31 replies to this topic

[stryder]

terry, on Sep 22 2005, 07:07 PM, said:

Obviously you've never seen any operating system in its alpha stage. Mac OS X for Intel is pretty much sophisticated and complete.

However, the librarys, kexts, compilers, etc., are all changing and evolving as Apple tweaks them.  Remember, they are now getting help from Intel for everything.  That may mean help with tweaking the code of the OS to be even more stable, or faster and more efficient.

In the glory days of OSX for PowerPC, when 10.0 came out, everytime an update was released (10.0.1, 10.0.2., etc.) it usually broke applications.  That was due to different things changing all the time.

Quote Ars:

Tiger also represents a milestone in Mac OS X's development process. Apple has promised developers that there will be "no API disruption for the foreseeable future." Starting with Tiger, Apple will add new APIs to Mac OS X, but will not change any existing APIs in an incompatible way. This has not been the case during the first four years of Mac OS X's development, and Mac developers have often had to scramble to keep their applications running after each new major release.

(end quote)

You can find the Ars review of 10.4 here:

http://arstechnica.c...macosx-10.4.ars

[Mr. Bond]

terry, on Sep 23 2005, 12:07 AM, said:

Obviously you've never seen any operating system in its alpha stage. Mac OS X for Intel is pretty much sophisticated and complete.

Quote

al·pha (ăl'fə)
n.
The first letter of the Greek alphabet.
The first one; the beginning.
Chemistry. The first position from a designated carbon atom in an organic molecule at which an atom or radical may be substituted.
Astronomy. The brightest or main star in a constellation.
The mathematical estimate of the return on a security when the return on the market as a whole is zero. Alpha is derived from a in the formula Ri = a + bRm, which measures the return on a security (Ri) for a given return on the market (Rm) where b is beta.

This is the first version of Mac OSX for intel, which has pretty much been released to anyone outside of Apple. It is very much incomplete, which can be seen through the lack of hardware/software support, and the many bugs and issues that exist withinin it. This was not released to developers for beta testing, but to give them a headstart at developing for the platform, and a rough outline as to how the final product will look and feel. It's like saying that house you bought doesn't have any finished rooms, but still looks like a house, so is pretty much the final product. These are what my views are, and yours may or may not differ, but I am not going to get into an argument over what each of believe should be considered alpha.

[buwie]

MrBond, on Sep 23 2005, 06:31 PM, said:

It is very much incomplete, which can be seen through the lack of hardware/software support

R U kidding? This is exactly what Apple doesn't care about, they only have to support their own hardware, remember? This release is way beyond alpha, no argument needed about that

[Pnutster]

terry, on Sep 24 2005, 07:42 PM, said:

The current public release of Mac OS X for Intel does not give a "rough outline" of Mac OS X's look and feel, but is almost on par with the production version for PPC that you can find in the stores.

Having 3 Macs running beside the Mac OS x86 I can pretty much agree... x86 look and feel is almost on par with PPC versions of OS X.

No doubt about it. Not an early Alpha stage. Years of x86 development are in this babe for sure!!! Didn't you hear Steve Jobs in his Keynote about this. Every single Mac PPC OS X version was 'shadowed' by an x86 version! From the first one on!!!

[Mr. Bond]

Quote

I find it quite startling that of all people a person like you who shows such gross incompetence is moderator on this board.

Being a moderator does not mean that I should have an opinion that everyone agrees with. The whole point of an opinions sub-forum, is so that people can state their opinions, whether people agree with them or not. I apreciate the fact that you, and others, don't agree with my opinion, as that is the purpose of this topic. Just because you believe I have an invalid opinion does not make you any more competent at moderating the forum than me.

Now, can we please return to the topic at hand?

[DannySmurf]

terry, on Sep 24 2005, 01:42 PM, said:

That's plain ridiculous. So you're insinuating a time frame in which five+ years are spent on developing in a crude alpha stage and then the product is suddenly pushed through beta testing and release candidates in about half a year?

Actually, it'll be about a full year by the time it's actually released with the Intel Macs in mid-2006. But yes, that's basically what's happening, and it's not unimaginable or even uncommon. Longhorn has been in development since 2001. We've seen (almost) four years of alpha builds, and the beta test that started a few weeks ago will conclude after about a year, and the product will be released. XP/Whistler had a similar (though shorter -- it was a much less ambitious project) development schedule. This is the way that software development generally goes. You don't spend six months in development and three years beta testing. When you do that, you get things like Windows Me.

I don't know how complete OSX for Intel actually is. I'm not an Apple developer -- neither are you. How sophisticated it is is certainly debateable (and if you're a long-time Windows user I certainly can't blame you for thinking that it IS sophisticated). However, until Apple says otherwise, it's incomplete, regardless of how well it works for you.

Reply to the original question: If it didn't boot, I guess you'd just stop using it, wouldn't you? But of course, by that time the actual OSX for Intel would be available, and you could use that instead.

[DannySmurf]

Quote

Longhorn is really a bad example. It has been scrapped completely and rebuilt from the ground up, with work starting to begin in summer 2004.

