Now, on this JLG issue, he (Daniel) is spot on. I don’t want to reveal too many personal details but suffice to say, BeOS could have been great. Could have been, because it was coded by very smart people. But therein lies the problem (which you can see with Linux) where smart codes are also clueless about what real humans want and need. There’s a place for experimental labs (like, Xerox PARC) to grow ideas, but there’s another kind of discipline needed to turn things into mature products. Now, mix this up with people that probably had less need for beautiful and profitable technology than being found out for the clueless posers that they actually are (and, while at it, ideally make a windfall killing off other ignorant loons), and you have the formula for wonderful failure.

I knew the whole can of worms would go nowhere when Palm (and then Access) bought the IP because I could see many of the same people involved.

And they are still in the industry…

http://www.nextarchive.net/

http://blackholeinc.com/specials/survey.shtml

http://www.rhapsodyos.org/home.html

http://www.stepwise.com/Articles/FirstImpressions/MacOSXServerIntro/index.html

http://web.mit.edu/afs/dev.mit.edu/user/cfields/apple/

Gassee’s major problem, and it was apparent even back when, was he was primarily interested in high end stuff (the Mac IIfx being the most glaring example). But those were the days when SGI was riding high and Apple and SGI were fighting for customers (or thought they were… instead Microsoft and 3D Studio were slowly eating SGI’s lunch).

Also remember that Apple was built on using engineering to build compelling products with ridiculous profit margins and then milking them for all they were worth. It also repeatedly suffered when those products were eventually undercut because it would never “stoop” to competing at the low end (witness the Apple ][ floppy drive (eventually undercut by the Commodore 64), or the Mac IIcx which, both of which were sold at something like 90% margins). This isn’t a model that appeared when Jobs left Apple. It was the model he left Apple with. The difference between Steve and the people who tried to replace him is that he knew when to change tactics.

BeOS was actually pretty innovative in many respects, while having some glaring oversights (no printer support *cough*). Its file system still looks like the Holy Grail compared to HFS+ with Spotlight (and the guy who wrote it was hired by Apple to do Spotlight, but I guess retrofitting the functionality is harder).

Be: pipedream. NeXT: today.

“given how much time it took Apple to turn NeXTStep into OS X, BeOS could have very well been built into a brilliant Apple OS. I know there are other factors involved, like the false illusion that Rhapsody was the light at the end of the tunnel helping the company, the Unix factor, the very fact that Apple needed Jobs to recreate the lost mystique, etc. But the thing is, Apple probably spent in OS X redevelopment far more than Gassee ever asked for selling them BeOS.”

I was at Apple in 1996-1997 when NeXT was acquired for $400 million. The deal was announced on Dec. 20, 1996 and finalized on Feb. 7, 1997: http://web.archive.org/web/*/product.info.apple.com/pr/press.releases/1997/q2/970207.pr.rel.next.html

Part of that press release reads:

“The first release of Rhapsody is expected to be launched to developers in mid to late 1997 and to customers within 12 months. A unified Rhapsody release is expected to be in the hands of customers by mid-1998. This will include compatibility with existing Mac OS applications, as well as provide a platform for next-generation computing.”

I worked on Rhapsody in 1997. The first developer release did indeed occur in August of 1997, a mere six months after the NeXT acquisition was finalized. The first customer release “within 12 months” never happened. Instead, because major developers such as Adobe said they were never going to port their software to an entirely new OS such as Rhapsody, in the spring of 1998 Apple announced the change in tactics that led to the Carbon APIs and Mac OS X.

Rhapsody was still finished, though, and released in March of 1999 as Mac OS X Server 1.0. Apple missed the “mid-1998″ target by a few months, but by that point they were splitting their resources between Rhapsody, the Carbon APIs for both classic Mac OS and Mac OS X, classic Mac OS itself (which Apple had expected to mostly stop working on), and all of the newer post-Rhapsody work on Mac OS X, such as Quartz.

When Mac OS X Server 1.0 shipped, it was actually exactly what Apple had promised for Rhapsody: “compatibility with existing Mac OS applications, as well as … a platform for next-generation computing.” Mac OS X Server 1.0 had the Blue Box environment, in which you could run a full version of Mac OS 8.5.1. The only integration the two OSes had, though, was a shared clipboard and shared disks.

So really, Apple DID achieve what they originally set out to do when they bought NeXT, and it only took about two years (Feb 1997 to March 1999). No doubt it would have happened sooner than that had they not split their resources so much among three different OSes in the spring of 1998.

If Apple had bought Be, Inc. instead, no doubt it would have taken much longer to turn it into as full as an OS as even Mac OS X Server 1.0 in March of 1999. Plus they still likely would have had to go back to the drawing board and come up with something like the classic Mac OS/Carbon/Mac OS X scheme to satisfy the major developers.

In the end, what probably would have happened would have been that a lot of Mac OS code would have been ported to BeOS to fill the holes it had in features that were completely absent, and it would have ended up as much more of a mish-mash, with none of the clearer separation between Carbon and Cocoa libraries in the present Mac OS X. In that, I imagine it would have wound up dragging in a lot more cruft from older code that would be impossible to get rid of.

I think an important point to note, as well, is that it was probably largely due to the robust existing code base of NeXTSTEP/OPENSTEP that Apple could have achieved what they did with changing directions midstream from Rhapsody to Carbon/Cocoa and Mac OS X, along with all of the subsequent technologies they’ve added to it.