Dharma: Cocoa on Windows... Again?

[steeldrivingjon]

Sorry your attitude hasn't changed. Besides, our conversation on the subject ended with the post that got you banned.

As I'm posting, I'm evidently not banned. I noticed today that I was evidently censored. I've been away on business all week, so hadn't posted.

The attitude you don't like appears to be that I know enough about OpenStep to call your statements into question, unlike most people hereabouts.

 

[rayz]

Me too. That's what I said - the *lame excuse* is "I don't wanna learn Objective-C",

Yep, and I said that I have not come across any developers who won't learn ObjectiveC because they are too lazy. Developers love learning new stuff; it's what makes the job interesting (IMO).

which is silly on its face, because learning the frameworks is far more work and would be required even if Apple used a more popular language.

Well, it still boils down to the same problem. why learn a framework for a such a small market, especially when the skills you pick up are not transferable anywhere else? Even if Dharma di show up, I'm not sure any non-Mac developer would risk it. Apple does tend to drop these technologies rather a lot.

The *actual* reason, if they bothered to consider it long enough, is the business stuff, and would exist regardless of the language used.

Ergo, if the root objections would exist regardless of language, then learning Objective-C is a non-issue.

Ah! I see what you're saying now!

But the framework and the language share the same problem. Cocoa is tied to ObjctiveC and ObjectiveC is tied to Cocoa. Apple has already tried to separate them using the Java bridge, which was a dismal failure.

Basically the problem is that for business reasons, developers won't learn ObjectiveC/Cocoa.

If ObjectiveC wasn't seen primarily as an Apple technology, it might be more popular (on the other hand, without real GC ...)

 

[Mechcozmo]

As I'm posting, I'm evidently not banned. I noticed today that I was evidently censored. I've been away on business all week, so hadn't posted.

The attitude you don't like appears to be that I know enough about OpenStep to call your statements into question, unlike most people hereabouts.

1. Censorship is what happens when you realize MacRumors is not a democracy. The mods have the power to do what they think is right to retain order. Chalk up a point for a dictatorship.

2. So far, the statements made by RacerX all seemed fine and dandy and your statements have been called into question. Added to the fact that you just insulted everyone here by saying that we didn't know enough to question about OpenStep, I'm a bit curious why you are seemingly trying to show that you know about OpenStep. The whole argument should be dropped, since it has degenerated this far into bannings, censorship, and overall annoyingness.

 

[rayz]

As I'm posting, I'm evidently not banned. I noticed today that I was evidently censored. I've been away on business all week, so hadn't posted.

The attitude you don't like appears to be that I know enough about OpenStep to call your statements into question, unlike most people hereabouts.

Actually, RacerX seems to have provided the evidence in his post.

I did not know for example, that Apple did indeed promise a Cocoa API for Windows.

And this is why 'laziness' is not the reason why folk don't learn ObjectiveC/Cocoa.

 

[RacerX]

As I'm posting, I'm evidently not banned. I noticed today that I was evidently censored. I've been away on business all week, so hadn't posted.

Oh... well, in case you missed it, you were banned.

Don't believe me? Well, Google seems to have cached a number of MacRumor pages (like this one) where you had the label "banned" under your name.

And as you seem to be unaware that you were banned, you may want to read the forum rules to insure that you are not... well, banned without you knowing it again.

The attitude you don't like appears to be that I know enough about OpenStep to call your statements into question, unlike most people hereabouts.

Well, then you would be wrong. What I don't care for is someone who is unable to maintain a civil debate.

If you care to retract your personal attacks (that is, remove them from all your posts) and provide a sincere (believable) public apology for those statements, then I'll give you the chance to try to prove... what ever it is you think you know about OpenStep.

Currently I don't even know that you have ever even touched NeXT software or hardware. To me you seem to be an anonymous poster who has anger management issues... and I have no need to discuss this issue further with someone like that.

You have the ability to start the discussion again... the ball is in your court.

 

[steeldrivingjon]

Currently I don't even know that you have ever even touched NeXT software or hardware. To me you seem to be an anonymous poster who has anger management issues... and I have no need to discuss this issue further with someone like that.

You have the ability to start the discussion again... the ball is in your court.

