We require a professional license for these other occupations because of how easy it is to be conned, and because of how much damage they can do. I consider some computer-related occupations to be just as dangerous.
Setting up a computer system or network can cause monetary dammage, but (except in specific circumstances) is not going to cause injury or death.
Where you can teach yourself C and learn your networking at home in your basement, you cannot become a doctor through self-education.
A doctor is a very specialized profession. Doing something wrong won't simply cost money, but potentially someone's life.
Electricians, building contractors, plumbers, etc require some sort of government-based certification because these things have the potential to cause injury if they aren't done correctly. In all but the most extreme cases, a bad programmer or systems/network engineer will only cause monetary dammage. In which case, the fault lies (at least partially) on the person who hired/contracted/out-sourced to the unqualified individual or company.
The government (in most countries) will protect you from making a bad decision that could cost your life. You can't start a medical practice without being licensed/certified, even if you post that fact. But to ask the government to protect you from losing money? That's above and beyond what the government is there to do (and I wouldn't want that to change).
Re:But why would non-geeks want to run Linux?
on
Linux for Non-Geeks
·
· Score: 2, Insightful
Actually, given how bad windows security has been lately, I'd recommend that most users not use windows unless [they're] geeks and know how to keep it clean, and free of Spyware.
I agree somewhat, but I do believe things are getting better.
On the one hand, I do know many people with (relatively) new XP-based machines that, upon a "system restore" cannot connect to the 'Net long enough to get the necessary updates. Of course I end up walking them through enabling the firewall, or (as was the case today with my cousin) aborting the shutdown that one of the XP/RPC-based worms causes...
...but, SP2 does promise to at least enable the firewall by default. Granted, a properly configured machine shouldn't require a firewall, but I wouldn't run a Windows box without one (and preferrably either a hardware-based firewall, or an external *nix box). But it's a great start.
Had XP (or 2000) had an enabled firewall by default, we'd have been much better off, all the way back to Code Red/Nimda up to the latest ones (I stopped keeping up any more, but there was an RPC based one, and one that hit LSASS.EXE, etc)... so like I said, it's a great start.
Also don't forget the Honeypot project a couple of (few?) years back. IIRC (which I probably don't) they connected a RedHat 6.something box to the 'Net, and it wasn't 30 minutes before it was rooted via an automated worm (or whatever)...
The point is this: Microsoft is *finally* getting some things right. They may be sacrificing some ancient backward-compatibility in favor of security, but they're finally doing some things right. Granted, every other OS in the world is far ahead in most of these areas, but it still is nice that MS is getting there...
It will be nice to not cringe whenever I hear a relative say "I'm finally getting a computer!"...
...my Honda is metric, and I have a set of metric tools to deal with that.
I personally grew up working on American cars (GM, Ford, Dodge) and using the "standard" measurement system exclusively. As I got into engineering-related areas, I've found it necessary to learn the Metric system, and the appropriate conversions.
As well, I've gotten to like foreign cars (Toyota specifically -- you can't kill them!), and I think you'll find most auto mechanics -- and I am not one -- easily capable of converting millimeters to fractions of an inch. Regardless of their normal mathematical skills, by association and eventual familiarity, these things are easily picked up. Not to mention, most American cars I've worked on recently seem to use either all metric or -- worse -- a combination of standards (too often the engine is Japanese while the rest is American...)
But the point of that was that the conversions aren't difficult, and (despite what many non-US people tend to imply) Americans are perfectly capable of learning to convert, or learning a new system of measurement.
It's not lack of intelligence or lack of will, but lack of necessity, that keeps most US citizens from converting. We all realize that 100 km == 60 miles, if only because of the jokes commedians make about driving into Canada and seeing a speed limit sign of 100. Plus most cars sold here display speed in both measurements, though admittedly the km portion is usually much less prominant......but again, it's lack of necessity. Lumber here is generally sold in "standard" units: an 8-foot 2x4 for example. It's an unnecessary difficulty to just decide to use metric units, when much of what you work with is non-metric. At that point it just adds an unnecessary (and potentially inaccurate or error-prone) conversion.
Granted, if the US government mandates the use of metric units, we'd have the necessary push. However, I suspect many would oppose a conversion being required by law in this country if it isn't shown to be absolutely necessary.
In the public (non-government) area, it becomes a "chicken vs egg" scenerio. No lumber yard is going to sell lumber measured exclusively in metric units, and no building contractor is going to ask for several 79mm X 157mm X 244cm boards to build a wall when 2x4x8' is the standard of measurement here in the US for such materials (I'm not sure my conversions were correct there btw).
Anyway, it's not a stupidity or laziness factor as many non-US people assume, rather it's that our current system is very much established and ingrained into our society that it's difficult to change without making laws to require the change, which most citizens would disagree is all that necessary. In engineering fields, it's a different story, and if you're an engineer you likely should know both systems (since you will likely deal with both). Even in electronics, some standard measurements are in "mils" (thousandths of an inch) while others are in millimeters...
Anyway I'm just ranting because I'm still awake for some odd reason.
BTW, this was not directed at the parent; the Honda/metric comment just inspired me to rant for a bit:)
I can't say for sure (I don't use SMS intentionally, and the bill isn't handy) but I am pretty sure I get charged for incoming messages as well as outgoing... hence my sister's most common message (she thinks she's funny): "This just cost you 10 cents dork".
