Slashdot Mirror
Linus speaks at Comdex
pq writes "CNET, via The NYTimes has
this coverage of Linus and his remarks at Comdex. iMacs, Transmeta, Open Source, the 2.4 kernel, Mindcraft, Mozilla, Linux Everywhere(tm) - this has it all.
" Also following the keynote, Transmeta's webpage announced
Crusoe (we mentioned that
last night).
In case anyone wants to view his keynote speech, it can be found at PC Week's website under the Monday tab. Good speech IMHO.
Upon what do you base your hypothesis that this involves reconfigurable circuitry? Not on
given that one could do that with a CPU with a fixed instruction set and a pile of software to do binary-to-binary translation of instructions from various instruction sets to the native instruction set...
...along the lines of what the Transmeta patents mentioned by various Slashdot articles have described.
I suspect that, if it's not already possible to have the kernel find a driver module given only, say a PCI vendor/device ID for the device (I seem to remember, possibly in a Kernel Traffic summary, discussion of doing something along those lines for USB), it could be done, so I doubt this is inherent to Linux.
Is that kernel-mode code? Or is it something that could be implemented in userland atop the Linux kernel and API libraries (or atop other UNIX-flavored OSes kernel and API libraries)? I have the impression that the print subsystem largely lives in userland in Windows NT (I'm not sure there's a strict distinction between userland and kernel code in Windows OT).
I.e.,
Linux (and other OSes to which source is made widely available) may allow you to build kernels more finely tailored to particular hardware than other OSes (although I don't know what you get with, say, Embedded NT), but that doesn't mean that they're necessarily better suited for embedded use than general-purpose desktop use - I've seen nothing to indicate that it's intrinsically impossible to build an OS that can both be built for a specific hardware configuration and be built as a more "generic" OS with the ability to load arbitrary drivers, or, if it is possible, that the resulting OS would necessarily suffer from being so built.
Transmeta are building a processor. For it to have any chance in the marketplace, obviously it needs an OS and Linux is a sensible choice. It seems to port well to platforms other than x86 (LinuxPPC, Alpha and others). It's stable, which is vital for embedded stuff, and it's easy to downscale.
However, processor manufacturers also need to ship compilers that are optimized for their chips. Intel do this very successfully. Their code profiler (VTune) also helps to generate code that takes advantage of every bit of the hardware. Because of these tools, Intel have a significant edge in benchmarks against other manufacturers who don't produce compilers.
Anyone know if Transmeta have been hiring compiler writers? Can we expect a set of Crusoe optimizations in egcs?
No! But it will run code for all OS flavors you can muster.
-- What you do today will cost you a day of your life.
This is a typical response from an Anonymous Coward:
CNET is a VERY "Microsoft oriented" site, and the author is clearly contrasting Linux with "normal" (read: Microsoft) OSes.
It is not that we constantly say "Uh, MS does that too!" it is that the FUD mongers continuously, vaguely accuse anything non-MS as being "inferior" when it is plain that it is in fact, superior in nearly every substantive way.
I too tire of "M$ Sux" posts, but that is not what mine was.
-Peter
>However, Linus has long since handed off much of >his responsiblity vis-a-vis the kernel to Alan >Cox and others.
However, doesn't Linus have to sign off on some of the major peices that get into the kernel, and also sign off on when a new Kernel level (e.g. 2.2 to 2.4) is ready to be announced?
This post brought to you by your friendly neighborhood MBA.
I usually don't believe in this whole "Linux is God" aspect that some members of the Linux community follow, but he does have some pretty good things to say usually (apparently he's a pretty bright kid =) ). So...
Ok boyos. here is what, in my opinion, is the best part of the article:
Torvalds steadfastly maintains the challenge and excitement
of producing a operating system--not competition with
Microsoft--is what motivates Linux developers. But he
couldn't resist a few jabs at the software giant.
...
"People see Linux as anti-Microsoft. In the press it looks that
way," Torvalds said. "But to me and all the developers I
know, it's not me vs. Microsoft." It's a matter of a fun
programming project, he said.
There it is in black and white. Some of the most hardcore Linux developers do it becuase they love it. It's a challenge. A problem to be solved. It's fun. The thrill of the chase if you will. And a little bit of pride. It's not the great apocalyptic david and goliath battle some make it out to be. It's about making the best system that they can.
Is competition necessary? Sure. If nothing else it shows places where improvement is needed. Case in point:
"While I was upset about Mindcraft for awhile, I took it as a
more positive thing after I got past the personal injury to my
pride," Torvalds said. "We just delved into it and fixed it.
We took this benchmark as a way of saying, 'Yes, Linux is
not the best at everything.' We fixed the area, and as a
result, Linux is doing extremely well on those kind of
benchmarks."
"You want to kiss the sky? Better learn how to kneel." - U2
"It was like trying to herd cats..." - Robert A. Heinlein
Sig:
Barbeque is a noun. Not a verb.
I usually don't believe in this whole "Linux is God" aspect that some members of the Linux community follow, but he does have some pretty good things to say
I get tired of this characterization of the Linux community, or even some of its members. It really doesn't have much basis in reality, and can generally be attributed to those who are opposed to the success of Linux for whatever reason. Indeed, it is a kind of personal FUD, aimed at the users and advocates of Linux more than anything else. While some may jokingly say something to the effect of "Linus is God", even the most ardent Linux advocate has his or her toung firmly in cheek.
There is a great deal of respect, gratitude, and good will toward Linus for what he has done and the contributions he has made and yes, even some hero worship. Compared to the adulations others (such as what Michael Jordan or Bill Gates get from their followings, for example), the adulations Linus Torvalds has gotten are generally quite moderate.
