So if you run a server with only port 80 open, you can't exactly do much with it. I guess content gets onto it via, um, HTTP PUT or WebDAV. Or in a production situation you'd have two NICs in it.
Either way, this is only testing the security of IIS's URI parser and of the request parser on the ASP engine. Granted, that makes for a safe webserver from the outside, but it doesn't speak for all the things happening on that assumed second NIC, like DCOM, NetBIOS, connections to databases and transaction monitors, Active Directory, any code running behid real-world ASPs more complex than a guestbook, DNS, WINS, SNMP, and so forth.
You can do the same thing on any OS, with the same near-invincibility on port 80. But you haven't really addressed the real-world vulnerabilities of a machine that would have other services running and conectivity to other hosts on other ports and interfaces.
This must be the bestest search engine ever, because the name says it is. You can't do better than "alltheweb". Everyone else might as well pack up and go home.
So with about 75% of an album's worth of Life House songs scattered around Who's Next, Psychoderelict and such, maybe the good news is that he didn't have to write more than a couple (if any) new songs to round out the project. And that's a good thing, seeing as he hasn't put out a record with even two paricularly good songs on it since Chinese Eyes more than 15 years ago.
This isn't exactly earth-shaking. For one thing, if you're going to run a $35,000-per-CPU app server against $100,000 in database licenses, and pay $3,000 per seat of their development tools, which in turn hook into $4,000 per seat of Java development tools, saving some money on initial hardware purchase and OS licenses isn't exactly most companies' top priority.
Second, the Netscape (neé Kiva) App Server and Sun (neé NetDynamics) App Server are being merged into a single product line, with the most likely outcome being a hybrid of the Netscape app server itself (fast and way scalable but has crappy tools) and the NetDynamics tools (widely loved and mature but sit on top of a merely-good app server).
A port of either to Linux in the short term would be a distraction and would see few sales. The next rev, when the single poduct emerges, will probably be another story.
With its 800x600 display and 32MB base RAM and no DVD-ROM, the iBook isn't suitable for corporate use. Duh.
It's their consumer laptop. In fact, it's arguably the first real strictly-for-consumer laptop ever. Yeah, most laptop vendors have a value product line, but they're just heavy models with small screens with nothing changed to target the young non-techie student or the home user who wants a laptop instead of a desktop to browse the web, do some writing, and send e-mail.
Hey, I'm a geek, so I don't really want one. I have big screen and disk requirements, ad I need a 2- or 3-button pointing device. But most people don't, and these sure are cute, maybe even cuter than an iMac. And they've got a handle and the wireless networking card, so they're easy to carry around without the fuss of a laptop bag. They're gonna sell zillions of 'em if they can ramp up production. And the New Apple seems to be able to do that.
An online store that only sells one kind of things is the norm these days, and it's patently silly.
Who wants to set up login accounts and billing profiles with different vendors for each type of product they buy? Who wants to use different interfaces for buying books, music, toys, shampoo and pizza?
No, all of this is going away, this whole silly "web browser" concept first and foremost. Along with the notion of "going to a web store".
I have a Palm VII. I have MySimon's comparison pricing applet for the Palm VII. I can stand in a bookstore and look up the best 25 prices on any book any time I want. And by this fall, I'm sure I'll also be able to order that book, or CD, or camera, or shoe, from any of those vendors.
This is all going to settle out into not just aggregation of pricing and catalog info as comparison-shopping services like MySimon already do. The next step will be aggregation of the buy process, so the buying will be done in a brandless environment through a higher-level aggregated shopping app. You won't know you bought that book from Amazon until the box arrives in the mail.
In the next few years, online shopping is going to assimilate into one large reverse auction experience, with companies like Amazon turning into little more than warehouse operations that participate in the bid process, unless they can offer value-adds like recommendations and promos for repeat buying. And the shopper won't even know it. From the shopper's point of view, they just know they're buying their book for a good price from one great big store.
Given this, expanding a lean pick-and-pack operation like Amazon's fulfillment center to do pick-and-pack of more goods in hopes of achieving economies of scale may be the only way to go.
It's no New Beetle or anything, but the new Rio looks pleasant enough. The clear ones look a lot like a Gameboy Pocket, IMHO not a bad industrial design to crib from as far as such things go.