I'm sure there's got to be *some* way to simply not accept the messages (either in the phone itself or via Cingular), but it hasn't become a problem yet aside from my sister having fun when I first got my phone... though I would be a bit miffed if it turned out I had to simply accept the messages and associated fees...
What's stopping someone from unleashing a real SMS attack (not like this goof in the article) and sending hundreds of thousands of messages to receipients who then have to pay for?
Nothing stops you from this, just like nothing stops you from prank-calling random numbers -- except of course that it's illegal to do so (not sure if/how this applies to prank-SMS though; that might even just fall under unsolicited calls? hm...)
Again, I'm not fully sure that I do pay for incoming messages, but it does seem that I read that somewhere in the contract, and that it was like 10 cents per message received, or $5 extra a month adds some 100 SMS messages.
One of the most important facts is: The child abuse was already done, when the pictures got posted.
Yes, that particular abuse. But think about the type of person who enjoys this type of porn. Would you want that person to babysit your child?
Besides, the fact that the illegal act itself (filming/photographing child porn) has already taken place doesn't make the publishing or viewing of such material any less illegal (or troublesome). And it certainly doesn't make me feel better about leaving my kids with someone who enjoys this type of content.
Text messages are less intrusive, people can answer them when they have the time.
I disagree personally, but I'm sure a lot of people agree with this. For me, if I don't want to be disturbed I'll put the phone in silent (vibrator) mode, and simply not answer it (or hit "ignore") if it's someone I don't want/need to talk to right now. If I really truly don't want to be disturbed by anyone the phone is turned off.
Text messages piss me off (my sister constantly sends me them) because they aren't part of my calling plan -- thus, each message costs me money, though I'm trying to find out if there's some way to tell Cingular that I simply don't want to accept them. Plus, typing on the keypad is difficult enough when adding a contact, much less trying to type out a sentance/paragraph that could have been covered with a 10 second phone call.
ISPs avoid competitive disadvantage by advertising services as "unlimited". ISPs regulate use by casting "over-use" as abuse which will be curtailed.
Actually (I used to work for a hosting provider), it seems that the latest trend is to offer "unmetered" bandwidth -- which I'm not entirely sure I get but I'm sure there's a loophole. I suspect (much like "unlimited") that once you hit a certain amount of data transfer, they suddenly start metering (which of course implies that you've been metered all along anyway)...
The host I worked for changed their plans in 1998, specifying a specific monthly data transfer included with each plan, and the cost per gig beyond that. It was hard to compete with "unlimited" or "unmetered" hosts, but the more tech savvy (majority of our customers) appreciated knowing what they were getting in advance.
As for unlimited text messaging -- how much could it possibly cost to process a text message? Seems it would be easier than a phone call (routing, etc), yet a phone call is most times cheaper or even free (from the same cell phones). That's what I don't get......and even though I don't have any text messaging included with my plan, I still receive them and it costs me 10 cents or so per message received (and I usually respond by calling the person and asking them to just call me in the future... much easier and cheaper that way...
I think this entire 'unlimited' offering is silly...
Agreed. If the company providing the service could truly offer "unlimited" service (can't even think of any examples, but I'm sure they exist) then it's fine. In most cases, however, "unlimited" simply doesn't fly, and you'll find (especially in the web hosting/ISP business) deep in the AUP/TOS something like "...unless you use more than x in one month...", eg, "unlimited as long as you stay within the limites".
I can't see that text messages could possibly cost that much to process (my provider (Cingular), where I do not have text messaging as part of my plan, charges 10 cents per message). It's simple ASCII text, generally very short, and has to use far less bandwidth than a phone conversaion. Yet, a phone conversation to the very same person you're text messaging with would be a lot cheaper (or pretty much free)... I think they're charging crazy fees simply because it's a new fad, and they can...
You're either can't be serious, or you're just too old to adopt new technologies.
Speaking for myself, I have no use for text messaging. My calling plan doesn't include it, so when I receive one (apparently I can't turn that off, and it costs me 10 cents), I usually message back with "Just call me dammit", since it is then free (or very cheap) compared to silly text messages.
I do agree that the idea of a land-line only simply doesn't work for many people (myself included). Cell phones are just so damned handy for so many reasons; I got mine the day after my car broke down and I was forced to walk a good bit, but since then I've come to really love having it. And (since I can never use the ungodly minutes included) it costs less than my basic home-phone plan (which I'm considering getting rid of)...
Somehow, users seem to have the blame pinned on them instead of developers. This seems unique to the software industry--even the auto industry has more clue.
I agree to a point. If a flaw in an automobile caused some sort of problem the manufacturer is certainly responsible regardless of what the driver may have done to trigger the problem.
However, there's also more responsibility required of the user (driver). I don't think computing should require a license or anything like that, but you don't drive a car unless you know what you're doing. There are many basic rules you have to learn.
PC users don't want to learn things like software updating, firewalls, what file types not to accept via email, etc. Just like many drivers choose not to learn (or blatantly ignore?) things like "keep right except to pass", or that the center turn lane is NOT a merge lane...
Really you should put Windows systems behind an external firewall (that is, a firewall that doesn't run on Windows), such as a hardware firewall or a *nix box of some kind.