Try searching Google Groups for "jon hendry nextstep"

The earliest post is from 1992, when I had an afs.com address.

Later posts are from my steeldriving.com address, from my NeXT consulting business from 1995 to 1997. (Hence, my current steeldrivingjon alias). There's a Usenet resume I posted around that time.

And then there's a brief mention in a 1993 NeXTWorld magazine column on user group activites:

Mike Matlock and Jon Hendry, with the support of Greg Anderson, their boss at Anderson Financial Systems, have been storming a variety of PC audiences with the NEXTSTEP message. These leaders of PhANG, the Philadelphia Area NEXTSTEP User Group, also led the infiltration groups by doing a demo for over 100 members of the Philadelphia Area Computer Society last year. They did a general NEXTSTEP-for-Intel demo on their own that drew 60 people, including 30 new prospects, in May. And they were scheduled to speak to 150 PC users at a July meeting of the South Jersey PC User Group.
http://www.simson.net/nextworld/93.6/93.6.Oct.Community8.html

 

[steeldrivingjon]

Yep, and I said that I have not come across any developers who won't learn ObjectiveC because they are too lazy. Developers love learning new stuff; it's what makes the job interesting (IMO).



Well, it still boils down to the same problem. why learn a framework for a such a small market, especially when the skills you pick up are not transferable anywhere else? Even if Dharma di show up, I'm not sure any non-Mac developer would risk it. Apple does tend to drop these technologies rather a lot.




Ah! I see what you're saying now!

But the framework and the language share the same problem. Cocoa is tied to ObjctiveC and ObjectiveC is tied to Cocoa. Apple has already tried to separate them using the Java bridge, which was a dismal failure.

Basically the problem is that for business reasons, developers won't learn ObjectiveC/Cocoa.

But as I said, the same lackluster interest existed in the days of C/C++ Mac programming.

So Objective-C/Cocoa is a red herring. Even with a familiar language, interest remains low. Which means it isn't the language or the language used in the frameworks. C++ frameworks did no better.

You probably just didn't hear as many complaints back then because MacOS wasn't as attractive to people on other platforms as it is now. ;^)

 

[steeldrivingjon]

Actually, RacerX seems to have provided the evidence in his post.

I did not know for example, that Apple did indeed promise a Cocoa API for Windows.

I knew. Hell, I worked on it. (Though RacerX seems to think there's some significant difference between OpenStep/NT and YellowBox for Windows, though YellowBox for Windows when announced was nothing more than OpenStep/NT. It was mostly a name change. Cocoa for Windows would require Quartz, which never made it to Windows; OpenStep/NT used NeXT's Display Postscript.)

RacerX and I quibble primarily on the success of NeXT's support for four-way fat binaries. I say it worked wonderfully. He seems to consider it a failure.

And this is why 'laziness' is not the reason why folk don't learn ObjectiveC/Cocoa.

On the other hand, this was in the late nineties or 2000, when Apple was in a dramatically different position than it is now - a very tenuous position, with shifting plans, and everything was in question - including whether or not they'd manage to ship a new OS, or if it would get bogged down as badly as Copeland.

I suspect the only people bitter about this these days are old NeXT developers who had business plans based on the availability of a freely distributable YellowBox for Windows runtime.

There aren't many such people.

I had an idea for a business (or at least a project) which was killed by this. It would have been akin to Wikipedia, but in addition to the site there would have been a desktop application which could be used for editing, collaborating, and for keeping local on-disk subsets of the main encyclopedia. I figured that if the website was done with WebObjects, then the website and the desktop app could share a certain amount of code.

To reach critical mass (IPO! Hey, it was the bubble.) it would have required that the desktop app be free, and would have required a Windows version. When YellowBox for Windows was killed (let alone made non-free) that made the idea sufficiently difficult to implement that I lost what interest I had.

I still own the domain I wanted to use: mycyclopedia.com

Wikipedia pretty much sucked up all the oxygen in that space, so the window of opportunity (if there was one) has long since shut.

 

[wrldwzrd89]

I knew. Hell, I worked on it. (Though RacerX seems to think there's some significant difference between OpenStep/NT and YellowBox for Windows, though YellowBox for Windows when announced was nothing more than OpenStep/NT. It was mostly a name change. Cocoa for Windows would require Quartz, which never made it to Windows; OpenStep/NT used NeXT's Display Postscript.)

