Welcome to the MacNN Forums.

piracy · Jul 25, 2002, 12:36 PM

Millennium,

You are doing this entire forum and everyone who reads it an immense disservice by saying what you did in your posts <a href="http://forums.macnn.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=46;t=007911" target="_blank">here</a>. It's even worse, since you're in a trusted position on this forum...to propagate something that is false will only create more difficulty that we already have here.

<blockquote>quote:<hr />But I stand by my words. This iMac was at the Apple Store at Tysons Corner, VA. As you enter the store, it's on the table to the right of the cash register, closest to the front of the store. Go there yourself if you don't believe me.<hr /></blockquote>I'm not sure what your point here is. I don't think anyone is disputing that Apple is running 6C98 on machines in its stores. After all, it's the build Jobs and other Apple folks used at Macworld, it was running on all the Macworld floor machines, and overall, it's a very nice, solid build. Apple wants people to see Jaguar! And 6C98 is a great build to do it with. But it's not GM, FC, or anything else. No one disputes that you saw it in an Apple store.

<blockquote>quote:<hr />And I still say there was more than one 4K78.<hr /></blockquote>This is probably the most irresponsible thing I've ever seen you say, and instantly made me lose respect for you as a technical expert here. Though this has been beaten to death numerous times, I feel I must again here for our new members:

For reference, see <a href="http://forums.macnn.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=3;t=003976" target="_blank">here</a> and <a href="http://docs.info.apple.com/article.html?artnum=106176" target="_blank">here</a>.

The very purpose of a build number in software development is to singularly and uniquely identify a build of a product, for numerous, numerous reasons: support, QA, customer issues, testing issues, version control, versioning of aggregate products, versioning of components within the product, etc. The build number is the single, unwavering reference to the exact state of the product you're working with. Its very existence is, by its nature, designed to be tied to one and only one unique build.

You should be, I hope, smart enough to understand that. Apple even went so far as to publicly state which build numbers were associated with which product versions. Even if you don't believe anything in the last paragraph, which is somewhat akin to believing the earth is flat and refusing to acknowledge all evidence to the contrary, that action alone proves that the build numbers are tied to one and only one build. Before you say that Apple doesn't have control over some builds that may have leaked and they're only referring to officially released software, why would apple say 4K78 == 10.0 when Apple had released 4K78 through two other legitimate channels other than retail sales (ADC and Software Engineering's Customer Seeding group)? If a customer's "About This Mac" box says 4K78, that's Apple's only way of knowing what precise version they're running. If there were multiple builds associated with single build numbers, how would a manufacturer ever support a product? In a typical software development environment, even rebuilding the product with NO CHANGES WHATSOEVER increments the build number (no, I'm not saying that some little fairy comes along and changes the number and that no one can stop it; I'm saying that it's a MATTER OF POLICY, and there's no logical reason to sabotage a project by having multiple builds with the same number...it would be disastrous to the product development cycle). For what it's worth, when Apple does make very small changes that don't merit and entirely new build number in its planning, it adds a lowercase letter to the end of the build, e.g., 5L17b.

Additionally, we've already <a href="http://forums.macnn.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=3;t=003976" target="_blank">proven</a> that the initial developer RC 4K78 and the retail boxed 4K78 are the same with an md5 checksum, which there is no known way of faking. I don't know what more you want.

Lastly, this isn't some kind of "well that's your opinion, and I have mine" type of deal. This is pure fact, and it's like you arguing that the sun is really a big ball of melted cheese vs. astounding known evidence and facts to the contrary. That alone isn't so bad, but you've got a lot of other people who read these forums that also believe the sun is really a big ball of melted cheese...not to mention believing in Apple conspiracy theories and "magical Steve Jobs non-debug builds" to boot.

There are no magical builds. There is not more than one build associated with a build number. Their very purpose is to uniquely identify a build. If you believe otherwise, well, you proabably believed a lot of stuff on Fox's TV show Millennium, too.

[ 07-25-2002, 12:44 PM: Message edited by: piracy ]

oranjdisc · Jul 25, 2002, 12:52 PM

Guys, it's time to unplug and have a beer.

Millennium · Jul 25, 2002, 01:02 PM

First you ask me to lock a thread, and then as soon as I do, you continue it?

I stand by my words. I have a legitimate developer seed of 4K78. I also have the final release of 10.0. I have compared them on the same machine. There is a difference. I don't claim to know how this is possible. Certainly two different files can have the same MD5 sum, but I see no point in Apple going to all the trouble of deliberately making that happen. But neither do I see any point in seeding a build, and then finally releasing that same build four months later without any new seeds in between. What is the point?

Very likely, they are the same source code, simply compiled with different options; that change could easily be considered insignificant enough to keep the build number. But there is a difference. I have seen it with my own eyes. While they may have the same number, there is no way the OSX CD I got on March 24th, 2001 and its four-month-older seed of the same number are the same build.

But enough of this. You guys don't want to discuss this, fine.