Slashdot Mirror

Check out this article from Slashdot, which has this post, which has this article in it. Ph33r.

Check out this article from Performance Computing. Describes the IA-64 and its advantages.

They were discussing this new bus architechture as an alternative to PCI, Sbus, MCA, etc. I get the impression that they were discussing something -different- to firewire/usb or the memory bus.

The architechture they described for a SMP system looked something like this:

(See Figure 3 from the article. I tried to do it as ASCII art, but preview says "that doesn't work")

The difference between this and the current layout for a PCI system is that the memory/channel controller is replaced by the PCI controller, and the switch is replaced by the PCI bus.

Personally, I see USB as a controller hanging off the switch, converting between the (high speed) I/O bus serial protocol and the (lower speed) USB protocol. The same would be true with most existing protocols: IDE, SCSI, Firewire, etc, if for no other reason than to take advantage of existing storage media.

3.Anyone have any idea how to fix the problem of dynamic IPs?

• Read Bill LeFebvre's article on Internet Black Holes to learn how the Real-Time Black Hole system uses DNS creatively. You can also go write to the source if you prefer. Here's an excerpt:

  The simplest way to get started using the MAPS RBL to protect your mail relay against theft of service by spammers is to arrange for it to make a DNS query (of a stylized name) whenever you receive an incoming mail message from a host whose spam status you do not know.

• Here's the abstract for TCP Splicing for Application Layer Proxy Performance, by Pravin Bhagwat et al.:

  Application layer proxies already play an important role in today's networks, serving as firewalls and HTTP caches -- and their role is being expanded to include encryption, compression, and mobility support services. Current application layer proxies suffer major performance penalties as they spend most of their time moving data back and forth between connections, context switching and crossing protection boundaries for each chunk of data they handle. We present a technique called TCP Splice that provides kernel support for data relaying operations which runs at near router speeds. In our lab testing, we find SOCKS firewalls using TCP Splice can sustain a data throughput twice that of normal firewalls, with an average packet forwarding latency 30 times less.

• Here's the abstract for Improving HTTP Caching Proxy Performance with TCP Tap:

  Application layer proxies are an extremely popular method for adding new services to existing network applications. They provide backwards compatibility, centralized administration, and the convenience of the application layer programming environment. Since proxies act as traffic concentrators, serving multiple clients at the same time, during peak load periods they often become performance bottlenecks. In this paper we present an extension of the TCP Splice technique called TCP Tap that promises to dramatically improve the performance of a HTTP caching proxy, just as TCP Splice doubled the throughput of an application layer firewall proxy.

• Cohen, A., S. Rangarajan, and H. Slye. On the Performance of TCP Splicing for URL-aware Redirection. In: Proceedings of the USENIX Symposium on Internet Technologies and Systems, pp. 117-125, October 1999.

  Recently, the focus of the work on NEPPI applications was mostly on high performance URL-aware switching using TCP splicing. TCP splicing is a technique for bridging TCP connections at the IP level within the kernel, thus avoiding the overhead of application-level copying between sockets as performed by programs such as proxies. URL-aware switching with TCP splicing can be utilized in layer 7 switches to achieve high performance content-aware redirection of HTTP requests. We have developed of prototype of a layer 4/7 switch based on NEPPI.

• A Mobile Networking System based on Internet Protocol(IP) Pravin Bhagwat, Charles Perkins. Proceedings of USENIX Symposium on Mobile and Location Independent Computing, August, 1993, Cambridge, MA.

  Due to advances in wireless communication technology there is a growing demand for providing continuous network access to the users of portable computers, regardless of their location. Existing network protocols cannot meet this requirement since they were designed with the assumption of a static network topology where hosts do not change their location over time. Based on IP's Loose Source Route option, we have developed a scheme for providing transparent network access to mobile hosts. Our scheme is easy to implement, requires no changes to the existing set of hosts and routers, and achieves optimal routing in most cases. An outline of the proposed scheme is presented and a reference implementation is described.

