Longhorn

7 Replies to “Longhorn”

1. Sorry Scot, but this is pretty unfair.

   WinFS does 100x more of what BeFS did, BFS is a child’s toy in front of WinFS. Even Havoc from Red Hat is at awe at these features: http://log.ometer.com/

   As for copying Apple, I don’t see it either. Be had the exact same idea (hardware accelerated desktop) before Apple did. Jason Sams was pushing for it in early 2000. And no, I would never say that Apple copied Be. Neither MS copies Apple in it. It is the natural evolution of technology. Expect hardware accelerated interfaces on all consumer products, e.g. mobile phones and PDAs in the coming years. Same goes for piles. Some things are just evolutionary steps that only makes sense to use them, even if someone else already does.

   And Longhorn itself has innovations not found anywhere else. MS *does* innovate, and plays catch up in other areas too. Same goes for Apple. It’s how the game plays, nothing more.

2. Oh, and something else regarding hardware accelerated desktops. JBQ (my husband, for those who don’t know :) has already done it in May 1999 while at Be, Inc.:

   He wrote a fully hardware-accelerated dynamic-composited UI in May’99 (with drop shadows and other such niceties) for the the kitchen appliance (remember that one?). The way he did it was faster in many ways than Apple’s QE, but not as flexible. But the core idea was exactly the same, though.

   I also remember talking about it in early 2001 in our bed time.. You know, we are both geeks, so before we go to sleep we talk about such stuff.. :D
   Anyways, back then he was geeking out with his hobby OS, and he had some basic VESA support. He wanted to add this facility to his OS too.

   To make the long story short, such ideas are just plain evolution, even if they sound so innovative to people who haven’t seen it before. Engineers with some vision (not everyone has that ;), will agree that hardware acceleration is a no brainer. And so, Jason Sams didn’t stole JBQ’s idea, Apple didn’t stole Jason’s, and MS didn’t stole Apple’s. It is just evolution.

3. Hi Eugenia – I actually agree with you about the evolution of technologies, and have made the same argument in the past. One car manufacturer invents power steering, then all of them have it. But another invents tinted windows, another invents better bumpers… they all feed into each other. I was being more playful about it, not meaning to sound indignant or anything. Hell, the whole Unix foundation of OS X is non-Apple.

   I wasn’t thinking about hardware acceleration. I was referring to things referenced in the Register piece like the very concept of the database-like filesytem (which I still maintain is a Be innovation, perhaps their greatest contribution to the history of computing, when the dust settles). I’m looking forward to seeing how Longhorn’s stab at this is similar or different from Be’s.

   And I was referring to Longhorn technologies that sound a whole lot like Rendezvous to me:

   “…and a Who is in my neighborhood service, informing users about who is in the vicinity with a PDA, laptop or smartphone. ”

   Of course, Rendezvous is just zeroconf, which Apple did not invent, but Apple was first to market it and package it and make it user friendly, as a real OS feature.

   But again, I don’t think it’s all that significant who invented something first — everyone will copy everyone ultimately anyway. What matters more is how well something is implemented, e.g. how much easier and more reliable 802.11 is in Mac OS than it is under Windows, where differing vendor implementations create chaos.

   It is fun to keep track and to poke fun though :)

4. >I’m
   > looking forward to seeing how Longhorn’s stab at this is similar or
   > different from Be’s.

   WinFS does not rely on user-space metadata to do cool stuff. The users won’t see the metadata, except in very few cases (e.g. mp3 attributes), like it is today with NTFS 5. Metadata support is there of course, but they are only accessible via an API that helps apps working well together.
   While BFS was fantastic for its time, today is useless compared to Reiser4, WinFS or XFS. It has no security, no ACLs, no mulituser support, nothing. Check the links at Havoc’s blog for more features found on WinFS.

   >But again, I don’t think it’s all that significant who invented something
   > first — everyone will copy everyone ultimately anyway.

   Agreed.

   I would be suspicious though if I see Expose’ on another OS. Expose’ can be done without hardware acceleration (just slower), but still, no one did it so far except Apple. So I do consider Expose an Apple’s innovation, even if it is a ‘simple’ utility and others might have thought of it in the past but where too lazy to implement it.

5. I remember hearing Microsoft talk about database-like file systems some 3 or 4 years ago – been looking forward to it ever since (mainly because BeOS made me realise what a great idea that is).

   Does this mean I can ditch iTunes now ;-)

6. We’ll see. One thing I don’t like about Windows is that it relies too much on the GUI (for the same reason I didn’t use MacOS for the longest time), and from what I see from the WinFS-links, this situation is only getting worse. Myself, I like to use both GUI and CLI.

   On the other hand WinFS apparently intends to solve data-organizational problems I simply don’t seem to have (like Windows in general never really appeared useful for my purposes), so what do I know.

7. A postscriptum: According to an article posted on today’s OSNews, Longhorn finally comes with a shell worth its name, which also includes commandline access to the WinFS features.

   Kinda like using Perl or Python as your command shell :-)