That's what Microsoft calls it, and for some reason people actually believe that that's what they've done. It's not. In reality, rebuilding an operating system as complex as Windows "from the ground up" would take a decade, at least. What they ACTUALLY did was swap out the XP core that they were using pre-2004, and plug in the Windows 2003 core. It was certainly a good move. But considering 1) the two systems are designed to be compatible right down to the binary level (even kernel-mode -- WDM -- drivers are swappable between the two) and 2) the sheer number of developers working on Windows, this swap was also a very trivial move from a development perspective.

Anyway, that's very off topic and you don't really have to reply to it. It just irritates me when people spread the Microsoft line about "rebuilding from the ground up" and other associated garbage.

Quote

Actually MrBond talked about Mac OS X still being pretty much in its "alpha stages".

He did. But I wasn't replying to his post (or even saying that he was right). I was replying to the fact that you said that OSX for Intel is pretty much complete. My point (which you ignored in your reply) is that OSX for Intel is not complete, and is likely nowhere near complete, and -- most importantly -- you are not in a position to make that judgement; neither am I, neither is MrBond, because none of us work for Apple. It will be complete when Apple says it is.

Whether you or I think it's an "alpha" or a "beta" is also irrelevant. That's a label that developers assign to their software, not users. I realize that users like to try, but since user experience (the only thing that users can judge) is NOT the determining factor in assigning development-stage labels, what any user thinks about the matter is not relevant.

[Big Kahuna]

domino, on Sep 5 2005, 10:33 AM, said:

Lots of speculations and "what if"s. I personally didn't install OSX86 to cheat Apple from there hard work. Many of us now know the capabilities of OSX86 and know that it is very crippled ware. But as soon as the real thing is released to the public, you betcha i'm gonna runout and get one. Can you imagine the speed and potentials of MacTel? It would be great to have USB, WiFi, Graphics, and sound working 100%. Well worth the hard earned cash IMHO.

oh believe me, i'm going to 'make the switch' when it comes out.. but i've always beta tested on windows, now i want to see what osx86's going to offer. i'm excited, and anxious about it.


Yes, I'm making the switch.

[DannySmurf]

terry, on Oct 4 2005, 03:02 PM, said:

Frankly said, I don't know if it really was a walk in the park. If my recollection is correct, it took Microsoft at least about a year to do the shift with the 5xxx releases and arrive where the latest 4xxx revisions have been before, in terms of functionality.

I'm not really sure. I know it took a year between the two sets, but that doesn't necessarily mean they spent all that time moving the kernel over. And honestly, if it took them that long to do that, I'd be shocked. The two kernels, while radically different on the inside, are designed to look and act almost identically to outside code ("outside" code including parts of the OS outside of the kernel). If Microsoft's legions of programmers struggled with a move like that for a year, I'd be VERY hesitant about running this next version of Windows.

Quote

Actually I haven't seen a single announcement from Microsoft claiming that.

Neither have I. But Microsoft's statements STRONGLY imply that. And I believe Jim Allchin used a very similar phrase not all that long ago (I might be mistaken). People pick up the implication and repeat it, without realizing what they're actually saying.

Quote

I'm inclined to think that at least those developers working with the developer transition kit to port their applications to Mac OS for Intel will have at least a bit of an insight into how far the OS has evolved. I simply cannot believe Apple is encouraging folks to start porting their apps to an operating system that isn't yet there while supplying them with some fake previews, whose very basics are still in a constant flux of changing dramatically. Developers need something definite to rely on.

You're right about all of that, of course. But I think you may not quite understand how software development on a modern operating system works. Developers need something definite to rely on, yes. We call that the "public interface." Basically, you call into the operating system at a certain point and you get some sort of response. The entry point and the response should be consistent. However, what the operating system does behind the scenes to generate that response does CAN be changed without changing the developer's/user's view of the system. For example, the network stack in Windows XP is quite different from that in Windows NT4. They don't work the same way at all. But a developer only has to write a single piece of code to work on both systems. The interface is the same. But there are six years of changes under the covers that make the XP stack work better, faster, etc.

Quote

So you're basically saying that simple users cannot differentiate at least rough stages of sophistication in a software product?

Sure they can. But as I said, the user experience does not determine whether a product is "alpha" or "beta," or vice versa. Those words just don't mean what you think they do (in fact, what most people think they do). They have nothing whatsoever to do with the quality of a product. The quality generally falls in line with where the product is in development. But not always. And the fact that a product is in the "alpha" stage does not mean that the quality is poor. For example, ICQ 98 (99?) Alpha. It was an "alpha" release, but it was a MUCH more stable, usable and finished product than its "gold" competition at the time (Powwow, if you're interested).

"Alpha" means that the product is still in active development: new features are still being added, and lots of new code is being checked into the project. That CAN (and a lot of the time does) mean that the product is unstable or rough around the edges, but it doesn't have to. "Beta" means that the product is feature complete, or very nearly so, very little new code is being checked in, and the main focus has switched to debugging.

Microsoft has confused that a bit over the years, and of course all of the amateur developers and OSS developers that make up meaningless version numbers for their apps like "Version 0.7 Beta 2 Release Candidate 1" don't exactly help. (that's an actual version number, BTW; it was for a calculator)