 64 bit versions of their code, so that probably wouldn't add to the size of the Mac version of Firefox 2

On PowerPC you'd only want applications that actually need to do 64-bit math or address more than 2 or 3 GB of memory to be 64-bit, but for x86 it's a different story because the extra registers that are available in 64-bit mode may make applications that have no use for 64-bitness in itself a lot faster.

So I'm curious as to what developers will do. Hopefully, they'll evaluate the performance of their code and compile for 32 or 64 bits depending on which is faster. But of course you want your app to work on older Macs too, so you may need to include:

• A PowerPC/no Altivec version for G3 Macs
• A PowerPC+Altivec version for G4 Macs
• A PowerPC 64-bit version for G5 Macs
• An x86/32 version for Core 1 Macs
• An x86/64 version for Core 2 Macs

18 MB may not seem so bad at some point in the future. :-)

Re:64 bit applications in Leopard

[iluvcapra]

Just an FYI: for those of you that really want smaller binaries, the lipo(1) program can take a universal binary and strip out the executable resources for architectures you know you won't need. lipo comes with your Mac and can be found at /usr/bin/.

I personally think such an approach is a recipe for disaster, but if you're trying to make a LiveCD or some sort of mini-installlation, it might be the only way to go.

Re:64 bit applications in Leopard

[Kichigai Mentat]

For a GUI version that does more than just that, try Monolingual, which identifies your architecture and offers to strip out incompatible binary data, and also offers to strip out unused locales.

Re:64 bit applications in Leopard

[Anonymous Coward]

On PowerPC you'd only want applications that actually need to do 64-bit math

No, Apple's gcc 4 branch emits 64-bit instructions in ILP32 mode which emit 64-bit math instructions for handling the long long type. In the PPC970 (G5) series these are as fast as 32-bit math instructions. Apple's gcc 4 branch also can emit G5-only instructions which make use of 64-bit GRPs and the like, and these can give some speed improvements in ILP32 code.

Apple has relevant documentation and the general tone is to avoid using PPC's LP64 mode unless absolutely necessary, since it probably will impede performance in most code.

The primary problem with LP64 mode in Tiger is that only libSystem has been built with LP64 support. Using "-m64" and linking only to libSystem is certainly possible, but if you reach out and want to use libtermcap or anything else, you're stuck, or have to build your own LP64 libraries. You can't mix LP64 and ILP32 modes in the same binary.

The EMT64 instruction set in Core 2 Macs provide a substantial speed boost for typical Mac application code and a number of frameworks, despite the overhead (especially with respect to cache and memory) of 64-bit longs and pointers, so having full LP64 support on Intel processors that support it is a win.

There probably won't be much built specifically for LP64 on the PPC. Anything that needs to run on a G3 system be using SIMD stuff (Apple calls it "Velocity Engine") that's mainly in frameworks from Apple or in isolated functions supplied by the developer; this has been common practice since the G4 was introduced, and reinforced when the 970s's VMX (what IBM calls it) turned out not to be 100% compatible with the 74xx's AltiVec (Motorola's term). Most such code is runtime selectable, so there has been little call for fat ("universal") binaries to cover variations in the PowerPC family of chips.

On the Intel side, it's much more likely that there will be "universal" support in many items of code, since for many things which are compute intensive, EMT64 is a win, and unfortunately Apple sold a number of systems which can not process EMT64 instructions at all, and developers may want their goodies running on those too.

Re:I wish MS would come out with something like th

[zlogic]

If development is made simpler, programmers can focus more on algorithms and see what's happening more clearly. That's why we have expensive RAD software like Matlab - instead of writing garbage collection and array resizing, developers focus on how to do the job most efficiently. And after that is found out, the app can be rewritten (even with some help from Matlab) in C++ or other language.

Don't forget F-Script

[argent]

Don't forget F-Script, which lets you write Cocoa applications in a Smalltalk-like language and environment.

