Realisticly, you'd be hard pressed to run even windows 98 with less than 32mb of ram on anything slower than a 200mhz pentium. Office XP is not a speed demon either, on a 500mhz celeron, it takes 20ish seconds or so to load up, and has a noticably laggy UI. I haven't ran openoffice lately, but I'm sure it's safe to say that it's no speed demon either.
I'm not convinced that either product is usable with the cited least required systems.
If you're referring to this virus, it polymorphs, meaning it changes it's basic makeup to evade being detected. AFAIK, it doesn't actually touch the virus scanner, it attempts to avoid it.
A serious no-no: Do not `introduce` the student to other students. There is nothing more embaressing than trying to be shoehorned by an adult into a group of kids where you clearly don't fit. Depending on the grade level of the "gifted student" it may be too late to integrate him into his peers. In many settings you'll see "clicks" where certain groups only hang out with each other. If the kid is alienated from the groups, it will be extremely difficult for him to find his place. Your best bet may be to drop him into some organized social activities. I used to play Magic: The Gathering, often. A local card shop holds tournaments. Not bad for building social skills, also i'm sure there are a lot of kids similar to him that hang out at those type of events. At least that way he would have something in common with the kids he interacts with.
Peel away the language? What are you talking about?
I'd argue the exact opposite. Who the heck would want to program in C when you have C++? Think of all the ADTs that C doesn't provide. Vectors, Stacks, Queues. C doesn't even provide a properly abstracted string! Not to mention templates; I often hear lack of templates as a serious complaint against java.
Java is a whole new can of worms. IMHO, language-wise, it provides a lot of things that C++ lacks. However, you have the overhead of running the Java VM, which has a track record for being bloated, cpu-intensive, and memory intensive. For end-user apps, i wouldn't want them to be stuck running the VM.
First, a default debian install, as long as you choose not to run dselect/tasksel, is MORE barebones than gentoo by far. A default debian install does not include gcc or ssh. Slackware uses BSD style init scripts (not traditional SysV init style). Although a rare problem, this causes compatibility problems in some apps. TGZ packages are not as common as debian's DEBs, nor as plentiful. Gentoo, every package update must be recompiled. A security bug in apache will require you to recompile all of apache. Debian's stable releases are supported with security backports for years on end.
Its only problem is real enteprise support for any of the distros, which I'm pretty sure is buyable from other people. Assuming you can get it with enteprise support, debian is THE distro for the enteprise.
Ignoring the legal side of things for a minute, it would be possible to block unrequested popups at the ISP level. The filter could run through a simple http proxy and merely strip out the offending javascript code.
Of course, the devil is in the details, I would venture to say that blocking all unrequested popups would be difficult with more obfuscated javascript code. Of course, the more complicated processing you perform, the more CPU you require. Unless you had a really really big proxy server cluster, or very few users, I imagine even casual web browsing would saturate any one server. Obviously, you would be text parsing every web page that passes through your proxy. That's some pretty intense workload.
The whole deal with users who don't want the feature, just don't make it a transparent proxy. With your ISP installation cd, you can automatically change the IE connection settings to use your special popup kiling proxy, but by default, it will just be regular internet. Saavy users could turn it off at their discretion.
OSS solutions are available. Just from a search at freshmeat, there's middleman, an http/https proxy which, among other things, can filter out popups.
I frankly would love to see something like this in production. You're absolutely right in that you cannot trust users with their own machines. Recommending alternative browsers and addons will invariably leave you supporting them, this is a huge pain in the neck and could be avoided with a centralized solution as you have suggested.
It tends to work better with some shows than others. At one point where i was recording Buffy the Vampire Slayer reruns (channel=FX), the thing would often skip commercials properly, but then also skip a few minutes into the show as well. Then when i'd record say, Justice League (channel=CartoonNetwork), it would skip the commericals perfectly. So yea, I'd say 80% is pretty accurate. I usually enable manual commerical skip, and just rewind a few steps if it overskips.
The problem with extending SMTP is obvious. Where do you keep backwards compatibility? Lets say you introduce encrypted e-mail. What happens when your email passes through an unencrypted relay? Lets say you don't allow passage through an unencrypted relay. You lose the ability to send email to some people.
So now lets say you introduce sender verification. You get the same type of relay issues. You could whitelist certain relays, but at that point, you're talking about whitelisting e-mail, highly annoying, not always feasable.
If you ask me, short of remodeling the entire internet, we're pretty much screwed when it comes to e-mail.
