macOS Carbon vs. Cocoa: A structured debate

[Eraserhead]

So you are saying EVERY program written in C is unreliable and not worth running?

Of course not, look lets compare two different markets, I don't think I'm an expert but anyhow. In scientific computing you use one of two languages C or Fortran, you use those as they are fast and well established, in this reliability is important but at the end of the day if you lose a few hours due to a software crash you can recode the software and try again, someone else can use the supercomputer while you are dealing with your crash and its not a big deal compared to the speed advantage which allows you to model the situation better in a given time period. Of course if you crash the whole computer then its a big deal, but they use UNIX so it's unlikely to occur.

Whereas in Finance compared to the amount of money involved the computations aren't that computationally expensive, and if the system goes down you can lose millions and millions of dollars very quickly. So they don't need quite as fast a computer but it needs to keep running over and over, so Java and C# are better languages for that.

They are also the most corrupt.

Well yes, but they are a very successful software company and make a lot of money, which was my point.

EDIT @Netscape, on the Mac in 1998 we had IE and Netscape, I used IE because it was a better product, now it isn't but it was actually better then.

 

[jsw]

That being said, is there some, definitive references, the functional equivalent of the old Inside Macintosh volumes, for Macintosh Cocoa programming?

Yes. The online examples and Xcode help are vastly superior to the old Inside Macintosh volumes, which were so comprehensive and dense that they actually made it more difficult to learn anything. I mean, seriously, they were over a foot of shelf space, printed on incredibly thin paper (with ink whose smell I still cannot forget), and that was for an OS substantially less complex than today's.

The example programs available should be more than sufficient. The one thing I'm waiting for are Objective-C books that are updated with the improvements to that language that are in Leopard.

 

[Fukui]

That being said, is there some, definitive references, the functional equivalent of the old Inside Macintosh volumes, for Macintosh Cocoa programming?

And, is there an expected "drop dead" date when Carbon apps won't run on a Macintosh?

Have you logged into apple.developer.com and gone to the documentation section under "Cocoa?" There are many reference papers for cocoa. Look on Amazon.com if you can't find enough info. Aaron Hilegass has some excellent stuff.....

If your looking for a single definitive reference, your not gonna find it, cocoa is evolving so fast with each release there has been no time to make a "definitive" guide... the best you can do is read all the latest material, starting with Intro to Objective-C (2.0 if you can access it), Intro to Cocoa Programming, Intro to Core Data etc...

 

[kainjow]

And, is there an expected "drop dead" date when Carbon apps won't run on a Macintosh?

It looks like Apple is still adding features and improving Carbon, but not at the rate of Cocoa. No one knows when it will be unsupported - probably not for a long time. But they will probably stop adding features to it "soon". I think it will go the way of QuickDraw: it will continue to be part of the OS for several more versions, but will become deprecated in favor of Cocoa.

 

[janey]

My check was for a great deal more than $2.56, which is basically chicken feed considering Knuth has people "working for free" as proof-editors, a pretty good scheme.

The value of a knuth check is not in how much the check is for (btw it's $2.56 for every mistake, not every check is for $2.56) but rather in that you found a mistake in one of his books, which are not nonexistant, but still pretty difficult to find. Hell, I can hardly read some of the chapters in TAOCP, let alone be in any position to find any mistake. Something about MIX assembly doesn't appeal to me.

...regardless, you'd have to be a total idiot to cash it, or to want it for the money.

 

[GothicChess.Com]

Have you logged into apple.developer.com and gone to the documentation section under "Cocoa?" There are many reference papers for cocoa. Look on Amazon.com if you can't find enough info. Aaron Hilegass has some excellent stuff.....

That was partly my point, seems like there are "too many"...

If your looking for a single definitive reference, your not gonna find it, cocoa is evolving so fast

... again, that is part of the problem, trying to catch a running train...

 

[kainjow]

That was partly my point, seems like there are "too many"...


... again, that is part of the problem, trying to catch a running train...

Just grab this book and start from there. Don't worry about Tiger/Leopard-only technologies. Just learn the language and the basic concepts first.

 

[GothicChess.Com]

The value of a knuth check is not in how much the check is for

There is no value to me at all. Do you put on your resume: "I have a Knuth check?" I mean, who cares, seriously.

I only mentioned I have a check from Steve Wozniak because he actually bought a software program that I wrote. He even mentions this on his website, woz.org, here:

http://www.woz.org/maclinks/cool.html

 

[janey]

There is no value to me at all....

Why on earth would you mention it on a resume?

You have little interest in the Knuth check, and we have little interest in your Woz check.

To quote you..."who cares, seriously"? It was irrelevant when it was first mentioned.

 

[GothicChess.Com]

It was irrelevant when it was first mentioned.

The relevance was, some clown was implying because I am not a diehard Cocoa programmer that I had never coded anything of significance.