F-Script is free and open-source.

Re:I wish MS would come out with something like th

[Trillan]

About the main difference in Visual Studio 2005's C++ is that Code Insight crashes a lot. I mean a lot. Thankfully, the Whole Tomato tool mentioned elsewhere in the thread doesn't.

Emphasizing the Wrong Features

[99BottlesOfBeerInMyF]

Ruby on Rails and iLife integration? The former is for 10.5 server (which most people here don't care about) and the latter was announced months ago. I submitted a link to this information four days ago, but with focus on some more important features:

• OpenGL 2.1
• Automatically spawning a thread for OpenGL programs that feeds the GPU, allowing those programs that are CPU bound up to two times the performance when using multi-core systems, without any more work on the part of developers.
• Application signing to determine trust levels
• Mandatory Access Controls, for sandboxing applications like SELinux does

It is these last two that are of real interest. Individually they are just adding more security features under the hood, which most people will never notice. In that case it is great, but nothing too new. Together, however, they could be the groundwork for just the type malware/spyware defense some security people have been hoping for for years.

Imagine a system where all unsigned code runs in a sandbox by default, without access to any files it does not create, the internet, or any important parts of the system. Realistically, people want to run software they don't trust. They will run it. Most people don't understand the idea of multiple users as a security mechanism. It does not make sense to them that you need to create a new user account to sandbox an application and it is painful from a usability standpoint.

This announcement could be the first indication of the first real, usable desktop that has the benefits of some of the most secure workstations on the planet. Who cares about RoR tools in OS X server?

Re:I wish MS would come out with something like th

[Anonymous Coward]

Maybe SlickEdit is what you are looking for... Their "Mac" version requires X11 though.

Features Added & Defaults Changed

[buckhead_buddy]

cerelib asked:

What changed in 10.2 to leave 10.0 and 10.1 owners out in the cold?

The reason is mostly one of developer convenience, IMHO.

The format of NIB files and the bundle structure changed in 10.2. For developers, this was a huge difference. Localizing an app (as I recall) under Puma (10.1) was time consuming. The binary-only NIB file format before 10.1 meant you either needed to localize your app in your code or by manually opening the NIBs in interface builder and changing things. 10.2 allowed for text based NIB files, supported new user-visible GUI items, and was more resilient against corruption (at least anecdotally). There was a backwards compatible file-format for 10.1 but it was not the default. A developer could make a backwards compatible app, but what developer is going to go out of their way to make their job harder? Even today, it's possible to develop bundles and NIBs that are 10.1 and 10.0 compatible, but it's just not a desirable thing for developers to do.

10.5 will offer developers what I consider to be "major laziness options" so I expect it will become a mandatory requirement for new versions of OS X apps. Some of Apple's private GUI items users know from the iApps and Pro apps will be available in Interface Builder so developers don't have to make an inferior knock-off. Garbage collection in Objective-C 2.0 will make debugging your model far easier in debug and development, even if you decide to go back to manual memory management for performance in your release app. More unification of Carbon and Cocoa (e.g. NSViews in a carbon windows). In the past, things like the NSDocument architecture were great but for many developers it wasn't enough to force a rewrite of their app from the ground up. They'd already done the hard stuff, so why start over? But with 10.5 there are compelling developer reasons for new apps to NOT be backwards compatible and for older apps to rewrite their model or interface layer in 10.5 only code (even if they can't justify starting over with a clean slate). The biggest justification for 10.5 only apps will be developer laziness :-)

Re:I wish MS would come out with something like th

[sofla]

This is admittedly off-topic, but...the fix for the "error dialog about not having a manifest file to load" is to generate a manifest (linker setting and also the new "manifest tool") and use the new "assembly" based deployment model. The release builds of CRT in VS 2005 have a runtime check which throws an exception if you don't have either a manifest resource (embedded) or .manifest file (detached) for your application. In other words, M$ is FORCING you to use their new deployment model now, whether you want to or not. And just for the sake of doing it - the .DLL's themselves don't care how they are deployed as they still run on pre-manifest systems.

