[aseem]

As Microsoft CSA, Ozzie had to live up to the legacy of Gates while steering the direction of hundreds of product groups. This is certainly no simple task.

Ultimately, he was doomed to fail. His job was to advocate the right technologies at the right time. Yet MS rarely makes the right choices in this area. Technology decisions are made by political GMs who want to hold on to their empires. They are made by VPs who don't want to sacrifice short term profits for a longer term vision.

Steve Jobs made an excellent point at the D conference this year about how Apple tries to pick technologies that are in their "Spring". These are technologies that are on their way up. Certainly Steve has the luxury of not worrying as much about backwards compatibility, etc. But he certainly has the courage of his convictions to pick a path that's best for his company and customers.

I don't think Ray had that kind of fortitude, and unfortunately, I don't think he had that kind of power. Ultimately, the CEO needs to push the technological vision of Microsoft. Leaving it to mid-level managers will only result in further mediocrity.

[lkijuhygtfd]

Technologies in their spring?

A BSD kernel, a gui from Next and a object orientated version of C from before C++ !

Technology-wise Apple basically takes a steam train, wraps a Bang+Olufson case around it and makes it emit the scent of roses!

Nice toys, very well made, but cutting edge technology doesn't underpin Apple's success.

[gvb]

Exactly! Jobs picked BSD (1977), the NeXT GUI (1985/1988), and an object oriented version of C (1986) that came before C++ (1983)! (OK, BSD had been around a little while by that time, and C++ predated Objective C.) They are mature technologies now, but were all picked relatively early in their life.

http://en.wikipedia.org/wiki/Berkeley_Software_Distribution

http://en.wikipedia.org/wiki/NeXT

http://en.wikipedia.org/wiki/C%2B%2B

[aseem]

You certainly note some fair examples for your point. However, I was referring to Apple's choosing of HTML5 instead of Flash. The easy choice that probably benefits mostly in the short-term is to play nice with Flash. Ultimately, Jobs and his team felt the best customer experience was to ditch Flash much to the malign of developers. Would Microsoft made such a decision? And who at Microsoft makes those decisions? Where does the buck stop?

[patrickgzill]

I would like you to explain a little more how C++ is more "cutting edge" than Obj-C, etc.

BSD is still used in Free/Open/Net/BSD; and Next's GUI was based on Display Postscript, which was easily tuned for PDF (PS and PDF are very similar languages)

What you will notice is that all of these choices were designed by very small teams, sometimes just 1 person was involved in the original design:

BSD - Bill Joy and a few others made the major design decisions

Next GUI - Keith Ohlfs, on top of DPS licensed from Adobe

Objective C - designed mainly by Brad Cox

You missed the Mach kernel, originally a small research project from CMU, Avie Tevanian worked on porting Mach to a multi-processor system in the early to mid 1980s.

[slantyyz]

Can't fault you on your points -- they're all solid, but I would say there's a glaring omission. Cutting edge user experience (and that includes usability) does underpin Apple's success.

[lkijuhygtfd]

Precisely - it's the user experience that puts Apple out ahead not cutting edge technology.

In fact their technology is and always has been (when did Mac-0S get real multi-tasking?) rather conservative.

[rimantas]

  Precisely - it's the user experience that puts Apple out
  ahead not cutting edge technology.

Would iPhone user experience be the same with resistive touchscreen? Were there any capacitive touchscreen phones before iPhone?

[jonhendry]

"(when did Mac-0S get real multi-tasking?)"

As soon as possible after Jobs' return.

[jonhendry]

They (and the MACH kernel) were new-ish when Jobs picked them in the 80s.

He kept them because they work well.

[rimantas]

Yep. Capacitive touchscreens, unibodies and state of the art battery tech don't count.

[ciupicri]

> Certainly Steve has the luxury of not worrying as much about backwards compatibility

Not all the time. Let's not forget about the Universal Binaries (a solution for the transition to Intel) and the Carbon API:

Carbon provides backward compatibility for existing Mac OS X software, while serving as a stepping stone for developers porting procedural applications from other platforms. [1]

[1] http://developer.apple.com/carbon/

[Zev]

Let's not forget about the Universal Binaries (a solution for the transition to Intel)

And from 68k to PPC.

[jonah]

Weren't those called "Fat Binaries"?

[Zev]

A "Universal Binary" is the exact same thing as a "Fat Binary," the only difference is that marketing gave it a new name.

If you look at the mach-o header of a binary that contains multiple architectures (initially, ppc/68k or now, i386/x86_64/ppc), the first 8 bytes will be either 0xcafebabe or 0xbebafeca, depending on your endianness. These values are also #define'd as FAT_MAGIC or FAT_CIGAM in <arch/fat.h>.

If you're on a Mac, you can check this by typing `open -h fat.h` into a Terminal and opening up the binary of a universal app in a hex editor.

[jonah]

Yep. But I don't believe the 68k/PPC ones were called "Universal Binary".

[Zev]

If it walks like a duck and quacks like a duck...

Please see the end of the first line in my previous post: ...the only difference is that marketing gave it a new name.

[jimbokun]

"His job was to advocate the right technologies at the right time. Yet MS rarely makes the right choices in this area. Technology decisions are made by political GMs who want to hold on to their empires."

So he had much responsibility and little authority, always a combination that guarantees failure.

[slantyyz]

Agreed, I never really saw him as much more than a figurehead.

In many ways, Microsoft is such an immovable beast, you can't expect many people outside of a Bill Gates to have the gravitas to make things happen.

[allwein]

>always a combination that guarantees failure.

Well, not always. I'm actually reading a book on the subject right now called "Results Without Authority"

http://www.amazon.com/gp/product/0814473431?ie=UTF8&tag=...

[auxbuss]

The CEO must push/sell the technological vision, but he must not make it. Not in a company the size of Microsoft.

I won't miss Microsoft, but their inability to define a vision, communicate, and execute it is voyeuristic to observe.

That said, Ballmer needs to walk, and do it quickly. It's over.

[rbanffy]

> I won't miss Microsoft, but their inability to define a vision, communicate, and execute it is voyeuristic to observe.

It's like watching a trainwreck in slow motion from a thousand different angles, isn't it?

(burn, karma, burn)

[elblanco]

> I don't think Ray had that kind of fortitude, and unfortunately, I don't think he had that kind of power.

I'm not even sure, looking at old internal Bill Gates emails and memos, that Bill had that kind of power.

[ergo98]

>Ultimately, he was doomed to fail.

Ozzie was essentially given reign to plow some green fields. He failed miserably at giving Microsoft any credibility or initiative.