Comment by breadwinner
7 days ago
General Magic and Magic Cap [1] are not mentioned in this article.
If you know anything about Microsoft, you know they don't innovate. That was true in the 90's when Bill Gates was running Microsoft, and it is true under Nadella. Anything they do is ALWAYS in response to a competitive threat. So what was the competitive threat that spurred Microsoft Bob? It was the "social interface" of General Magic's Magic Cap operating system. When that flopped Microsoft cancelled Bob.
I think General Magic might've been onto something. From what I've seen, BOB looked so childish that it was likely insulting to adult users. General Magic had a crisper UI (partly because of the b/w nature of the devices) that felt more like the iconography of a late-80s copy machine.
I've got an old Sony Magic Link; one of the devices running the Magic Cap software. Both suffered from similar problems.
For starters the spacial interface is so cumbersome it makes all interactions with the system tedious. The first time you walk through the system it's cute but when you need to painstaking navigate to a particular room to do something it's just frustrating.
The hardware could not keep up with the demands of the interface. The PCs that shipped with Bob (in my experience) could not run it without paging and thus slowed to a crawl running it. Launching a program from Bob just resulted in interminable waits while the disk thrashed. The Magic Link is painfully slow and does not demonstrate the OS well at all.
Magic Cap was really no less insulting to users than Bob. It wasn't as cartoony but its tediousness wasted your time. The sluggishness of the hardware did not help. Even the early Newton MessagePads were snappier devices and their UI didn't make you tediously navigate through a virtual space.
> If you know anything about Microsoft, you know they don't innovate.
Longhorn failed because MS tried to be too innovative within the scope of a single OS release. In some ways Microsoft was more innovative under Ballmer than Nadella--see e.g. the radical Midori OS that Nadella killed, or Microsoft Research's highly productive Silicon Valley campus that Nadella shut down.
> Microsoft Research's highly productive Silicon Valley campus
What are some notable examples of innovations that came from MSR Silicon Valley campus? In my opinion nothing notable has ever come out of MSR, regardless of campus. Microsoft is needing to rely on third parties such as OpenAI because MSR seems incapable of contributing anything notable to AI.
The Seattle / UW related branch of MSR definitely had a few wins in the graphics department.
As https://www.microsoft.com/en-us/research/blog/meteoric-rise-... reminded me, they were on like 20% of the SIGGRAPH papers in 1996. And several have stood the test of time, like Hugues's Progressive Meshes paper (https://dl.acm.org/doi/10.1145/237170.237216). A little later, the spherical harmonics paper (https://dl.acm.org/doi/10.1145/566570.566612) was basically all MSR (I think Jan maybe interned there while at MPI? It's been a long time...).
That isn't to defend anything in AI, but it's not the case that they had no impact. The oral history thing there claims that the first grammar checking in Office 97 came from their NLP work.
Off the top of my head: differential privacy (which is now being deployed at scale), differential dataflow (Materialize), shared logs (Corfu/Tango, productized by VMWare). I'm sure there's many more that would be easy to google.
> In my opinion nothing notable has ever come out of MSR, regardless of campus.
One notable thing that came out of the Cambridge (England) campus was the implementation of generics in the CLR.
1 reply →
This idea may just have needed better technology. We should try these things again.
There's a bunch of stuff that didn't really work that well 30 years ago that we use today