• A Mobile Host Protocol Supporting Route Optimization and Authentication IEEE Journal on Selected Areas in Communications, special issue on "Mobile and Wireless Computing Networks," 13(5):839-849, June 1995. c IEEE. Andrew Myles Department of Electronics

  Host mobility is becoming an important issue due to the recent proliferation of notebook and palmtop computers, the development of wireless network interfaces, and the growth in global internetworking. This paper describes the design and implementation of a mobile host protocol, called the Internet Mobile Host Protocol (IMHP), that is compatible with the TCP/IP protocol suite, and allows a mobile host to move around the Internet without changing its identity. In particular, IMHP provides host mobility over both the local and wide area, while remaining transparent to the user and to other hosts communicating with the mobile host. IMHP features route optimization and integrated authentication of all management packets. Route optimization allows a node to cache the location of a mobile host and to send future packets directly to that mobile host. By authenticating all management packets, IMHP guards against possible attacks on packet routing to mobile hosts, including the interception or ...

• RFC 2230 has some words that might be relevant here:

  Dial-Up Host Example

  This example outlines a possible use of KX records with mobile hosts that dial into the network via PPP and are dynamically assigned an IP address and domain-name at dial-in time.

  Consider the situation where each mobile node is dynamically assigned both a domain name and an IP address at the time that node dials into the network. Let the policy require that each mobile node act as its own Key Exchanger. In this case, it is important that dial-in nodes use addresses from one or more well known IP subnets or address pools dedicated to dial-in access. If that is true, then no KX record or other action is needed to ensure that each node will act as its own Key Exchanger because lack of a KX record indicates that the node is its own Key Exchanger.

  Consider the situation where the mobile node's domain name remains constant but its IP address changes. Let the policy require that each mobile node act as its own Key Exchanger. In this case, there might be operational problems when another node attempts to perform a secure reverse DNS lookup on the IP address to determine the corresponding domain name. The authenticated DNS binding (in the form of a PTR record) between the mobile node's currently assigned IP address and its permanent domain name will need to be securely updated each time the node is assigned a new IP address. There are no mechanisms for accomplishing this that are both IETF-standard and widely deployed as of the time this note was written. Use of Dynamic

  DNS Update without authentication is a significant security risk and hence is not recommended for this situation.

As my sig file used to say "Windows hasn't increased computer literacy, it's lowered the standard".

See Sun press release. For Toy Story 2, they used 120 E4500 with 14 UltraSparc-II's each - total of 1680 CPUs, along with 4.5terabytes of storage. List price of around $30M I guess, though I presume they got some kind of discount ^-^. btw, one of the requirements was for the render-farm to be pretty compact. Performance Computing magazine have a review of the E4500 here. Pixar used Sun kit for their previous stuff too. If they do a Toy Story 3, by then the UltraSparc-IV should be out, which'll be about 5x faster in FP than current top-end UltraSparc-II's.