Some Apple employees loaded Mac OS X Server onto one of the early iPods and connected it to a desktop Mac. Then, they booted to it. It ran.
I hope that all the folks who always seem troll on Apple product, saying that all they do is slap on some pretty exterior, jack up the prices, and market, market, market, will think for a moment and appreciate the depth of this product.
How does that constitute depth? It's a firewire hard drive! My cheap, ugly archos jukebox studio is a usb hard drive. If my PC could boot from a usb hard drive, I could do the exact same thing with windows.
I've found I can usually fit 20 or so of my songs on a cd. I have a 4x burner. It takes 15 minutes to burn a cd. If I bought music from itunes, i'd probably buy in albums, I could rack up maybe 100 songs in no time. That's 5 cds, taking 75 minutes to burn. This doesn't include the time it would take to rip and reencode every single song.
There's no question that apple is restricting the use of the music. The question is if this is acceptable or not.
I think this is the wrong kind of comparison to be making. I for one never saw linux as a complete windows replacement. The desktop side of things is sort of a slowly evolving blob, as far as I've ever seen. Luckily "how it looks," isn't the reason why most people claim to use linux.
Apple is a corporation. MacOS is (for the most part) closed source. That means that when you purchase software from them, you are stuck getting updates and fixes from them. I wouldn't say apple is as bad as microsoft, not yet anyway, but lots of stuff points out that Apple is working towards locking in their users. For example, the fiasco about security updates to the older versions of OSX a few days ago.
I always hear people chastising Microsoft about their evil DRM-enhanced future. I don't see why people don't notice Apple doing it RIGHT NOW. Look at iTunes. You can burn your music, or you can put it on your iPod. I have an Archos mp3 player. I can't put music I buy from iTunes on it even though I've purchased the music. I'm by no means putting down Apple's use of DRM, after all, they have to make money somehow, but it's important to realize that they are just another corporation, and in parallel, they are just out to make money.
With open source you don't have to rely on a central source for fixes, you can fix it yourself, you can modify the appliations to suit your needs and whatnot. MacOS X may look nice, but it's no develper's heaven. That's what linux is for.
OSI is good. I usually work down. Start by profiling all the servers, basic hardware spec on each one, applications running on each, if unix -> snag a copy of all major conf files, list of major server apps... windows -> major server apps/basic configuration info. Sometimes (if deemed necessary), make a step-list of things to accomplish basic tasks, like adding users to an AD or unix box. Afterwards, maybe a visio diagram showing all the hubs/switches and server connections to each. Then down to each hub/switch, labelling ports.
I've got one of those... I'm not really happy with it. The membrane is like extra tough, so it makes the keys feel like they stick. It's not like the buckling spring ones either, where the keys feel looser. The keyboard has this sort of tight sticky feel to it...
It's really a huge scam. Athlon 2500 is really 1.83ghz. It is definetly resonable competition to the the the 2.5ghz range p4s, but still, it's not a 2.5ghz processor. I occasionally sell machines I build for people. I have to tell them: No, it's not a pentium 4, it's an athlon xp with speed roughly equivilant to a 2.5ghz pentium 4. It just makes it sound cheaper. No suprise that people view it as "budget" parts.
The one thing that nearly everyone fails to realize is that linux is not a collective group effort. Like slashdot, it's made up of lots of different groups of people, all with their own unique beliefs and thoughts. "Linux" can't just pull their act together. There will be no call to arms. There will be no removal of choice. The reason that you can't embed an avi that plays via xine in kpresentator isn't because the developers didn't think of it as a potential feature. It's also not because everyone is at war. Integration doesn't occur simply because it's hard. I'm sure you're all familiar with people who "resign from the scene." It's because most of the developers have a political agenda, and are stubborn and lazy. They're not gonig to go out of their way to try to integrate with a window environment that simply isn't their own project. As far as I've seen, the extent of cooperation is just when someone decides they want a feature, they implement it, and send a patch in. KDE isn't going to magicly spawn integration with Gnome. Someone has to write the patches. Standards are pointless when they are not followed.
I've grown sick and tired of these call to arms articles assuming they know the solution to everything. If they want the features, write the patches. If you want more polished applications, write them. You didn't pay for them, you don't deserve the right to complain.
They originally cited "Security Reasons" for this whole fiasco. Gaim says new version of protocol is over SSL. Maybe you need a bit more faith in the devil, eh?
I was interning in a mid-sized business in NYC, they run NT4/2000 exclusively. I was in charge of all of the gopher jobs, patching machines, maintaining small VB apps, etc. When people starting making a big deal about the RPC flaw, the sysadmin decided he was going to patch all of the machines, just to be safe. Easy for him to say, since he didn't have to do it.