Getting a check from the co-founder of Apple who purchased something I had personally authored completely shot down his pithy remarks.

Knuth has no relevance compared to Wozniak with regard to this forum.

Your mention was the one that was irrelevant.

 

[janey]

The relevance was...

Nope, it was still pretty irrelevant. It's like saying I'm a die hard mac fan and even more so cause I met Woz. Hm, not really.

Doesn't give it much significance either.

 

[GothicChess.Com]

Just grab this book and start from there. Don't worry about Tiger/Leopard-only technologies. Just learn the language and the basic concepts first.

Thanks for that tip, and thanks for contributing something MEANINGFUL to this discussion, unlike some of the blatant pissant trolls who have nothing better to do.

I'll check it out, although I think the learning curve for this Cocoa stuff is probably fairly steep.

 

[SC68Cal]

Jesus, you are such a whiner Chess.

You complain about picking up a new programming language? Are you serious?

You must have slept through Computer Science 101, where they tell you "Don't worry about the specific implementation of the language, you need to learn the concepts behind the programming language."

In the space of a few months I went from some Java, to PHP & MySQL and learning some basic C as a hobby.

I don't have any problems picking up a new language. BECAUSE THEY ALL FOLLOW THE SAME FUNDAMENTAL CONCEPTS.

No wonder you're writing a chess program.

 

[Eraserhead]

I'll check it out, although I think the learning curve for this Cocoa stuff is probably fairly steep.

I love the book, you can do the extensions to stretch yourself or not depending on how you feel.

FWIW I didn't find it too difficult and beforehand the only programming I had done was a bit of Applescript and about 6 months of Java.

No wonder you're writing a chess program.

It does seem like a fairly good Chess program though .

 

[GothicChess.Com]

It does seem like a fairly good Chess program though .

In the history of computer chess, it is well known that Ken Thompson's $600,000 (1982 prices) Belle program was the first ever artificial intelligence player to make Master in United States Chess Federation play.

What is not well known, is that when I was still 20, my own Macintosh chess program, The Sniper, was the first software program to break the 2200 Master barrier, just 4 years after Thompson. (Belle was a hardware program, a dedicated machine, funded by Bell Labs.)

It did so on a Macintosh 512 KE, running on an 800K floppy disk that also had the operating system on it as well as the program! (I didn't have a computer with a hard drive yet, can you even imagine??) And, I did this in my spare time, while double majoring in Electrical and Computer Engineering

Later, other hardware machines broke master, like the Fidelty Mach III and the Fidelty Mach IV Master, but The Sniper was about 5 years ahead of the next software program to break Master.

 

[Eraserhead]

What is not well known, is that when I was still 20, my own Macintosh chess program, The Sniper, was the first software program to break the 2200 Master barrier.

If you are as good a programmer as what you've said implies you'll have no problems learning Cocoa, and I suspect you'll like it, pick up the Hillegass book and give it a try.

 

[garethlewis2]

Gothic you want an example.

Take ADP. They bought the company I worked at. Wilco.

Look at Wilco on ADP's site. Then look at Calypso.

See which one has all the customers and makes the money. It isn't ADP and it isn't Wilco.

All investment banks moved over from C to Java or C# to get away from the common fact that nearly 50% of bugs in C programs are memory related issues. With C++, 50% of problems are subtle memory related issues, usually concerning when to use references, pointers and copy constructors. If you aren't using an equivalent of garbage collection in C++ your program will always contain subtle memory errors, unless you are writing the noddiest of programs.

 

[GothicChess.Com]

Gothic you want an example.

Take ADP. They bought the company I worked at. Wilco.

Look at Wilco on ADP's site. Then look at Calypso.

See which one has all the customers and makes the money. It isn't ADP and it isn't Wilco.

All investment banks moved over from C to Java or C# to get away from the common fact that nearly 50% of bugs in C programs are memory related issues. With C++, 50% of problems are subtle memory related issues, usually concerning when to use references, pointers and copy constructors. If you aren't using an equivalent of garbage collection in C++ your program will always contain subtle memory errors, unless you are writing the noddiest of programs.

ADP prints about 1 out of every 3 paychecks on the continent of North American, and payroll is not even their core business. I'm sure they're making some money!

You can take the cushiest development environment in the world, in the hands of a bad programmer, you will still crash with graphic animation and sound

I'm sorry, but if I can solve a checkmate in 268 moves in C (see http://www.gothicchess.com/javascript_endings.html ) there ain't no way some rinky-dink little financial emulation is gonna make me run out of memory.

 

[GothicChess.Com]

If you are as good a programmer as what you've said implies you'll have no problems learning Cocoa, and I suspect you'll like it, pick up the Hillegass book and give it a try.

I'm just lazy now in my old age of 40...

I'll definitely pick the book up after my next vacation.