RacerX and I quibble primarily on the success of NeXT's support for four-way fat binaries. I say it worked wonderfully. He seems to consider it a failure.



On the other hand, this was in the late nineties or 2000, when Apple was in a dramatically different position than it is now - a very tenuous position, with shifting plans, and everything was in question - including whether or not they'd manage to ship a new OS, or if it would get bogged down as badly as Copeland.

I suspect the only people bitter about this these days are old NeXT developers who had business plans based on the availability of a freely distributable YellowBox for Windows runtime.

There aren't many such people.

I had an idea for a business (or at least a project) which was killed by this. It would have been akin to Wikipedia, but in addition to the site there would have been a desktop application which could be used for editing, collaborating, and for keeping local on-disk subsets of the main encyclopedia. I figured that if the website was done with WebObjects, then the website and the desktop app could share a certain amount of code.

To reach critical mass (IPO! Hey, it was the bubble.) it would have required that the desktop app be free, and would have required a Windows version. When YellowBox for Windows was killed (let alone made non-free) that made the idea sufficiently difficult to implement that I lost what interest I had.

I still own the domain I wanted to use: mycyclopedia.com

Wikipedia pretty much sucked up all the oxygen in that space, so the window of opportunity (if there was one) has long since shut.

This is very interesting, steeldrivingjon. I was completely unaware of these NeXT-related happenings you and RacerX unearthed for me. I have to thank everyone else (besides myself, of course) that contributed to the discussion of NeXT.

As for Yellow Box/Cocoa on Windows, judging by what I've read its fate depends on what Apple ends up doing with Quartz. Apple could make a "QUARTZ.DLL" and give it to Windows users, I suppose...there's only one problem with that: DirectX already uses QUARTZ.DLL for other purposes. I'm sure Apple will find a solution agreeable to all involved.

 

[Fukui]

As for Yellow Box/Cocoa on Windows, judging by what I've read its fate depends on what Apple ends up doing with Quartz. Apple could make a "QUARTZ.DLL" and give it to Windows users, I suppose...there's only one problem with that: DirectX already uses QUARTZ.DLL for other purposes. I'm sure Apple will find a solution agreeable to all involved.

Well, quicktime 7 has CoreVideo so porting quartz shouldn't be a problem either... the only problem is windows vista: OpenGL will run pretty badly unless its in full-screen mode, so Quartz-2D-Extreme/CoreVideo couldn't run as fast as the Vista native graphics... cocoa apps would be crippled visually.

 

[rayz]

Well, quicktime 7 has CoreVideo so porting quartz shouldn't be a problem either... the only problem is windows vista: OpenGL will run pretty badly unless its in full-screen mode, so Quartz-2D-Extreme/CoreVideo couldn't run as fast as the Vista native graphics... cocoa apps would be crippled visually.

When writing fo OSX, do you make OpenGL calls directly, or is there a Cocoa API for it?

 

[wrldwzrd89]

When writing fo OSX, do you make OpenGL calls directly, or is there a Cocoa API for it?

My understanding is that the Quartz layer handles all the OpenGL stuff. You make calls to the Quartz API (which can be used in both Carbon and Cocoa) to do the graphical stuff using OpenGL.

 

[rayz]

My understanding is that the Quartz layer handles all the OpenGL stuff. You make calls to the Quartz API (which can be used in both Carbon and Cocoa) to do the graphical stuff using OpenGL.

So could the same calls be mapped onto whatever Vista will use as its presentation layer?

 

[wrldwzrd89]

So could the same calls be mapped onto whatever Vista will use as its presentation layer?

I would think they could, assuming the needed wrapper code was already in place to enable this. However, as has been mentioned earlier, limitations of Windows Vista may cause sub-optimal performance if this is enabled.

 

[kainjow]

What if secretly, all this time, Apple has had a separate Quartz project that works perfectly with DirectX for when Cocoa on Windows is released?

 

[JFlac]

Not that I have any knowledge of Taoism of any sort, but did anyone notice Apple's new "Universal" (Binary) Logo? That's this old funny Yin and Yang thing (see the Wikipedia page for more info about the Yin and Yang in Taoism).