It was pointless. The machines were firewalled off and there was only 1 guy with a laptop. But I wasn't the sysadmin. I was but a lowly gopher. I had no problem with the windows 2000 machines. The NT4 machines, were another story. They hadn't been patched above SP6a, and many had been under heavy use for a undefined length of years.
Out of the maybe... 30-40 machines I patched, I remember 2 severe issues.
After installing the Critical Updates Package on one machine, explorer crashed every time the user logged in. The sysadmin was forced to downgrade the machine back to IE5.0 and remove a few Critical Updates.
After installing the July 2001 Culmulative Security patch, one machine failed to boot. Some error about one of those NT* files failing to load, IIRC. The sysadmin had to recover the important stuff, and ghost his hard drive.
2 out of 35ish may seem like a small margin of error, but the sysadmin told me that he had once worked at a very large company that deployed updates via group policy and rendered every machine in the entire business useless.
If I discovered a new poison that could be easily replicated with over the counter products, do I protect the public by releasing the recipie for it, or for the antidote?
Who would buy an antidote without knowing the problem it cures? To solve a problem, don't you have to KNOW what the problem is in advance? What kind of CS monkey is going to start reimplementing all of their hash table stuff because someone invents a new and potentially slower method of doing it? Obviously, you need to have a reason to begin making changes in software.
Assuming they could get past all the potential technical hurdles regarding security and authentication, we still are basically saying that a private company can alter/damage the contents of a computer legally without any coordination with law enforcement. That scares me.
I don't see why a private company shouldn't have the right to remotely wipe the hard drives of computers that they legally own. How different is this from remote administration of a server? Now it's just in hardware, rather than software.
Are you suggesting that whoever is currently in possession of the laptop has the right to do whatever they want with it?
I'm also the target demographic. Almost none of my friends run XP. I can tell you that none of the "features" of this new app seem to have any worthwhile purpose. I know that while i'm chatting with my friends, I don't want my messaging program to interfere with things that I'm doing. And have a DJ session online? By god, I don't want to listen to that crap my friends listen to...
Slashdot Mirror
Realisticly, you'd be hard pressed to run even windows 98 with less than 32mb of ram on anything slower than a 200mhz pentium. Office XP is not a speed demon either, on a 500mhz celeron, it takes 20ish seconds or so to load up, and has a noticably laggy UI. I haven't ran openoffice lately, but I'm sure it's safe to say that it's no speed demon either.
I'm not convinced that either product is usable with the cited least required systems.
If you're referring to this virus, it polymorphs, meaning it changes it's basic makeup to evade being detected. AFAIK, it doesn't actually touch the virus scanner, it attempts to avoid it.
A serious no-no: Do not `introduce` the student to other students. There is nothing more embaressing than trying to be shoehorned by an adult into a group of kids where you clearly don't fit. Depending on the grade level of the "gifted student" it may be too late to integrate him into his peers. In many settings you'll see "clicks" where certain groups only hang out with each other. If the kid is alienated from the groups, it will be extremely difficult for him to find his place. Your best bet may be to drop him into some organized social activities. I used to play Magic: The Gathering, often. A local card shop holds tournaments. Not bad for building social skills, also i'm sure there are a lot of kids similar to him that hang out at those type of events. At least that way he would have something in common with the kids he interacts with.
Peel away the language? What are you talking about?
I'd argue the exact opposite. Who the heck would want to program in C when you have C++? Think of all the ADTs that C doesn't provide. Vectors, Stacks, Queues. C doesn't even provide a properly abstracted string! Not to mention templates; I often hear lack of templates as a serious complaint against java.
Java is a whole new can of worms. IMHO, language-wise, it provides a lot of things that C++ lacks. However, you have the overhead of running the Java VM, which has a track record for being bloated, cpu-intensive, and memory intensive. For end-user apps, i wouldn't want them to be stuck running the VM.
First, a default debian install, as long as you choose not to run dselect/tasksel, is MORE barebones than gentoo by far. A default debian install does not include gcc or ssh. Slackware uses BSD style init scripts (not traditional SysV init style). Although a rare problem, this causes compatibility problems in some apps. TGZ packages are not as common as debian's DEBs, nor as plentiful. Gentoo, every package update must be recompiled. A security bug in apache will require you to recompile all of apache. Debian's stable releases are supported with security backports for years on end.
Its only problem is real enteprise support for any of the distros, which I'm pretty sure is buyable from other people. Assuming you can get it with enteprise support, debian is THE distro for the enteprise.