 

[Compile 'em all]

rinky-dink little financial emulation

There is no such thing as little financial applications. If you think that since you coded a chess app you can easily implement an application that supports scalability/replication/transaction_management/pooling then you are dreaming.

Every programming language is good at doing somethings and not others. C is a good choice for a coding a kernel but not a good one for a web application. A good programmer will use the right tools to get the job done.

 

[GothicChess.Com]

There is no such thing as little financial applications. If you think that since you coded a chess app you can easily implement an application that supports scalability/replication/transaction_management/pooling then you are dreaming.

Every programming language is good at doing somethings and not others. C is a good choice for a coding a kernel but not a good one for a web application. A good programmer will use the right tools to get the job done.

If you were correct, I would agree with you

I have written programs that play checkers, chess, and Gothic Chess, a much harder version of chess with more pieces [see http://www.GothicChess.com for example, and if not interested for the chess game, then maybe the 6'5" tall blonde that I hired will catch your eye ]

I also owned a software development company on Long Island and a computer consulting business. I could code financial management apps standing on my head underwater faster than most secretaries could type.

Trust me, financial apps are just lengthy and tedious to maintain. Chess programs require a much higher degree of precision and are much more detail-oriented.

Once you write your first chess program, let me know.

 

[garethlewis2]

So how did you define your pricing models?

Considering they have a varied amount of data, and you are looking to exploit the market.

The only people I know that have done this earn 9 million a year and work for Nomura, Meryl Lynch, Goldman Sachs and RBS. The 9 million is their salary and they can earn more than 30 million in bonuses. These guys could not code a financial simulation in the time it takes their secretary to type a letter. The mathematics required is so complex they don't hire anyone unless they have PHd in Theoretical Physics or Mathematics. The coding part is done by drones, but the interesting part of modelling a gigantic simulation with millions if not billions of variables is modelled by PHds. If you could write something like this and were not working for an IB, then you are without a doubt the smartest person on the planet.

 

[GothicChess.Com]

So how did you define your pricing models?

Considering they have a varied amount of data, and you are looking to exploit the market.

The only people I know that have done this earn 9 million a year and work for Nomura, Meryl Lynch, Goldman Sachs and RBS. The 9 million is their salary and they can earn more than 30 million in bonuses. These guys could not code a financial simulation in the time it takes their secretary to type a letter. The mathematics required is so complex they don't hire anyone unless they have PHd in Theoretical Physics or Mathematics. The coding part is done by drones, but the interesting part of modelling a gigantic simulation with millions if not billions of variables is modelled by PHds. If you could write something like this and were not working for an IB, then you are without a doubt the smartest person on the planet.

I told you a million times, don't exaggerate

I have published a few artificial intelligence papers, appeared in a textbooks in some CompSci Departments' libraries, two domains usually reserved for PhDs, of which I have none.

The Wright Brothers were bicycle mechanics, yet they invented the field of Avaiation.

Albert Einstein was a Swiss patent clerk, yet he published "Relativity" and demonstrated that the Newtonian Physics metaphor was lacking a great deal.

I think in today's day-and-age people are way too hung up on "credentials", but here are a few of mine:

The book I am in:
Advances In Computer Games 10 (ISBN 1-4020-7709-2)
http://www.springer.com/west/home?S...HORTCUT=www.springer.com/prod/b/1-4020-7709-2

My papers:

http://www.gothicchess.com/80.pdf
80-SquareChess. ICGA Journal, Vol.27, No.2,pp.81-95.
Notable for having solved a "Holy Grail" of chess -- how to mathematically compute the values of any chess piece, given a rectangular board of any non-zero dimension.

The Perfect 7-Piece Checkers Database.
ICGA Journal, Vol.26, No.4,pp. 229-238.
Notable for proving that it is possible to be in a winning checkers position with as few as 7 pieces on the board, and every other program in the world could not carry out the win, since the solution was far beyond a contemporary horizon of search (253 plies total). The only solution was to map every one of the 19 billion 7-piece endings all the way to mate, not "just" know if they were theoretical wins, losses, and draws. This undermined 40 years of A.I. thinking, and had a great deal of impact among researchers that have always used checkers as their software "Drosophilidae".

 

[lazydog]

I've been following this thread because I thought I could learn something about the future directions of Mac development, but I can't remain silent anymore… sorry!

I think in today's day-and-age people are way too hung up on "credentials", but here are a few of mine:

Oh man… this is such an uncool thing to do!!!

As for Chess programs… I'm sure that most of the regular contributors to this forum could write a half decent chess program if they wanted to.

b e n

 

[GothicChess.Com]

As for Chess programs… I'm sure that most of the regular contributors to this forum could write a half decent chess program if they wanted to.

b e n

Will all due respect, you have no clue what you are talking about! While there are many chess programs out there, that in no way trivializes what is required in terms of the skillsets necessary to make an outstanding one, let alone one that wins a Computer World Championship.