Yay assemblies! They're like frameworks, except without versioning and without the development time support! More effort to build and you get nothing out of it.

Re:I wish MS would come out with something like th

[Anonymous Coward]

You don't need to use much of a "deployment model" - just put the CRT DLLs (msvcr80.dll, etc) in the same directory as the application, and put Microsoft.VC80.CRT.manifest in there too. WinXP+ will still use the system's copy of the CRT if it's available (presumably because that's more likely to be patched), but it will fall back on the one in the application directory if necessary; and Win2K will just ignore the manifest and load the DLLs like normal. But nobody seems to explain that method - they all say "use the Microsoft Installer and the CRT merge modules" which is just a pain.

Cat Names

[buckhead_buddy]

akincisor wrote:

I'm sure they will run out of big cat names before [Mac OS X 10.10] :)

They've already deviated from the "big cat" naming trend. Pumas (10.1) are classified as small cats because they can not roar. Apple has applied for cat related trademarks including Cougar, Lynx, and Lion. That'd take us through 10.8. What about 10.9? I'm guessing:

• Sphinx - Monumental, mythic, and knowledgeable
• Tabby - Friendly and fully domesticated
• Simba - Co-marketing with Disney gone wild
• Sabertooth - Extinct (good name for a concurrent compatability release with 11.0)

Re:Cat Names

[XxtraLarGe]

Mac OS X 10.9 Liger - Bred for its skills in magic.

Developer DVD

[green pizza]

The Mac OS X developer tools are on the same DVD as Mac OS X itself. And, on new machines, the developer tools installer is already on the hard drive. So the truth is somewhere in the middle.

Comment removed

[account_deleted]

Comment removed based on user account deletion

More features less compatibility

[dvhh]

Of course new features are mostly welcome in most OS, but it can gives more headache for dev's to keep backward compatibility ( but of course most of the time, this is a feature that is most of the time replaced by a "require a newer version" dialog box). Of course modern api gives more security, usability and ease of programming but at what cost ? Isn't OSX secure enough for most user that don't feel the need to upgrade the OS but must because their software require a newer OS version (and their OS a newer hardware)?

Re: Link working?

[tehcyder]

If you just stuck with one OS and one browser maybe it'd reduce the load on the servers and you'd give people a chance to read the article

As the GP was obviously trying every possible combination of OS and browser simultaneously?

Re:I wish MS would come out with something like th

[NuShrike]

It is already so painful to even write code in Xcode. Sure the interface building is nice, but what's the point if you can't even get decent and relevant code completion and lookup that even begins to resemble VS intellisense 2002!

Don't even get me started on Xcode's crappy debugger (gdb-based) that can't even automatically give me reliable and decent variable value expansion and such. Why does I have to dig to find what a CFStringRef, or any ref, points to what actual data, sometimes? Why do have to dig through layers just to know what my basic_string<> is? Why does the debugger sometimes map to no code line except the very bottom of some random file while I'm stepping? Why does the debugger act strangely and go off into lalaland if I'm off by a CFRelease or CFRetain? Why do I even have to deal with CFRetain/Release and CFStringCreateCopyWithKitchenSinkInsideTheTitanic in today's world of namespaces, constructors and such?

Only thing that seems to offset all this hair-pulling is distcc for quick recompiles, with no such no easy alternative in MS world.

Re:I wish MS would come out with something like th

[AaronBrethorst]

Can you elaborate on why you feel that Xcode and Interface Builder are nicer to use than Visual Studio?

Re:I wish MS would come out with something like th

[andi75]

> Only thing that seems to offset all this hair-pulling is distcc for quick recompiles, with no such no easy alternative in MS world.

There's Xoreax Incredibuild, which we used at my old job. Reduced compile times for a full build from +30 minutes to less than 4...