Absolutely agreed. I am glad that many cable/DSL modems have built-in firewalling (or at least NAT). I'm personally behind a Linux box with iptables, and the only time I worry is when I bring in a friend's PC to my local network to fix it (almost always an infected machine) (and I disconnect my Windows box from the network before connecting the infected box).
What's hilarious is when a company is "protecting" a Windows network with a firewall running on -- you guessed it -- Windows. A friend's workplace was hit hard with one of the recent worms because the firewall machine became infected, and proceeded to infect the entire network. That's some great protection...
In SP2 MS claims it will be on by default. That will of course cause many problems for people, but it will hopefully stop the rapid spread of some worms too.
Yes, it may cause problems for *some* users, but the vast majority of XP users have absolutely no need for any services to be listening on a public interface. Those who do should easily know how to disable the firewall or allow the specific port(s) they need.
It should have been on by default all along. Think of all the worms that would never have happened if, say, Windows 2000 and XP had a simple firewall on by default.
The negative side (users complaining because they didn't know to turn off the firewall when it's in their way) would have caused far far less PR problems for Microsoft than that caused by the many remote exploits in the last few years.
There's been no shortage of buffer-overflow style attacks against Linux. The difference is a) there's far fewer machines out there to target and b) the users of those machines are far more likely to either have taken preventative measures or know how to identify and fix exploited machines.
Also don't forget c) there is a lot more diversity in Linux machines. You can't count on the fact that a significant percentage of Linux machines are running a particular service or program (or a particular vulnerable version thereof).
In Windows you can't be running any other version of LSASS.EXE than the one provided; every XP box in the world will have this vulnerability until patched.
My advice would be not to bother if you live in an area where Comcast and SBC "Yahoo"... are your only choices.
Off topic, but when it comes to Cable/DSL providers, I've always read mixed results from every provider. I suspect it comes down to who, specifically, is running the local equipment...
I'm with Comcast currently and the service has yet to have an issue (about 6 months now). I had Sprint/Earthlink DSL back in Florida, Road Runner cable (Time Warner) also in FL, and Bell South DSL in Georgia. Road Runner gave the most problems, almost always due to "upgrades", but even then it wasn't frequent enough to complain (with a "residential" account).
Note, I do run my own local DNS cache as I found out (early on) that in most cases when service was down it was simply DNS not working (in many cases the physical connection is provided by one company (eg, Sprint) and DNS/email/etc by another (Earthlink in the Sprint case)).
The one thing that bothers me about VOIP is that (so far, in my experience) a cable connection requires power. My DSL always worked w/o power (modem, switch, router etc on UPS) where a cable connection simply didn't work without power. But, with my cell phone being cheaper all around, I am tempted to get rid of the POTS line. Cheap as the cell service is, I've not yet seen a need to go VOIP personally.
Windows is a stupid choice for a headless compute node...
On that I will agree; unless there is a very specific reason to use Windows for a cluster (or server or...), I can think of no reason to have an OS that requires a video card (and drivers), and prompts the first time you boot without a pointing device connected, on a system that requires no interface or direct interaction.
...just as Linux is a stupid choice for a home desktop.
On this I don't agree. For you perhaps. For me even, in most cases: I run Linux (and FreeBSD) on my servers, and Windows 2000 and XP on my desktops (laptop is dual-boot XP/Fedora). However, there are plenty of good reasons to go with Linux (or BSD) on a desktop system.
I would agree that as a pre-install, or on a desktop for a user who doesn't know Linux (and will be angry that they can't run the latest Windows-based spyware-riddled game) it's not a great choice. But I wouldn't just generalize that "Linux is a stupid choice", because there are times where Linux is a good choice.
I've set more than a couple of desktop users up with Linux -- specifically, unsophisticated people who only need to check email and browse the web, and are using older machines (that I donated in most cases). And in all of these cases, the user really didn't notice any difference (and they constantly ask if the latest virus they heard about on the news affects them).
Not true. You assume people are intelligent, able to install things, and interested in doing so in the first place.
I agree with all but the intelligent part. It's not that people aren't intelligent, it's more that they have little or no interest in learning that much about computers. For those of us where computers are a large part of our lives and/or careers, these things seem obvious, just like a doctor might think some complex medical procedure is obvious...
It's rather improbable that the average person has the skills or the motivation to hunt for the correct tool.
That's a much better spin on it that I can agree with fully. Just remember, skills != intelligence. Nor does lack of motivation imply lack of intelligence; lack of interest perhaps...
Add to that the fact that most people are quite lazy. If the page looks weird they just forget about browsing the site and go somewhere else.
I do that myself, but in my case most times the page was designed for IE exclusively or requires flash, whereas I use Mozilla w/o the flash plugin. These days I find much less rendering problems, either because sites are more compliant (doubtful) or because Mozilla and IE interpret buggy/incomplete code in similar ways...
If Microsoft actually continued to add standard compliant technology to their browser we'd have a situation where we could actually adopt new technologies at a sensible rate. As it is now this is hardly the case.
Agreed fully. IE has, in my opinion, stagnated since version 5 or so, and no major feature enhancements (that I would use anyway) since 4.0. I forget which version implemented Microsoft's idea of CSS2 support, but in any case it's still not complete. Plus it lacks so many useful features that most other browsers have (disallowing unrequested popups, tabbed interface, etc)...