Is competition necessary? Sure. If nothing else it shows places where improvement is needed.
Absolutely. We know this, the DOJ knows this, and the Judge knows this. Others appear not to grasp this concept quite as well.
The Future of Human Evolution: Autonomy
I'd still like to see Linux as a desktop alternative. Simply so that I have a choice of what I want to run on my machine. That said, I don't really mind if Linux will never make it as the desktop OS of choice for non-techs. However, if Linux will only fill a small niche in this market, the problem will be that developers will not focus their efforts on running their products on Linux, and as a result, I will get poorer applications support just because I chose not to run Windows (or whatever else the desktop OS of choice may be). I mean, I will always be able to run Linux on my machine if I wanted to; but the problem is, I would be somewhat excluded from other desktop PC users because there will be limited support for applications on my Linux box and limited portability of files between my box and other machines, etc..
I guess, in my mind, the "ideal" picture would be to have Windows, Linux, and whatever else out there, to each have roughly the same share of the market. Then we won't have the MS monopoly problem, there will be healthy competition that ensures quality amongst the systems, and I will have the freedom of choice which OS I want to run, and, because each OS will be roughly equally important, applications will be equally supported for all platforms, and I won't feel like I'm using a "niche" OS as opposed to a "mainstream" OS.
mikre he sophia he tou Mikrosophou.
IDG
matt
When Linus graduated, he was asked why he didn't take a job at RedHat devoting his full time to writing the kernel. His response was that he didn't want to be swayed by his employer's interests into decisions that would harm the kernel's functionality. He joined Transmeta to keep his employment completely seperate from his Linux hobby.
Well what he's doing today is very close to what he wanted to avoid. Transmeta developes an embedded processor. Every time he speeks he chants about getting Linux into embedded systems and today he actually said Linux is more practical for embedded systems than a desktop. You have to wonder how much pressure he experiences working for an embedded systems company to divert his efforts from other kernel improvements to fitting the kernel into Transmeta's agenda.
No no no, you have it wrong,
There will not be a crusoe source arch branch, just a configuration branch in, "make config," that will read:
[ ] Use Crusoe Processor
|
|-[ ] Run Crusoe in x86 mode
|-[ ] Run Crusoe in ppc mode
|-[ ] Run Crusoe in sparc mode
:)
-AP
*COUGH!* Either Linus' times are "absolute", in which case there's not a kernel that's been "on schedule" in the history of Linux, OR Linus isn't promising anything but is giving a tentative guesstimate, in which case no Linux timetable exists.
Second, I'll believe Transmeta's release date when I see a "Crusoe" arch in the Linux kernel.
Uhhhhhhh.... Hooooooooooooooooold on a moment! This looks suspicious! A delay in 2.4, until about the same time Crusoe will be unleashed? And no hard freeze yet? Hmmmm. Are you thinking what I'm thinking?
"I think so, Brain, but I've never seen a pink giraffe in a tutu."
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
This week's editorial on Advogato has some more discussion of Linus's talk, from the specific viewpoint of free software developers.
Everyone is welcome to read, but Advogato limits posting to members of the free software development community. It's my hope that this will make Advogato a more useful resource for developers.
LILO boot: linux init=/usr/bin/emacs
|-[]Run Crusoe in Robinson mode
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
Can you believe the FUD mongering?
"On the down side, Linux's upgrade schedule--which never is etched in stone--has slipped a little in the last few months."
What is the author contrasting this with?
Oh, I know, it's that "etched in stone" NT 5 . . . I mean Windows 2000 (or is it 2001?) release date.
-Peter
An interesting thing I got out of his keynote, was the fact that the Linux Kernel will scale both UP and DOWN. . . Will other OS's do this? I guess I had never seen it from that point of view. Wouldn't you want to develop on your desktop, deploy to either clusters or handhelds? The fact that Linux will scale like this, is reason enough for me to develop for the OS. Personal note, when Linus mentioned Mozilla, I was the one who started the "smattering of applause" in the audience...
1.3L, 3 moving parts, 280 HP, no Turbos, wanna Race? RotaryNe
Linus's most admirable attribute, which is also IMHO the reason he's so successful with Linux, is that when beat by his competitors, it doesn't cause him to hate them, but rather it causes him to work harder and make a better product.
IMNSHO many Linux zealots has a lot to learn from Linus... Linux is not about "we hate MS", but it's about "how we can make something better than MS, or anyone else for that matter, can". If Linux was merely about hating MS and trying to "take over" MS's market, it'd be no different from MS abusing its monopoly to squish out competitors. It was quite sad for me to read the posts about the Mindcraft benchmark results -- most of those posts show that the attitude of some people here seems to be the "I hate MS" attitude. If we hate MS more just because Linux lost to NT in the benchmarks, and we accuse Mindcraft of having staged the whole thing, etc., are we not the same as MS spreading FUD against its competitors? Rather, we should take the negative result as an indication that Linux has room for improvement, and we should use our energy to improve it, not waste our time by shouting "not fair".
mikre he sophia he tou Mikrosophou.
On upside.com...
A bit about his "good nature", etc... click here.
--globalnap.net, product of pure caffeine--
Transmeta's Crusoe is for mobile applications right? When we think of mobile we usually think of portable right? What if Transmeta doesn't mean portable as in transportable (i.e. a laptop or palmtop) but portable as in "gcc has been ported to many different platforms"? "A whole new world of mobility" would then mean the processor itself is mobile as in not static. Mobile as in portable as in porting apps to itself on the fly. Would that then mean a Crusoe desktop is possible?
Just food for thought.
There comes a time in every man's life when he must say, "No mother! I do not want any more Jell-O!"