Applix Anyware, the all-Java port of Applix's butt-ugly Unix office suite, has been around for some 3 years now. I remember trying the demo back in '96.
It was pretty nifty and quite usable, if predictably sluggish and butt-ugly. Now, what with the much faster JVMs and much faster CPUs we have nowadays, I'd imagine it's pretty snappy and makes an interesting solution for shops that want to deploy thin clients. But I'll bet it's also still butt-ugly.
The license terms look similar to those for MySQL. That is, it's free of charge when a person, company or org puts it on its own machines, regardless of who uses the machines. Payment and/or negotiation are required for redistribution.
As with MySQL, you seem to be welcome to build resale solutions around it without anyone getting paid, so long as your app leaves it to the customer to obtain and install rdist themselves separately.
The terms are weird and tortuous, but they do not seem to require payment for commercial or business-to-business use per se.
If a name as crappy as "Celeron" ended up being no obstacle to sales of countless millions of systems, then the even crappier but slightly less goofy "Athlon" should do just fine.
They should make TV commercials consisiting of nothing but one person, in close-up, spending 30 seconds puzzling over how to pronounce "Athlon".
Let me get this straight. BeOS is being pitched as a snappy multimedia OS for intensive image and audio processing tasks. It has fewer consumer apps than Linux and no mainstream web browser, but this is made up for by its wonderful suiatbility to multimedia production. Fine.
So here it is being sold on $500 PCs with low-end CPUs and low-end audio and video hardware to a market that generally goes with 15-inch monitors.
Not a bad idea. Yet another variation on things found most recently in the likes of MS ChromeEffects, but also seen in Apple's RDF browser, Perspecta's thingamawhatsises, the Brain, and, eek, Gopherspace 3D, that wonderful last triumph of, y'know, gopher, among many others. Nice to see some of these ideas moving off the VGA monitor.
Raster wants to do Stunt Programming, death-defying, convention-breaking, high-bandwidth stuff. Redhat has always been about cleaning up Linux for institutional, mainstream consumption. It wasn't a great fit. It's good that he's moved on, though he's been mighty childish and unprofessional about it.
JWZ is also a Stunt Programmer. On the Mozilla project, he was given marching orders to be a project leader, not a Stunt Programmer. The Mozilla project faltered. He left, frustrated. Since then, the Mozilla project has become much more goal-focused, its frequent milestone releases coming with clear goals. The difference is like night and day, reflected in everything from the crisp, punctual status reports and the daily inventories of showstoppers.
The successor to NS Communicator is still a long way from release, but the new discipline evident across the board on the Mozilla project shows how much dithering there was in the absence of strong project leadership.
This Raster guy and that Zawinski fella have done damn fine stuff, and have plenty of damn fine stuff ahead of them. But both were ill-suited to the jobs they were in, and should stop bellyaching.
Last time I checked, ICQ was just as slow and unreliable as it was before AOL bought it. The ICQ website is still a confusing, tangled mess, certainly no slicker than it was before AOL bought it. The ID numbers are as unweildy as ever, and the client software is still free of charge.
So what's poisoned about it? The fact that AOL bought it? Phooey.
E is slick. E gives good demo. E is user-hostile and though it's GNOME compliant, it's in no way GNOME-compatible.
RedHat did what they could. If they couldn't get Raster to show any interest in making sense out of E's UI or doing anything to make it or its configuration user-friendly, at least RH had the good sense to ship it with the only readable theme anyone bothered to make.
I thought the Hobbit-Klingon look and feel of most E themes and widget sets summed it up well. Yes, E has the nicest wipes and slides and zooms around, and is pretty lean about it. Not only is Raster uninterested in making it a usable environment for end-users, though: he's pretty clearly hostile to it. The idea of anyone making E usable for someone who isn't a Unix hacker or a CompSci major or a trekkie who likes to read API docs in his/her spare time seems to disgust him.
This same disease seems to affect the GNOME project to a lesser degree: who ever heard of an end-user environment that makes it easy to not only get rid of your main toolbar (the panel), but also makes it diffiult to make it come back by default? Why such an allergy to a trash can? Is two-stage deletion for wimps? Is the hatred for 15 years of UI design convention that strong? At least the GNOME project seems to finally be ineterested in some of these usability issues.