Now, maybe I'm reading too much into it, but "Dharma" is also a Taoist term (it's used in other religious and philosophical variants, but another page on the Wikipedia links it to Taoism in the Lost mythology, which is what started this rumor).

Is this logo a clue that more is coming?

 

[DeathChill]

I thought it was supposed to be revealed on the first day that the Intel Mac's are released? Oh well, maybe it will be revealed later, or not at all.

 

[David Sharpe]

Maybe

Guys, this is 100% fake. The guy put the project name as Dharma, and signed the email "John Locke, somewhere near Hawaii".

Watch LOST on ABC, and tell me it isn't fake.

Maybe the intent was to purposely use Lost reference to make us all question the validity of the letter.

You do realize the computer they use to input the numbers is an Apple ][.

hmmm

 

[Sunrunner]

Maybe the intent was to purposely use Lost reference to make us all question the validity of the letter.

You do realize the computer they use to input the numbers is an Apple ][.

hmmm

Maybe its all one big conspiracy and Steve Jobs is really an alien who is going to take over the world....

 

[yarghblargh]

I want to first apologize for resurrecting a thread which has been dead so long.
Now that that's over, let's get to business...

I'm a windows developer and have a substantial (well, comparatively speaking at least) mac marketbase who could benefit from my products. Has this rumor become true over the past year, or was it just hype?

 

[zweigand]

never materialized...

 

[peharri]

I want to first apologize for resurrecting a thread which has been dead so long.
Now that that's over, let's get to business...

I'm a windows developer and have a substantial (well, comparatively speaking at least) mac marketbase who could benefit from my products. Has this rumor become true over the past year, or was it just hype?

If it were to happen, it would be with the release of Leopard, and wouldn't be announced until Leopard is announced. So I wouldn't expect any news on that front for a while.

In the meantime, Java isn't as bad as it used to be (actually, these days it's pretty good) and applications written in recent versions of Java appear to look pretty integrated with the system. Whichever system that might be. As long as you code them properly at least.

Now, I know what you're thinking. Java. No native APIs. Just a subset of the goodies that both operating systems have to offer. Unfortunately, Cocoa on Windows, if it ever appears, will be the same. It's not really in Apple's best interests to, say, port CoreAnimation to Windows.

Supposedly the Windows Quicktime library is another alternative. As the original source for Carbon, it has a lot of undocumented Carbon APIs implemented, enough for developing a major application. But there's also a lot of glue code you'd presumably end up writing yourself.

If I wanted to attack both markets, I'd go with Java. That's the way the world is going anyway. Some careful programming and nobody will know the application you've written isn't native.

 

[iSee]

never materialized...

Yeah. If you look at iTunes on Windows, it's pretty clear Apple has some kind of cross-platform layer cooked up. Probably based on Objective-C and Cocoa API's. But I don't know whether Apple will ever release it or not. It is always easier to get these kinds of things working for certain apps or certain sub-sets of apps. It is quite another to get it working 100% for many, many apps.

And think about the support burden. Either Apple would have to throw it out there without much support, in which case it won't be widely used and could hurt Apple's rep. Or they would have to charge developers to use it.

So the only way Apple could justify something like this is if they saw it driving a lot of people to the Mac platform. I don't think it would have a huge impact, though. The Mac has a good selection of apps available in most of the areas where this is likely to increase the number of available choices.

It just seems to me like this would be expensive for Apple to support, and won't necessarily help sell a lot more Macs.

 

[peharri]

Yeah. If you look at iTunes on Windows, it's pretty clear Apple has some kind of cross-platform layer cooked up. Probably based on Objective-C and Cocoa API's.

iTunes is a Carbon application, so most probably iTunes for Windows is compiled against the standard Quicktime API, which incorporates a huge amount of Carbon (Carbon's origins start with the Quicktime for Windows APIs)

 

[yarghblargh]

Rats.
Perhaps the community should wrench the idea fropm Apple's hands (if it indeed ever was in them). I mean we (as in the multi-box community) have developed an entire PPC emulator (several, come to think of it), but we can't write a compiler?