• Make sure that in optimizing the program interface for the two-minute beginner, you haven't pessimized it for the two-year daily-user.
• Keep it touch-typist friendly.
  • Let me keep my eyes on the screen at all times, not on the input peripherals.
  • Mimimize the context switches between mouse and keyboard. It slows me down. I can type much, much faster than I can mouse around.
  • Minimize all required mouse use, because it causes RSI. Let me keep my hands on the homerow as much as possible, not dancing around the funny keys that require me to look down to find, like HOME, END, PAGEUP, etc. Put those on real keys.
• No prior Windows knowledge expected, required, nor in fact, even beneficial.
• All programs, configurations, library functions, and interfaces must be completely documented.
• Never make me do anything tedious and repetitive, like holding some an arrow key or a mouse for a long time just to move a large distance.
  • I shouldn't have to read the library code to figure out how Gtk works, nor existing themes to figure out how to make a new one
  • nor should I have to click on happycons to get some dribbled out set of web pages for how to run or configure a program
  • The documentation should searchable, indexable, typesettable, and printable.
  • Follow POSIX 1003.2 requirements that all commands have a minimal manpage.
• Scriptability. Automatability. All the knobs need to be exposed either via raw text files or else normal CLI programs.
• Respect for the user's existing preferences where applicable.
  • X defaults -- If I have *visualBell: on, then that should suffice for all applications.
  • stty settings -- If I think ^H is what I want to erase a character, don't make me use DEL or ^?, or worse still, the BACKSPACE key (which sends a ^H anyway) yet not Control-H). And if I have my werase set to ^W, pay attention to that, too.
  • Preferred editor -- if I have an editor setting in my environment, don't make me learn a new one just for your program. Most toolkits' text widgets have insultingly idiotic editing abilities -- pop up my preferred editor instead, or at least, give me that option. Perhaps prefered newsreader, shell, mailer, etc should come into play as well, but the sorry excuse for an editor is the most annoying thing.
• A way to leverage existing knowledge of words. This may sound bizarre, but nothing is more frustrating to this Unix user than to have a program pop up a set of seventeen tiny graphical stickpin icons. Don't make me guess what your cutes idea of a neat bitmap for an exit or reload or search button is. Allow me the option of using real words, not happycons. And allow for keyboard shortcuts on all functionality.
• Don't make me suffer through a tedious manual search through scads of cascading menus each time I want to find something. There is no way to search, analyse, or print a cascading menu system. This is insane. A common mechanism provided by the low-level toolkit and managed by the desktop or window manager must be invented. Life is too short for hunt and peck. For example, the window manager could provide a way to search the menus of the current focussed program for a particular text string. That way you never have waste your life on an idiotic recursive but linear visual search.

Okay, so let's say the Judge rules that MS has used it's Monopoly power for illegal, anticompetive ends and then M$ is somehow no longer made nearly the power it is now, by breakup or something similar. How would this be good? Increased government regulation in this arena can have no positive outcomes. The government is like anti-Midas: anything it touches turns to shit. Do you honestly think Linux would have grown to the power it is today without Microsoft? Let's say windows only had a 35% desktop market share, and the rest of the market was split up between Be, MacOS, some Unix derivatives, and some other new crappy windows-like operating systems. I do not believe that the OSS community would have had the motivation to come together and produce something like GNU/Linux without something like M$ to push against. I firmly believe a lot of OSS hacking is done for it's "fuck M$" value. Look at Applix, pride of those who would have Linux replace windows 9X on my mom's desktop. Without MS Office in it's gilded place, would OSS really have done this on its own? No, OSS would have told my mom to learn vi. Microsoft's position effectively functioned as a highly competitive one for OSS. M$ provided the benchmark and the motivation for OSS to exist. This is nothing new: AT%T's Unix, and effectively, highend OS, monopoly of times long since passed spurned the writing of Andrew S. Tannenbaum's "Operating Systems: Design and Implementation," which an article posted on slashdot only last thursday cites as the beginning of the OSS movement in general. All of this came from being disgruntled with a corporation on power, which is a form of competition. This probably wouldn't have happened without this impetus of dissatisfaction. If I was Jon Katz and Wired was paying me to write this response, I'd feel compelled to call it something like, "New Competition in the Connected Era". If the government artifically reduces M$ ability to compete, by breaking it up or draining it's coffers with class-action rebates, how will this help the cause of increasing software quality? It won't make Linux run better. It won't make anyone go out and write better code. The quality of software will be furthered by old fashioned competition. People switched to Linux when it got better than Windows. How will M$ having less money or power make Linux get better? It will mean the Fed will need more money, departments, committees bureas, and red tape to "protect" the rest of the high-tech industry from the de-fanged M$. It will slow the entire process of innovation down. M$ should have been defeated in the market arena. It was inevitable. It's almost laughable how out of touch there strategy appears to be now. Some things should have been regulated by the Fed: namely, M$ violation of their contract with Sun regarding the Java standard. That is illegal. That will harm the consumer and slow the advance of technological innovation, but if the Fed was to force M$ to comply with the contract, the bricks of the M$ wall would continue to fall as they have been. They are already losing it: WinCE is a laughable in the PDA market compared to PalmOS, and will likely never really compete against things like Jini. Pervasive tiny connectivity is one of the Next Big Things, and /. head can tell you that. M$ is nowhere. They are still not realizing that "The Network Is the Computer" as Sun puts it. You know how many IT depts are gonna jump on 0 admin thin clients? Prolly lots. IT is really just a process of trying to minimize headaches, and the thinner the clients, the less headaches you have. Also, they have approximately 0 wireless strategy at this point. The next big thing is the total of these: de-power the desktop (making it impossible to run a million licensed copies of lard-ass Word all over your company), simplify it, connect your small devices like phones and organizers, and maintain pervasive, wireless, two way connectivity. After that, increase the bandwidth and decrease the size. THis is the clear path. Linux is moving towards it as a community, Sun, Handspring, and others are as corporations, /. is as a geek guild, M$ is in the dark! Who needs the Fed to take out their fangs when they are about to be trampled in the stampede anyway? jeb.

