The price of the Sun Ray is normally $500/each, and 20% is a pretty normal educational discount, so $400 isn't even slightly surprising. Of course, they still need a Sun server (at least until someone reverse engineers them and makes a Linux or FreeBSD solution) and they need monitors (but they probably have those already from their PCs).
You're right in that clustering is not anything new -- one of the best implementations is in Digital's OpenVMS, which is pretty old-school if you're counting in internet years.
But clustering is very different from the examples you give. It's not running different services on different machines. It is taking a bunch of machines and making them act as one.
Beowulf-style clusters are one way of doing this, but there's a limit to how many nodes you can connect that way and still get performance increases. It scales up, but probably not to thousands of nodes. Now, the LCC people obviously haven't built anything to prove that they can do better, but it sounds like they may have a theoretical improvement.
And, it's only hinted in the article ("satisifies both commercial data processing and HPC requirements"), but it's possible also that this technology is not only fast, but unlike Beowulf also provides improved robustness.
This is all vapor now of course. But we'll see. The people working on this have some important projects to their credit.
The WinCE machines don't get it. The Windows UI doesn't scale down gracefully, and trying to force that paradigm on to a palm computer doesn't work gracefully. PalmOS, for all its flaws, is light and responsive. And, the minimalist UI philosophy is far better suited to the size of the device.
Since it will be GPL'd, we're very likely to see it in the kernel. But it won't be soon -- there's no code there yet, and we're almost at feature-freeze for the 2.3/2.4 kernel. And I doubt it will be the default: ext2 is pretty good, is relatively lightweight, and most importantly, is tried and true. And, for the future, the ext3 project has a lot of equally exciting features -- and since it comes from "inside", from the same people who've been working on the current filesystem, it may have better integration, not to mention better mindshare.
I don't know about the second Linux World Conf/Expo, but the first one really failed in the Birds-of-a-Feather sessions. They were all on the same night, in the same room. If done right, this can (and should) be a central part of a conference.
I always find it amazing when people list AOL's purchase of Netscape as anything other than a clear and crushing Microsoft victory. I mean, this was the company that was supposed to change the landscape of the computing world.
I'd also like to see something like Photoshops "Variations" tool, which lets you preview and choose betweeen various image-adjustment options. This might be doable in script-fu....
My coding skills aren't up to it, or else I'd help, but what I'd really like to see is an option to make your cursor = the size of the brush you're using, as photoshop does. That way, you can see what you're doing!
Companies need to develop a policy on this kind of thing. Although the current law may allow corporations wide latitude, you're opening yourself to all sorts of trouble otherwise. (Moral and morale trouble, if not legal.)
Since that doesn't seem like it's the case where you are, SAGE's Code of Ethics for sysadmins might be personally helpful, at least.
Why share the serial ports? Wouldn't it be better to make the Linux box in charge of the ISDN connections (bringing them up and dropping them as bandwidth is needed) and then connect the Macs to it via a TCP/IP lan? Or am I missing something?
It's a distortion of (I think) Plato, something like "Man is the world writ small / The world is man writ large". I think the "world writ large" phrase is fairly common on its own, though.
--
Re:I think you're confusing shared with static
on
Mozilla M9 Released
·
· Score: 2
$ ls -l apprunner -rwxrwxr-x 1 mattdm mattdm 58133 Aug 24 20:31 apprunner $ strip apprunner $ ls -l apprunner -rwxrwxr-x 1 mattdm mattdm 25384 Aug 27 08:30 apprunner
$ ls -l viewer -rwxrwxr-x 1 mattdm mattdm 190161 Aug 24 20:31 viewer $ strip viewer $ ls -l viewer -rwxrwxr-x 1 mattdm mattdm 122068 Aug 27 08:30 viewer
Slashdot Mirror
--
But clustering is very different from the examples you give. It's not running different services on different machines. It is taking a bunch of machines and making them act as one.
Beowulf-style clusters are one way of doing this, but there's a limit to how many nodes you can connect that way and still get performance increases. It scales up, but probably not to thousands of nodes. Now, the LCC people obviously haven't built anything to prove that they can do better, but it sounds like they may have a theoretical improvement.
And, it's only hinted in the article ("satisifies both commercial data processing and HPC requirements"), but it's possible also that this technology is not only fast, but unlike Beowulf also provides improved robustness.
This is all vapor now of course. But we'll see. The people working on this have some important projects to their credit.
--
--
--
--
Since it will be GPL'd, we're very likely to see it in the kernel. But it won't be soon -- there's no code there yet, and we're almost at feature-freeze for the 2.3/2.4 kernel. And I doubt it will be the default: ext2 is pretty good, is relatively lightweight, and most importantly, is tried and true. And, for the future, the ext3 project has a lot of equally exciting features -- and since it comes from "inside", from the same people who've been working on the current filesystem, it may have better integration, not to mention better mindshare.
--
--
--
--
(looks like the preview window is too small to be useful, but that's fixable...)
--
The researchers found that 89 percent of award-winning ads match as few as six formulas, which they called "creativity templates"...
Could it be that the problem isn't supergood computers, it's that human designed-advertising is uncreative drivel? Kinda sounds like it.
--
This might be doable in script-fu....
--
--
--
--
--
--
--
--
--
Since that doesn't seem like it's the case where you are, SAGE's Code of Ethics for sysadmins might be personally helpful, at least.
--
--
--
--
$ ls -l apprunner
-rwxrwxr-x 1 mattdm mattdm 58133 Aug 24 20:31 apprunner
$ strip apprunner
$ ls -l apprunner
-rwxrwxr-x 1 mattdm mattdm 25384 Aug 27 08:30 apprunner
$ ls -l viewer
-rwxrwxr-x 1 mattdm mattdm 190161 Aug 24 20:31 viewer
$ strip viewer
$ ls -l viewer
-rwxrwxr-x 1 mattdm mattdm 122068 Aug 27 08:30 viewer
--