Ignoring the legal side of things for a minute, it would be possible to block unrequested popups at the ISP level. The filter could run through a simple http proxy and merely strip out the offending javascript code.
Of course, the devil is in the details, I would venture to say that blocking all unrequested popups would be difficult with more obfuscated javascript code. Of course, the more complicated processing you perform, the more CPU you require. Unless you had a really really big proxy server cluster, or very few users, I imagine even casual web browsing would saturate any one server. Obviously, you would be text parsing every web page that passes through your proxy. That's some pretty intense workload.
The whole deal with users who don't want the feature, just don't make it a transparent proxy. With your ISP installation cd, you can automatically change the IE connection settings to use your special popup kiling proxy, but by default, it will just be regular internet. Saavy users could turn it off at their discretion.
OSS solutions are available. Just from a search at freshmeat, there's middleman, an http/https proxy which, among other things, can filter out popups.
I frankly would love to see something like this in production. You're absolutely right in that you cannot trust users with their own machines. Recommending alternative browsers and addons will invariably leave you supporting them, this is a huge pain in the neck and could be avoided with a centralized solution as you have suggested.
It could be the java plugin, whenever that loads all my cpu resources spike and IE freezes for a few seconds, and i'm on a 2.4ghz P4 w/ 512mb ram.
eesh.
It tends to work better with some shows than others. At one point where i was recording Buffy the Vampire Slayer reruns (channel=FX), the thing would often skip commercials properly, but then also skip a few minutes into the show as well. Then when i'd record say, Justice League (channel=CartoonNetwork), it would skip the commericals perfectly. So yea, I'd say 80% is pretty accurate. I usually enable manual commerical skip, and just rewind a few steps if it overskips.
The problem with extending SMTP is obvious. Where do you keep backwards compatibility? Lets say you introduce encrypted e-mail. What happens when your email passes through an unencrypted relay? Lets say you don't allow passage through an unencrypted relay. You lose the ability to send email to some people.
So now lets say you introduce sender verification. You get the same type of relay issues. You could whitelist certain relays, but at that point, you're talking about whitelisting e-mail, highly annoying, not always feasable.
If you ask me, short of remodeling the entire internet, we're pretty much screwed when it comes to e-mail.
I've found I can usually fit 20 or so of my songs on a cd. I have a 4x burner. It takes 15 minutes to burn a cd. If I bought music from itunes, i'd probably buy in albums, I could rack up maybe 100 songs in no time. That's 5 cds, taking 75 minutes to burn. This doesn't include the time it would take to rip and reencode every single song.
There's no question that apple is restricting the use of the music. The question is if this is acceptable or not.
I think this is the wrong kind of comparison to be making. I for one never saw linux as a complete windows replacement. The desktop side of things is sort of a slowly evolving blob, as far as I've ever seen. Luckily "how it looks," isn't the reason why most people claim to use linux.
Apple is a corporation. MacOS is (for the most part) closed source. That means that when you purchase software from them, you are stuck getting updates and fixes from them. I wouldn't say apple is as bad as microsoft, not yet anyway, but lots of stuff points out that Apple is working towards locking in their users. For example, the fiasco about security updates to the older versions of OSX a few days ago.
I always hear people chastising Microsoft about their evil DRM-enhanced future. I don't see why people don't notice Apple doing it RIGHT NOW. Look at iTunes. You can burn your music, or you can put it on your iPod. I have an Archos mp3 player. I can't put music I buy from iTunes on it even though I've purchased the music. I'm by no means putting down Apple's use of DRM, after all, they have to make money somehow, but it's important to realize that they are just another corporation, and in parallel, they are just out to make money.
With open source you don't have to rely on a central source for fixes, you can fix it yourself, you can modify the appliations to suit your needs and whatnot. MacOS X may look nice, but it's no develper's heaven. That's what linux is for.
OSI is good. I usually work down. Start by profiling all the servers, basic hardware spec on each one, applications running on each, if unix -> snag a copy of all major conf files, list of major server apps... windows -> major server apps/basic configuration info. Sometimes (if deemed necessary), make a step-list of things to accomplish basic tasks, like adding users to an AD or unix box. Afterwards, maybe a visio diagram showing all the hubs/switches and server connections to each. Then down to each hub/switch, labelling ports.
Seems to work alright..
I've got one of those... I'm not really happy with it. The membrane is like extra tough, so it makes the keys feel like they stick. It's not like the buckling spring ones either, where the keys feel looser. The keyboard has this sort of tight sticky feel to it...