We don't need this. We need real Linux apps.

A "Winix app" is a program that happens to run under Unix but doesn't fee like it. All the social cues are wrong. It's like a hard-core Apple Mac user being given an MS-DOS program that has been ported to his system, where "ported" means no more than making the code run. It's still not going to work for him. And he'll hate it.

What are the the Unix cultural cues that are missing from Winix "apps", or Windows cues that are present in Winix "apps" but missing from Unix programs? It's a lot of little things. Here are just a few:

• Trying to be all things, not one thing, and thereby ending up being huge and bloated, not small and functional.
• Not being usable as a tool called from other programs, nor being able to call other tools using this programs.
• Requiring a bitmapped display with bus-speed updates for proper operation, making them useless over a modem connection or on a vt100-style terminal.
• Can't be run in a batch script or cron job, because it can't be automated.
• Not having keyboard support, or good keyboard support, for basic operations.
• Relying upon the mouse for too much when the keyboard would be more efficient.
• Not having regular expressions for searching. You're lucky if you get a case-sensitive string comparison.
• Not having any manpages even though these are basic commands, and all commands must have manpages for searching and indexing and printing.
• Being forced to learn yet another set of unsearchable, unmemorizable, unsortable hieroglyphics (iconic ideograms) rather than alphabetic commands.
• The ubiquitous "toolbar" at the top of every program.
• Lack of config files in simple text formats for easy parsing and generation by other tools.
• Inability to configure out the eye-cruft, or to avoid the seven levels of mouse clicks needed to get to a particular selection. There's no ease of long-term use, only short-term use.
• Complete ignorance of the user's preferences from stty settings for interrupting, line editing, etc.
• General disdain for shell globbing conventions, especially tilde expansion.

If you run Word Perfect under Linux, it's still not a Unix program. It's a Winix "app". It doesn't feel like Unix. xv is a Unix program, but ee is a Winix "app". KDE and CDE mostly comprise Winix "apps", not Unix programs. Netscape is a Winix "app", not a Unix program.

Linux users in particular seem to be happy with Winux "apps". Even more frighteningly, they seem happy to crank them out from scratch. I'm not completely certain I understand that phenomenon. Perhaps this is because they never really got into Unix programs in the first place. These things certainly aren't Unix friendly.

I don't see the mindless porting of Windows "apps" to Unix systems as being particularly useful to Unix people. The resulting "Winix apps" will never be proper programs and tools. They weren't designed that way. They may well appeal to the mindless masses who've been trained to accept a completely distinct set of cultural computing cues, and this strategy might get more non-technical users to employ the platform, but it does nothing to make happy those of us who are already comfortable with Unix. In fact, it often has quite the opposite effect.