Raster, OTOH, seems to have nothing but open contempt for people who dare think of computers as a tool to run prewritten apps on. Not everyone working for a living has the spare time to become a programmer on the side so they can configure their windowing environment. Most people would regard a coherent GUI control panel for what is, after all, a GUI, as something sensible. Not as a sellout.
You can't stream to a Palm VII
on
PDA+MP3 Player
·
· Score: 3
The Palm VII works like a pager. It does not establish a persistent net connection; nor, at between $0.17 and $0.30 per kilobyte, would you want it to.
Okay, well, if you hook a modem to any Palm's serial port, I guess you could get the data transfer part right. But a Palm gets its combination of speed and long battery life by using a very slow CPU (a 16MHz 68020-derivative) running a very lean OS and lean apps.
Putting aside for the moment that the Palm doesn't have audio circuitry (it clicks a speaker, just like a soundcard-less PC or an Apple II), I'm more than a little skeptical that that 16MHz CPU is going to be able to decode MP3s, much less do so while doing anything else.
~2 hour battery life when audio plays
on
PDA+MP3 Player
·
· Score: 1
The Casio E-100 is a neat toy. Fast, has color, stereo sound, etc. Problem is, I saw an online review that said battery life under constant use drops from 6 hours to about 2 hours when you play audio.
Yeah, the PDA-with-audio approach is probably going to be compelling a couple of years down the road, but right now nobody's got a device that gets the battery issue right.
Since the Palms prove monochrome PDAs can run on extremely low power, a viable approach might be the one taken by the Palm series: the PDA functions are powered by a capacitor that gets charged at necessary intervals from the AAA batteries. The wireless subsystem of the Palm VII, as I understand it, essentially does the same thing, charging a rechargaeable battery roughly daily from the same AAA batteries. The only things powered directly by the AAAs are the backlight and maybe the speaker.
The problem with the E-100 in this respect is that the device has three very high drain components: the color screen, the high-speed CPU it uses to decode the audio, and the audio circuitry itself. A viable device might work as follows:
- Low-power PDA subsystem, probably with a monochrome or low-color (like Gameboy Color) display for now, and a slow CPU with a lean OS (Palm or EPOC, not the present WinCE).
- Dedicated circuitry for hardware decoding of audio files; this piece would run at a higher speed, but only when in use.
- Isolated amplifier circuitry, possibly with a separate power source (perhaps shared with the decoder circuitry).
Yes, it has MS Office 95 and 97 filters, and in my experience it does a markedly better job than WordPerfect Office at opening MS Office files. But it's not really compatible. It won't touch fast-saved documents, has trouble with longer ones, requires macros (if your business uses any) to be rewritten, and it's an even bigger memory hog than MS Office.
All this said, it's a decent office suite with a lot of great features, a nice interface, and damn fine cross-platform support. But it won't coexist comfortably with other office suites any better than any other office suite. A business or institution really can't mix office suites; for all the filters in the world, their file formats are still too far apart for everyday use.
A word processor that can "usually" open Word files is useless. When you're sent a Word file, the only acceptable word processor is one that can always open a Word file, even with clipart, drawings, equations, a glossary, embedded spreadsheets, and so forth.
You're not smarter than the people you work for. You're naive. There are a lot of things I dislike about MS Office, but even if StarOffice were suddenly so free that it was GPL'ed, it can cost a fortune to migrate a running business from one office suite to another.
If you have a department running Linux or some other such OS on the desktop and you need access to the company's standard office suite, WinFrame might make sense. And if you're on one of the major commercial Unixes, there's always SoftWindows.
What you really want to do is play Miner 2049er -- on top of an Apple II emulator -- on top of vMac -- on top of WINE -- on top of lxrun -- on top of Solaris -- on top of vmware
no networked administration?
on
NOS Crossroads
·
· Score: 1
Good grief. I didn't know Linux was so hard to manage over a network, and that I have to cobble together perl scripts that write logfiles to a shared volume to monitor my pile of machines.
I guess I should free up some space and get rid of all those SNMP agents I have running and scrap the NSS and PAM stuff that unifies configuration and lets the system participate transparently in things like NT domains.
It's going to kill me to decomission those old Pentium Linux servers I've got running and replace them with NT boxes. They seemed to be running so nicely these six months since I last booted them.
I guess Mr. Gassse's accountant told him he's got to stop pouring all his own money into it if he wants to be able to afford food and shelter after he turns 60.