As for HTML, JPEGs and PDF's being all we need... you really haven't done much in terms of real world web design have you? Sure, you'll get the information across, but that just won't cut it in the current market (this claim is naturally not without exceptions.)
Standard and simple technologies are sufficient for many purposes, but in general I agree that we shouldn't be stuck with 1998 technologies. I like knowing that I can use certain advanced features, but in more than one case I've had to pull a cool CSS trick out of a page after finding out that IE doesn't support it (or doesn't support it properly). Often the work-around is to re-implement it using JavaScript, which isn't worth the effort IMO.
Just look at the new features we've gained over the last five years (that are actually being used). Then look at the five years prior to that (1994 to 1999) and you'll notice that things have stagnated around the time IE became stagnant. Granted we're still seeing many technologies becoming more widely used (eg, CSS), but only because the market generally has to catch up. By the time we get anything new (at the mercy of Microsoft most likely), we'll see actual development stagnate for a while, while we wait for the new tech to mature, and for there to be enough users with a newer browser...
At least that's my opinion, based on observation... however I do feel that HTML/CSS is currently pretty darned flexible if you learn the right tricks; it could be worse (it's not like we're stuck with plain ASCII text).
In summary, I fully agree with you and wanted to comment on a few points, and to point out that I don't feel that it's lack of intelligence, but rather specific knowledge that many users don't care to know (nor should they have to, ideally). Computers are tools to most people...
I do believe you used intelligence unintentionally to mean knowledge, as the other points in your post seem to agree.
Oh, and I hate PDF. More specifically, the idea (portability) is great, they print nice, but I cannot stand the Acrobat viewer (I wonder if there's another Windows viewer out there somewhere...)
(I really should get to bed soon; I noticed I'm rambling quite a bit here...)
I tested and can only send up to 5 MB. I tried a 9 and then a 7 MB attachment, but each time it told me I had exceeded the 10 MB limit - which obviously wasn't the case...
Email attachments are encoded (usually base64), which adds quite a bit of overhead. Someone else said it was around 30%, but I believe it may be more than that for binary attachments...
So your 7MB message, when encoded, likely does in fact exceed 10MB...
Isn't redirecting information to/dev/null technically destroying it?
I think the point was that there's a difference between destroying data before it's requested, and destroying it when/after it's requested.
Document/data retention policies are written for just this purpose. If the FBI (or whoever) asks for data, and you (as a business policy) don't normally keep that data (whether at all, or beyond some set period of time), then you are generally okay. If OTOH the data is destroyed (or comes up "missing") when it's requested, you may be in for some legal trouble.
There is no law that I'm aware of requiring one to keep data/logs, except in specific cases. However, it should be provable that the data is missing in accordance to an already established policy if you want anyone to believe that the data wasn't destroyed simply because it was requested...
If an application generates some logging info, then you have it in your possesion, sending it to/dev/null is therefore destruction of data you own.
And the point is to be consistant. If you always direct your log information to/dev/null, and when asked for some specific portion of data you mention your policy, you shouldn't have anything to worry about. It's only when you normally *do* keep data, but the requested data is suddenly "not available", that you begin to look suspicious.
Granted it's somewhat suspicious in itself when one doesn't keep any logs that are normally kept (for statistics/accounting purposes, etc), but that alone (if consistant) isn't grounds for accusation AFAIK...
Either way by making yourself insecure like this you are just adding to the many problems of the Internet.
Nail on head. Please refer to the SMTP protocol and open relays as a prime example.
I think if one is running a network of any kind -- free or otherwise -- one should be responsible for it. If you're not prepared to be responsible, don't offer the service.
I don't think it's very responsible to (as a bad analogy, as most are) leave a loaded weapon out in the open in the hopes that if someone used it for an illegal purpose that you'd be immune to any consequences.
It's one thing to have an open network due to ignorance, and another to just say "hey, if someone does something bad with it, I'm not responsible". Knowingly allowing people to use your resources for "bad" things (illegal or otherwise) is asking for trouble IMO.
Whether it's illegal or not isn't the issue here. I'm not on any kind of moral high-horse, but I don't see this as any differnt from running an open SMTP relay, or leaving a gun out in the open... in either case -- whether by ignorance or using it as an excuse to escape liability -- it's wrong. In my opinion anyway...
Personally, popup ads will tend to mean I'll never buy something from the company, both on principle (stop shitting over my desktop popup advertisers!) and because I worry about the potential ethics of said company (if they are prepared to kick my door in and force me to watch ads, will they run off with my wallet also?). However from time to time, I'll click out of curiosity on a tasteful non intrusive banner, specifically the google style text ones that arent in the face and are informative.
I couldn't agree more with this. I haven't seen a popup ad or animated Flash ad in quite a while (switched to Mozilla since 0.95, Flash/Java disabled) but one company I refuse to do business with is X10. Sure their little cameras look cool, but before I switched from IE, their ads were (at the time) so freakin' annoying and pervasive that I just can't see giving them any sort of money.
Google has it right. Their ads are text based, and they are picky about what kind of ads they will even show. I've been on both sides, and as an advertiser I once received a notice that the word "FREE" with all caps (as in, "FREE download") was potentially problematic. If they're that picky about that, I'm sure the ads I see go through the same type of checking before they're shown to me. In fact I pay more attention to Google ads (always clearly marked as such) far more than any other type of ad, based solely on their reputation with me personally.