At least, that's what i think
It's really a huge scam. Athlon 2500 is really 1.83ghz. It is definetly resonable competition to the the the 2.5ghz range p4s, but still, it's not a 2.5ghz processor. I occasionally sell machines I build for people. I have to tell them: No, it's not a pentium 4, it's an athlon xp with speed roughly equivilant to a 2.5ghz pentium 4. It just makes it sound cheaper. No suprise that people view it as "budget" parts.
The one thing that nearly everyone fails to realize is that linux is not a collective group effort. Like slashdot, it's made up of lots of different groups of people, all with their own unique beliefs and thoughts. "Linux" can't just pull their act together. There will be no call to arms. There will be no removal of choice. The reason that you can't embed an avi that plays via xine in kpresentator isn't because the developers didn't think of it as a potential feature. It's also not because everyone is at war. Integration doesn't occur simply because it's hard. I'm sure you're all familiar with people who "resign from the scene." It's because most of the developers have a political agenda, and are stubborn and lazy. They're not gonig to go out of their way to try to integrate with a window environment that simply isn't their own project. As far as I've seen, the extent of cooperation is just when someone decides they want a feature, they implement it, and send a patch in. KDE isn't going to magicly spawn integration with Gnome. Someone has to write the patches. Standards are pointless when they are not followed.
I've grown sick and tired of these call to arms articles assuming they know the solution to everything. If they want the features, write the patches. If you want more polished applications, write them. You didn't pay for them, you don't deserve the right to complain.
They originally cited "Security Reasons" for this whole fiasco. Gaim says new version of protocol is over SSL. Maybe you need a bit more faith in the devil, eh?
True story,
I was interning in a mid-sized business in NYC, they run NT4/2000 exclusively. I was in charge of all of the gopher jobs, patching machines, maintaining small VB apps, etc. When people starting making a big deal about the RPC flaw, the sysadmin decided he was going to patch all of the machines, just to be safe. Easy for him to say, since he didn't have to do it.
It was pointless. The machines were firewalled off and there was only 1 guy with a laptop. But I wasn't the sysadmin. I was but a lowly gopher. I had no problem with the windows 2000 machines. The NT4 machines, were another story. They hadn't been patched above SP6a, and many had been under heavy use for a undefined length of years.
Out of the maybe... 30-40 machines I patched, I remember 2 severe issues.
After installing the Critical Updates Package on one machine, explorer crashed every time the user logged in. The sysadmin was forced to downgrade the machine back to IE5.0 and remove a few Critical Updates.
After installing the July 2001 Culmulative Security patch, one machine failed to boot. Some error about one of those NT* files failing to load, IIRC. The sysadmin had to recover the important stuff, and ghost his hard drive.
2 out of 35ish may seem like a small margin of error, but the sysadmin told me that he had once worked at a very large company that deployed updates via group policy and rendered every machine in the entire business useless.
Beware the updates.
it's not uncommon that companies use embedded linux load balancer box'en in front of their servers
Win9x = 9x kernel
WinXP = NT kernel
Almost completely different beast.
Plus, what's to say this bug isn't simply a problem with vmware?
If I discovered a new poison that could be easily replicated with over the counter products, do I protect the public by releasing the recipie for it, or for the antidote?
Who would buy an antidote without knowing the problem it cures? To solve a problem, don't you have to KNOW what the problem is in advance? What kind of CS monkey is going to start reimplementing all of their hash table stuff because someone invents a new and potentially slower method of doing it? Obviously, you need to have a reason to begin making changes in software.
Assuming they could get past all the potential technical hurdles regarding security and authentication, we still are basically saying that a private company can alter/damage the contents of a computer legally without any coordination with law enforcement. That scares me.
I don't see why a private company shouldn't have the right to remotely wipe the hard drives of computers that they legally own. How different is this from remote administration of a server? Now it's just in hardware, rather than software.
Are you suggesting that whoever is currently in possession of the laptop has the right to do whatever they want with it?
ActiveX support is only for windows, this has been in development for some time, as seen here.
Assuming nothing is changed, the activex control support is a plugin and is thus, optional. For now....
I have two friends that both have to constantly reboot their linksys routers.... Never had a problem with my D-Link
I'm also the target demographic. Almost none of my friends run XP. I can tell you that none of the "features" of this new app seem to have any worthwhile purpose. I know that while i'm chatting with my friends, I don't want my messaging program to interfere with things that I'm doing. And have a DJ session online? By god, I don't want to listen to that crap my friends listen to...