Slashdot Mirror
So if you run a server with only port 80 open, you can't exactly do much with it. I guess content gets onto it via, um, HTTP PUT or WebDAV. Or in a production situation you'd have two NICs in it.
Either way, this is only testing the security of IIS's URI parser and of the request parser on the ASP engine. Granted, that makes for a safe webserver from the outside, but it doesn't speak for all the things happening on that assumed second NIC, like DCOM, NetBIOS, connections to databases and transaction monitors, Active Directory, any code running behid real-world ASPs more complex than a guestbook, DNS, WINS, SNMP, and so forth.
You can do the same thing on any OS, with the same near-invincibility on port 80. But you haven't really addressed the real-world vulnerabilities of a machine that would have other services running and conectivity to other hosts on other ports and interfaces.
This must be the bestest search engine ever, because the name says it is. You can't do better than "alltheweb". Everyone else might as well pack up and go home.
So with about 75% of an album's worth of Life House songs scattered around Who's Next, Psychoderelict and such, maybe the good news is that he didn't have to write more than a couple (if any) new songs to round out the project. And that's a good thing, seeing as he hasn't put out a record with even two paricularly good songs on it since Chinese Eyes more than 15 years ago.
Second, the Netscape (neé Kiva) App Server and Sun (neé NetDynamics) App Server are being merged into a single product line, with the most likely outcome being a hybrid of the Netscape app server itself (fast and way scalable but has crappy tools) and the NetDynamics tools (widely loved and mature but sit on top of a merely-good app server).
A port of either to Linux in the short term would be a distraction and would see few sales. The next rev, when the single poduct emerges, will probably be another story.
With its 800x600 display and 32MB base RAM and no DVD-ROM, the iBook isn't suitable for corporate use. Duh.
It's their consumer laptop. In fact, it's arguably the first real strictly-for-consumer laptop ever. Yeah, most laptop vendors have a value product line, but they're just heavy models with small screens with nothing changed to target the young non-techie student or the home user who wants a laptop instead of a desktop to browse the web, do some writing, and send e-mail.
Hey, I'm a geek, so I don't really want one. I have big screen and disk requirements, ad I need a 2- or 3-button pointing device. But most people don't, and these sure are cute, maybe even cuter than an iMac. And they've got a handle and the wireless networking card, so they're easy to carry around without the fuss of a laptop bag. They're gonna sell zillions of 'em if they can ramp up production. And the New Apple seems to be able to do that.
There is an FSF-sanctioned web browser. The web browsing mode in Emacs.
That's because northern Michigan is culturally Candadian.
An online store that only sells one kind of things is the norm these days, and it's patently silly.
Who wants to set up login accounts and billing profiles with different vendors for each type of product they buy? Who wants to use different interfaces for buying books, music, toys, shampoo and pizza?
No, all of this is going away, this whole silly "web browser" concept first and foremost. Along with the notion of "going to a web store".
I have a Palm VII. I have MySimon's comparison pricing applet for the Palm VII. I can stand in a bookstore and look up the best 25 prices on any book any time I want. And by this fall, I'm sure I'll also be able to order that book, or CD, or camera, or shoe, from any of those vendors.
This is all going to settle out into not just aggregation of pricing and catalog info as comparison-shopping services like MySimon already do. The next step will be aggregation of the buy process, so the buying will be done in a brandless environment through a higher-level aggregated shopping app. You won't know you bought that book from Amazon until the box arrives in the mail.
In the next few years, online shopping is going to assimilate into one large reverse auction experience, with companies like Amazon turning into little more than warehouse operations that participate in the bid process, unless they can offer value-adds like recommendations and promos for repeat buying. And the shopper won't even know it. From the shopper's point of view, they just know they're buying their book for a good price from one great big store.
Given this, expanding a lean pick-and-pack operation like Amazon's fulfillment center to do pick-and-pack of more goods in hopes of achieving economies of scale may be the only way to go.
Who's the dull, "unwired" guy now?
Anyone got an old sampling keyboard with a floppy drive?
.aiff or .au player?
Or a Newton MP100 with, say, an
It's no New Beetle or anything, but the new Rio looks pleasant enough. The clear ones look a lot like a Gameboy Pocket, IMHO not a bad industrial design to crib from as far as such things go.