Targetting people who you know don't want your ads (getting around popup blockers or spam filters, or ignoring the Do-Not-Call lists, etc) is simply not a good marketing move IMO. You can't sell to people who have (in some way) stated that they have no interest in your ads...
However, I have no problem with "normal" ads. Ads that aren't annoying (many TV/radio commercials are as bad as popup windows/Flash ads), that simply state what you're offering and possibly why I might want/need your service, are the types of ads I would respond to. Google's ads are almost always in this category, as are some (but not many) radio/TV ads, and even some normal static/animated banner ads on the 'Net. Bogus error messages or bogus contests ("You are visitor xxx..." or "Punch the Monkey") are even worse IMO...
The point is, if you have to try that hard to get me to notice your product, your product likely isn't good enough on its own to warrant a look. Other times (X10) the product itself may be worth the monetary cost, but I refuse to do business with a company who thinks this type of behavior is acceptable. Supporting such a company is (directly or indirectly) supporting these annoying advertising techniques, thus making them appear effective, and contributing to the problem.
Hell, I forgot the topic at hand here, but once I get started on advertising ('Net and elsewhere) I tend to rant...
GOTO's make spaghetti code. It is very hard to trace through, especially if the code is uncommented. If you end up with a bug, you will have a very hard time trying to trace through your GOTO's to find it.
I disagree. While GOTOs are often abused, they do have their place. I use them in C, infrequently (maybe 2 total in my main C project).
But the key is the uncommented bit: why does GOTO necessarily mean it's uncommented? The two GOTOs I use in my main project are well commented, and are only used in case of an audio card/driver failure. It's easy to follow the program flow, and obvious even if you don't read the comments (which in my project is often more than the code itself).
There's nothing inherent about the GOTO feature that is bad -- rather, it's an easily abused feature. That doesn't mean there aren't good uses for it. I can think of good uses for GOTO in BASIC, C, and other languages. I can also think of many abuses of other features of hundreds of languages...
I think that a company trying this would shoot itself in the foot. Every element would need to be mined from the US, all of the "Intellectual Property" would need to be US-based, and I mean every single little piece.
It wasn't all that long ago that Toyota had commercials showing the fact that most Toyotas sold in the US are actually built in the US. Of course much of the steel came from Japan, but many American car companies had by that point also started using Japanese steel; it was cheaper and higher quality in many cases...
...but I don't believe that Toyota tried to imply that every component was USA-born either, they were mostly stressing the fact that they created a lot of jobs for US workers since at the time, many people wouldn't buy foreign vehicles for that very reason.
Of course these days people don't seem to care about that, so the campaign hasn't been necessary. People buy cars made in Japan, Korea, and elsewhere if they find a good deal (and I think that's a good thing, really).
I don't like my browser or mail client doing things I'm not explicitly aware of. I cannot use Windows with a clear conscience because of IE's and Outlook's persistent security failures.
Just because one uses Windows does not mean they are limited to IE and OE. I personally run XP on my main desktop machine, but I use Mozilla. I've successfully converted more than a handfull of people to Mozilla or Firefox or Opera -- and all Windows users (of course there are the stubborn people like my dad who will inevitably end up using IE). There's nothing that says you ever have to use IE or Outlook Express just because they are preinstalled; on a new install, I fire up IE one time, to go to Mozilla.org. That, and Windows Update...
I don't so much find Windows to be inferior. It's just that Linux and the canon of open source software built upon it make so much more sense financially, socially, and from an engineering standpoint.
I agree to an extent on the engineering standpoint. On a server I'd never ever use Windows (is Hotmail still on BSD? Or Microsoft's own DNS servers?) but for me, Windows does have its place.
My stance is this: use the right tool for the job. Sure, I hope to be able to afford a Mac one day soon, but at the moment Windows runs my desktop (while Linux and/or FreeBSD runs my other four home systems and my servers). Windows is, for me personally, the best tool for the job on the desktop. On my servers, Linux or FreeBSD (depending on the task) are the best tool.
I have no interest in Linux over-taking the desktop. It does a great job for the things I use it for, but Windows does (and I'm sure the Mac does) a fine job on a desktop machine. Security problems, sure, but I patch my WinXP box and I patch my Linux boxes and both are, with proper patching/upgrading, pretty damned stable.
I guess the main problem I had with your post was the implication that one who uses Windows necessarily has to use IE or Outlook, which is simply not true. There are many alternatives to those particular applications, and (at least since 2000/XP) the OS itself is pretty solid...
Slashdot Mirror
We require a professional license for these other occupations because of how easy it is to be conned, and because of how much damage they can do. I consider some computer-related occupations to be just as dangerous.
Setting up a computer system or network can cause monetary dammage, but (except in specific circumstances) is not going to cause injury or death.
Where you can teach yourself C and learn your networking at home in your basement, you cannot become a doctor through self-education.
A doctor is a very specialized profession. Doing something wrong won't simply cost money, but potentially someone's life.
Electricians, building contractors, plumbers, etc require some sort of government-based certification because these things have the potential to cause injury if they aren't done correctly. In all but the most extreme cases, a bad programmer or systems/network engineer will only cause monetary dammage. In which case, the fault lies (at least partially) on the person who hired/contracted/out-sourced to the unqualified individual or company.