This isn't Unix as Literature anymore. We're in the the post-literate age of populist pablum instead. But hey, that's both the price and prize of popularity. You're supposed to enjoy it. Well, unless you're one of those Unix types, that is.

http://www.perform ancecomputing.com/reviews/software/9904c.shtml

You're also toeing the line dangerously close to Godwin's law, with your silly rants about "Sovietism". Grow up, please.

Alright, I admit I may have had rant switched on without realizing it, and for that I apologize. Godwin's Law is a good point, but I was attempting to make a point with the Soviet remark. My point was simply that by casting decision-making and problem-solving choices into dialog boxes, dropdown lists, radio boxes, OK buttons and so forth, the OS manufacturers rob the user/administrator of the option to make creative solutions to problems. It was actually an analogy I borrowed from Thomas Scoville's article "Elements of Style: UNIX as Literature". He was discussing NT, but if you'll read the article, you'll clearly see how I thought his analogy applied directly to OS X.

By the by, your calling me to the mat on Godwin's Law was quite appropriate, but let's bag phrases like "grow up". No reason to get personal just because someone feels passionate about something.

like FreeBSD. However, your "several years ahead" of the old BSD 4.4 system argument is flawed, considering MacOS X uses FreeBSD (and NetBSD) code, plus its own enhancements drawn from almost TEN YEARS of experience with Mach (the NeXT boys DO work at Apple now, remember?)

You'll pardon me good sir, but where did you get this information? Certainly not from Apple, in their white papers, they claim OS X to be based on BSD 4.4 and Mach 2.5, but make no claim to possessing Net or Free BSD code in their system. The terms "based on" to me say that their arcitecture is based largely on the old 4.4 release. I stand by my original statement, but would be happy to retract if you can find me a reliable source to prove me wrong.

The NeXT boys do work at Apple now, but NeXT flopped, if I may remind you. A friend of mine had a son who worked at NeXT, too. From what I heard, it was the classical Job's style of management that helped keep the company from going too far. It'll be interesting to see how his style affects Apple now that he's back.

Judging things on technical grounds must be seperate from religious grounds. Don't like the proprietary GUI? fine. But
don't start telling me that "it's not easy to use", or it's technically inferior. You have already shattered your credibility
in claiming that FreeBSD is as easy to use as MacOS - it plainly isn't.

I never claimed it was as easy to use as the Mac, I simply said it was easy. The first time I installed FreeBSD it took me the time to download it from the Internet, basically. To get X running took another hour, perhaps. I would never claim it was as easy to use as the Mac, because that level of simplicity does reduce usability and render a product "technically inferior", to use your words, not mine. Again, I reference you to the Scoville essay above.

In the end, there's a level of irreducible complexity to computers that makes a GUI into a waste of time. I used to be a GUI nut, but then I realized what power and flexibility the command line offers. It's the same reason when a book gets turned into a movie the people who loved the book kavetch that the movie was horrible, it was glossy, and it lacked depth. (ie the Tom Clancy books, the John Grisham books, an so forth). Don't get me wrong, I like movies...but a good book is far more rewarding and enriching. Thus the GUIONLY mentality of Mac people annoy me, and encite me to rant. ;) The GUI, like other media such as Television and Movies, promote the stiffling of thought. They garner a response mentality, where one reacts to situations; rather than a striving mentality, where one actively thinks about a problem and tries to solve it.

As to the title of your rebuttle, I think it says more than I could say about certain matters.

It's right on the mark regarding Sun's popularity in industry, their popularity as the leader of Anyone But Microsoft, their ability to take advantage of open standards and open source, and the criticisms brought upon them by competitors who accuse them of being just as bad as Microsoft if they could get away with it.

The funny part was that the article (in the April issue of Performance Computing) says that NT can barely reach 99% up time (that's 10 work days of unscheduled down time), which is why it's so laughable a choice for a data center server. I'm not sure though if they're comparing apples with apples (e.g., clustering servers).

FWIW...