Applix Anyware, the all-Java port of Applix's butt-ugly Unix office suite, has been around for some 3 years now. I remember trying the demo back in '96.
It was pretty nifty and quite usable, if predictably sluggish and butt-ugly. Now, what with the much faster JVMs and much faster CPUs we have nowadays, I'd imagine it's pretty snappy and makes an interesting solution for shops that want to deploy thin clients. But I'll bet it's also still butt-ugly.
The license terms look similar to those for MySQL. That is, it's free of charge when a person, company or org puts it on its own machines, regardless of who uses the machines. Payment and/or negotiation are required for redistribution.
As with MySQL, you seem to be welcome to build resale solutions around it without anyone getting paid, so long as your app leaves it to the customer to obtain and install rdist themselves separately.
The terms are weird and tortuous, but they do not seem to require payment for commercial or business-to-business use per se.
If a name as crappy as "Celeron" ended up being no obstacle to sales of countless millions of systems, then the even crappier but slightly less goofy "Athlon" should do just fine.
They should make TV commercials consisiting of nothing but one person, in close-up, spending 30 seconds puzzling over how to pronounce "Athlon".
Let me get this straight. BeOS is being pitched as a snappy multimedia OS for intensive image and audio processing tasks. It has fewer consumer apps than Linux and no mainstream web browser, but this is made up for by its wonderful suiatbility to multimedia production. Fine.
So here it is being sold on $500 PCs with low-end CPUs and low-end audio and video hardware to a market that generally goes with 15-inch monitors.
Who is a low-end Be system for?
Not a bad idea. Yet another variation on things found most recently in the likes of MS ChromeEffects, but also seen in Apple's RDF browser, Perspecta's thingamawhatsises, the Brain, and, eek, Gopherspace 3D, that wonderful last triumph of, y'know, gopher, among many others. Nice to see some of these ideas moving off the VGA monitor.
Raster wants to do Stunt Programming, death-defying, convention-breaking, high-bandwidth stuff. Redhat has always been about cleaning up Linux for institutional, mainstream consumption. It wasn't a great fit. It's good that he's moved on, though he's been mighty childish and unprofessional about it.
JWZ is also a Stunt Programmer. On the Mozilla project, he was given marching orders to be a project leader, not a Stunt Programmer. The Mozilla project faltered. He left, frustrated. Since then, the Mozilla project has become much more goal-focused, its frequent milestone releases coming with clear goals. The difference is like night and day, reflected in everything from the crisp, punctual status reports and the daily inventories of showstoppers.
The successor to NS Communicator is still a long way from release, but the new discipline evident across the board on the Mozilla project shows how much dithering there was in the absence of strong project leadership.
This Raster guy and that Zawinski fella have done damn fine stuff, and have plenty of damn fine stuff ahead of them. But both were ill-suited to the jobs they were in, and should stop bellyaching.
Last time I checked, ICQ was just as slow and unreliable as it was before AOL bought it. The ICQ website is still a confusing, tangled mess, certainly no slicker than it was before AOL bought it. The ID numbers are as unweildy as ever, and the client software is still free of charge.
So what's poisoned about it? The fact that AOL bought it? Phooey.
E is slick. E gives good demo. E is user-hostile and though it's GNOME compliant, it's in no way GNOME-compatible.
RedHat did what they could. If they couldn't get Raster to show any interest in making sense out of E's UI or doing anything to make it or its configuration user-friendly, at least RH had the good sense to ship it with the only readable theme anyone bothered to make.
I thought the Hobbit-Klingon look and feel of most E themes and widget sets summed it up well. Yes, E has the nicest wipes and slides and zooms around, and is pretty lean about it. Not only is Raster uninterested in making it a usable environment for end-users, though: he's pretty clearly hostile to it. The idea of anyone making E usable for someone who isn't a Unix hacker or a CompSci major or a trekkie who likes to read API docs in his/her spare time seems to disgust him.
This same disease seems to affect the GNOME project to a lesser degree: who ever heard of an end-user environment that makes it easy to not only get rid of your main toolbar (the panel), but also makes it diffiult to make it come back by default? Why such an allergy to a trash can? Is two-stage deletion for wimps? Is the hatred for 15 years of UI design convention that strong? At least the GNOME project seems to finally be ineterested in some of these usability issues.