The government (in most countries) will protect you from making a bad decision that could cost your life. You can't start a medical practice without being licensed/certified, even if you post that fact. But to ask the government to protect you from losing money? That's above and beyond what the government is there to do (and I wouldn't want that to change).
Actually, given how bad windows security has been lately, I'd recommend that most users not use windows unless [they're] geeks and know how to keep it clean, and free of Spyware.
...but, SP2 does promise to at least enable the firewall by default. Granted, a properly configured machine shouldn't require a firewall, but I wouldn't run a Windows box without one (and preferrably either a hardware-based firewall, or an external *nix box). But it's a great start.
I agree somewhat, but I do believe things are getting better.
On the one hand, I do know many people with (relatively) new XP-based machines that, upon a "system restore" cannot connect to the 'Net long enough to get the necessary updates. Of course I end up walking them through enabling the firewall, or (as was the case today with my cousin) aborting the shutdown that one of the XP/RPC-based worms causes...
Had XP (or 2000) had an enabled firewall by default, we'd have been much better off, all the way back to Code Red/Nimda up to the latest ones (I stopped keeping up any more, but there was an RPC based one, and one that hit LSASS.EXE, etc)... so like I said, it's a great start.
Also don't forget the Honeypot project a couple of (few?) years back. IIRC (which I probably don't) they connected a RedHat 6.something box to the 'Net, and it wasn't 30 minutes before it was rooted via an automated worm (or whatever)...
The point is this: Microsoft is *finally* getting some things right. They may be sacrificing some ancient backward-compatibility in favor of security, but they're finally doing some things right. Granted, every other OS in the world is far ahead in most of these areas, but it still is nice that MS is getting there...
It will be nice to not cringe whenever I hear a relative say "I'm finally getting a computer!"...
...my Honda is metric, and I have a set of metric tools to deal with that.
...but again, it's lack of necessity. Lumber here is generally sold in "standard" units: an 8-foot 2x4 for example. It's an unnecessary difficulty to just decide to use metric units, when much of what you work with is non-metric. At that point it just adds an unnecessary (and potentially inaccurate or error-prone) conversion.
:)
I personally grew up working on American cars (GM, Ford, Dodge) and using the "standard" measurement system exclusively. As I got into engineering-related areas, I've found it necessary to learn the Metric system, and the appropriate conversions.
As well, I've gotten to like foreign cars (Toyota specifically -- you can't kill them!), and I think you'll find most auto mechanics -- and I am not one -- easily capable of converting millimeters to fractions of an inch. Regardless of their normal mathematical skills, by association and eventual familiarity, these things are easily picked up. Not to mention, most American cars I've worked on recently seem to use either all metric or -- worse -- a combination of standards (too often the engine is Japanese while the rest is American...)
But the point of that was that the conversions aren't difficult, and (despite what many non-US people tend to imply) Americans are perfectly capable of learning to convert, or learning a new system of measurement.
It's not lack of intelligence or lack of will, but lack of necessity, that keeps most US citizens from converting. We all realize that 100 km == 60 miles, if only because of the jokes commedians make about driving into Canada and seeing a speed limit sign of 100. Plus most cars sold here display speed in both measurements, though admittedly the km portion is usually much less prominant...
Granted, if the US government mandates the use of metric units, we'd have the necessary push. However, I suspect many would oppose a conversion being required by law in this country if it isn't shown to be absolutely necessary.
In the public (non-government) area, it becomes a "chicken vs egg" scenerio. No lumber yard is going to sell lumber measured exclusively in metric units, and no building contractor is going to ask for several 79mm X 157mm X 244cm boards to build a wall when 2x4x8' is the standard of measurement here in the US for such materials (I'm not sure my conversions were correct there btw).
Anyway, it's not a stupidity or laziness factor as many non-US people assume, rather it's that our current system is very much established and ingrained into our society that it's difficult to change without making laws to require the change, which most citizens would disagree is all that necessary. In engineering fields, it's a different story, and if you're an engineer you likely should know both systems (since you will likely deal with both). Even in electronics, some standard measurements are in "mils" (thousandths of an inch) while others are in millimeters...
Anyway I'm just ranting because I'm still awake for some odd reason.
BTW, this was not directed at the parent; the Honda/metric comment just inspired me to rant for a bit
Isn't the point that searching in your favourite search engine may very well turn up a page with the incorrect conversion factor?
;)
Well, if your favorite search engine happens to be Google, the search engine itself will do the math for you.
But that's just Google...
Who knows what I'll do with all this storage. I guess I won't have to empty my Spam folder as often......
In Yahoo mail, messages in the Bulk folder don't count toward your storage quota. I believe this is a relatively new feature.
I was quite surprised to see the 100MB upgrade this morning, and my first thought was "ah, now they have to compete with GMail"...
Are you telling me your SMS is receiver pays!?
I can't say for sure (I don't use SMS intentionally, and the bill isn't handy) but I am pretty sure I get charged for incoming messages as well as outgoing... hence my sister's most common message (she thinks she's funny): "This just cost you 10 cents dork".
I'm sure there's got to be *some* way to simply not accept the messages (either in the phone itself or via Cingular), but it hasn't become a problem yet aside from my sister having fun when I first got my phone... though I would be a bit miffed if it turned out I had to simply accept the messages and associated fees...
What's stopping someone from unleashing a real SMS attack (not like this goof in the article) and sending hundreds of thousands of messages to receipients who then have to pay for?
Nothing stops you from this, just like nothing stops you from prank-calling random numbers -- except of course that it's illegal to do so (not sure if/how this applies to prank-SMS though; that might even just fall under unsolicited calls? hm...)
Again, I'm not fully sure that I do pay for incoming messages, but it does seem that I read that somewhere in the contract, and that it was like 10 cents per message received, or $5 extra a month adds some 100 SMS messages.
One of the most important facts is: The child abuse was already done, when the pictures got posted.
Yes, that particular abuse. But think about the type of person who enjoys this type of porn. Would you want that person to babysit your child?
Besides, the fact that the illegal act itself (filming/photographing child porn) has already taken place doesn't make the publishing or viewing of such material any less illegal (or troublesome). And it certainly doesn't make me feel better about leaving my kids with someone who enjoys this type of content.
Text messages are less intrusive, people can answer them when they have the time.
I disagree personally, but I'm sure a lot of people agree with this. For me, if I don't want to be disturbed I'll put the phone in silent (vibrator) mode, and simply not answer it (or hit "ignore") if it's someone I don't want/need to talk to right now. If I really truly don't want to be disturbed by anyone the phone is turned off.
Text messages piss me off (my sister constantly sends me them) because they aren't part of my calling plan -- thus, each message costs me money, though I'm trying to find out if there's some way to tell Cingular that I simply don't want to accept them. Plus, typing on the keypad is difficult enough when adding a contact, much less trying to type out a sentance/paragraph that could have been covered with a 10 second phone call.
ISPs avoid competitive disadvantage by advertising services as "unlimited". ISPs regulate use by casting "over-use" as abuse which will be curtailed.
...and even though I don't have any text messaging included with my plan, I still receive them and it costs me 10 cents or so per message received (and I usually respond by calling the person and asking them to just call me in the future... much easier and cheaper that way...
Actually (I used to work for a hosting provider), it seems that the latest trend is to offer "unmetered" bandwidth -- which I'm not entirely sure I get but I'm sure there's a loophole. I suspect (much like "unlimited") that once you hit a certain amount of data transfer, they suddenly start metering (which of course implies that you've been metered all along anyway)...
The host I worked for changed their plans in 1998, specifying a specific monthly data transfer included with each plan, and the cost per gig beyond that. It was hard to compete with "unlimited" or "unmetered" hosts, but the more tech savvy (majority of our customers) appreciated knowing what they were getting in advance.
As for unlimited text messaging -- how much could it possibly cost to process a text message? Seems it would be easier than a phone call (routing, etc), yet a phone call is most times cheaper or even free (from the same cell phones). That's what I don't get...
I think this entire 'unlimited' offering is silly...
Agreed. If the company providing the service could truly offer "unlimited" service (can't even think of any examples, but I'm sure they exist) then it's fine. In most cases, however, "unlimited" simply doesn't fly, and you'll find (especially in the web hosting/ISP business) deep in the AUP/TOS something like "...unless you use more than x in one month...", eg, "unlimited as long as you stay within the limites".
I can't see that text messages could possibly cost that much to process (my provider (Cingular), where I do not have text messaging as part of my plan, charges 10 cents per message). It's simple ASCII text, generally very short, and has to use far less bandwidth than a phone conversaion. Yet, a phone conversation to the very same person you're text messaging with would be a lot cheaper (or pretty much free)... I think they're charging crazy fees simply because it's a new fad, and they can...
You're either can't be serious, or you're just too old to adopt new technologies.
Speaking for myself, I have no use for text messaging. My calling plan doesn't include it, so when I receive one (apparently I can't turn that off, and it costs me 10 cents), I usually message back with "Just call me dammit", since it is then free (or very cheap) compared to silly text messages.
I do agree that the idea of a land-line only simply doesn't work for many people (myself included). Cell phones are just so damned handy for so many reasons; I got mine the day after my car broke down and I was forced to walk a good bit, but since then I've come to really love having it. And (since I can never use the ungodly minutes included) it costs less than my basic home-phone plan (which I'm considering getting rid of)...
Somehow, users seem to have the blame pinned on them instead of developers. This seems unique to the software industry--even the auto industry has more clue.
I agree to a point. If a flaw in an automobile caused some sort of problem the manufacturer is certainly responsible regardless of what the driver may have done to trigger the problem.
However, there's also more responsibility required of the user (driver). I don't think computing should require a license or anything like that, but you don't drive a car unless you know what you're doing. There are many basic rules you have to learn.
PC users don't want to learn things like software updating, firewalls, what file types not to accept via email, etc. Just like many drivers choose not to learn (or blatantly ignore?) things like "keep right except to pass", or that the center turn lane is NOT a merge lane...
Really you should put Windows systems behind an external firewall (that is, a firewall that doesn't run on Windows), such as a hardware firewall or a
*nix box of some kind.
Absolutely agreed. I am glad that many cable/DSL modems have built-in firewalling (or at least NAT). I'm personally behind a Linux box with iptables, and the only time I worry is when I bring in a friend's PC to my local network to fix it (almost always an infected machine) (and I disconnect my Windows box from the network before connecting the infected box).