Raster, OTOH, seems to have nothing but open contempt for people who dare think of computers as a tool to run prewritten apps on. Not everyone working for a living has the spare time to become a programmer on the side so they can configure their windowing environment. Most people would regard a coherent GUI control panel for what is, after all, a GUI, as something sensible. Not as a sellout.
The Palm VII works like a pager. It does not establish a persistent net connection; nor, at between $0.17 and $0.30 per kilobyte, would you want it to.
Okay, well, if you hook a modem to any Palm's serial port, I guess you could get the data transfer part right. But a Palm gets its combination of speed and long battery life by using a very slow CPU (a 16MHz 68020-derivative) running a very lean OS and lean apps.
Putting aside for the moment that the Palm doesn't have audio circuitry (it clicks a speaker, just like a soundcard-less PC or an Apple II), I'm more than a little skeptical that that 16MHz CPU is going to be able to decode MP3s, much less do so while doing anything else.
The Casio E-100 is a neat toy. Fast, has color, stereo sound, etc. Problem is, I saw an online review that said battery life under constant use drops from 6 hours to about 2 hours when you play audio.
Yeah, the PDA-with-audio approach is probably going to be compelling a couple of years down the road, but right now nobody's got a device that gets the battery issue right.
Since the Palms prove monochrome PDAs can run on extremely low power, a viable approach might be the one taken by the Palm series: the PDA functions are powered by a capacitor that gets charged at necessary intervals from the AAA batteries. The wireless subsystem of the Palm VII, as I understand it, essentially does the same thing, charging a rechargaeable battery roughly daily from the same AAA batteries. The only things powered directly by the AAAs are the backlight and maybe the speaker.
The problem with the E-100 in this respect is that the device has three very high drain components: the color screen, the high-speed CPU it uses to decode the audio, and the audio circuitry itself. A viable device might work as follows:
- Low-power PDA subsystem, probably with a monochrome or low-color (like Gameboy Color) display for now, and a slow CPU with a lean OS (Palm or EPOC, not the present WinCE).
- Dedicated circuitry for hardware decoding of audio files; this piece would run at a higher speed, but only when in use.
- Isolated amplifier circuitry, possibly with a separate power source (perhaps shared with the decoder circuitry).
Yes, it has MS Office 95 and 97 filters, and in my experience it does a markedly better job than WordPerfect Office at opening MS Office files. But it's not really compatible. It won't touch fast-saved documents, has trouble with longer ones, requires macros (if your business uses any) to be rewritten, and it's an even bigger memory hog than MS Office.
All this said, it's a decent office suite with a lot of great features, a nice interface, and damn fine cross-platform support. But it won't coexist comfortably with other office suites any better than any other office suite. A business or institution really can't mix office suites; for all the filters in the world, their file formats are still too far apart for everyday use.
A word processor that can "usually" open Word files is useless. When you're sent a Word file, the only acceptable word processor is one that can always open a Word file, even with clipart, drawings, equations, a glossary, embedded spreadsheets, and so forth.
You're not smarter than the people you work for. You're naive. There are a lot of things I dislike about MS Office, but even if StarOffice were suddenly so free that it was GPL'ed, it can cost a fortune to migrate a running business from one office suite to another.
If you have a department running Linux or some other such OS on the desktop and you need access to the company's standard office suite, WinFrame might make sense. And if you're on one of the major commercial Unixes, there's always SoftWindows.
Can we now look forward to a review of Around the World in 80 Days? Or I, Robot?
What you really want to do is play Miner 2049er
-- on top of an Apple II emulator
-- on top of vMac
-- on top of WINE
-- on top of lxrun
-- on top of Solaris
-- on top of vmware
Good grief. I didn't know Linux was so hard to manage over a network, and that I have to cobble together perl scripts that write logfiles to a shared volume to monitor my pile of machines.
I guess I should free up some space and get rid of all those SNMP agents I have running and scrap the NSS and PAM stuff that unifies configuration and lets the system participate transparently in things like NT domains.
It's going to kill me to decomission those old Pentium Linux servers I've got running and replace them with NT boxes. They seemed to be running so nicely these six months since I last booted them.
I guess Mr. Gassse's accountant told him he's got to stop pouring all his own money into it if he wants to be able to afford food and shelter after he turns 60.