What's hilarious is when a company is "protecting" a Windows network with a firewall running on -- you guessed it -- Windows. A friend's workplace was hit hard with one of the recent worms because the firewall machine became infected, and proceeded to infect the entire network. That's some great protection...
In SP2 MS claims it will be on by default. That will of course cause many problems for people, but it will hopefully stop the rapid spread of some worms too.
Yes, it may cause problems for *some* users, but the vast majority of XP users have absolutely no need for any services to be listening on a public interface. Those who do should easily know how to disable the firewall or allow the specific port(s) they need.
It should have been on by default all along. Think of all the worms that would never have happened if, say, Windows 2000 and XP had a simple firewall on by default.
The negative side (users complaining because they didn't know to turn off the firewall when it's in their way) would have caused far far less PR problems for Microsoft than that caused by the many remote exploits in the last few years.
There's been no shortage of buffer-overflow style attacks against Linux. The difference is a) there's far fewer machines out there to target and b) the users of those machines are far more likely to either have taken preventative measures or know how to identify and fix exploited machines.
Also don't forget c) there is a lot more diversity in Linux machines. You can't count on the fact that a significant percentage of Linux machines are running a particular service or program (or a particular vulnerable version thereof).
In Windows you can't be running any other version of LSASS.EXE than the one provided; every XP box in the world will have this vulnerability until patched.
My advice would be not to bother if you live in an area where Comcast and SBC "Yahoo" ... are your only choices.
Off topic, but when it comes to Cable/DSL providers, I've always read mixed results from every provider. I suspect it comes down to who, specifically, is running the local equipment...
I'm with Comcast currently and the service has yet to have an issue (about 6 months now). I had Sprint/Earthlink DSL back in Florida, Road Runner cable (Time Warner) also in FL, and Bell South DSL in Georgia. Road Runner gave the most problems, almost always due to "upgrades", but even then it wasn't frequent enough to complain (with a "residential" account).
Note, I do run my own local DNS cache as I found out (early on) that in most cases when service was down it was simply DNS not working (in many cases the physical connection is provided by one company (eg, Sprint) and DNS/email/etc by another (Earthlink in the Sprint case)).
The one thing that bothers me about VOIP is that (so far, in my experience) a cable connection requires power. My DSL always worked w/o power (modem, switch, router etc on UPS) where a cable connection simply didn't work without power. But, with my cell phone being cheaper all around, I am tempted to get rid of the POTS line. Cheap as the cell service is, I've not yet seen a need to go VOIP personally.
Windows is a stupid choice for a headless compute node...
...), I can think of no reason to have an OS that requires a video card (and drivers), and prompts the first time you boot without a pointing device connected, on a system that requires no interface or direct interaction.
...just as Linux is a stupid choice for a home desktop.
On that I will agree; unless there is a very specific reason to use Windows for a cluster (or server or
On this I don't agree. For you perhaps. For me even, in most cases: I run Linux (and FreeBSD) on my servers, and Windows 2000 and XP on my desktops (laptop is dual-boot XP/Fedora). However, there are plenty of good reasons to go with Linux (or BSD) on a desktop system.
I would agree that as a pre-install, or on a desktop for a user who doesn't know Linux (and will be angry that they can't run the latest Windows-based spyware-riddled game) it's not a great choice. But I wouldn't just generalize that "Linux is a stupid choice", because there are times where Linux is a good choice.
I've set more than a couple of desktop users up with Linux -- specifically, unsophisticated people who only need to check email and browse the web, and are using older machines (that I donated in most cases). And in all of these cases, the user really didn't notice any difference (and they constantly ask if the latest virus they heard about on the news affects them).
Not true. You assume people are intelligent, able to install things, and interested in doing so in the first place.
I agree with all but the intelligent part. It's not that people aren't intelligent, it's more that they have little or no interest in learning that much about computers. For those of us where computers are a large part of our lives and/or careers, these things seem obvious, just like a doctor might think some complex medical procedure is obvious...
It's rather improbable that the average person has the skills or the motivation to hunt for the correct tool.
That's a much better spin on it that I can agree with fully. Just remember, skills != intelligence. Nor does lack of motivation imply lack of intelligence; lack of interest perhaps...
Add to that the fact that most people are quite lazy. If the page looks weird they just forget about browsing the site and go somewhere else.
I do that myself, but in my case most times the page was designed for IE exclusively or requires flash, whereas I use Mozilla w/o the flash plugin. These days I find much less rendering problems, either because sites are more compliant (doubtful) or because Mozilla and IE interpret buggy/incomplete code in similar ways...
If Microsoft actually continued to add standard compliant technology to their browser we'd have a situation where we could actually adopt new technologies at a sensible rate. As it is now this is hardly the case.
Agreed fully. IE has, in my opinion, stagnated since version 5 or so, and no major feature enhancements (that I would use anyway) since 4.0. I forget which version implemented Microsoft's idea of CSS2 support, but in any case it's still not complete. Plus it lacks so many useful features that most other browsers have (disallowing unrequested popups, tabbed interface, etc)...
As for HTML, JPEGs and PDF's being all we need... you really haven't done much in terms of real world web design have you? Sure, you'll get the information across, but that just won't cut it in the current market (this claim is naturally not without exceptions.)
Standard and simple technologies are sufficient for many purposes, but in general I agree that we shouldn't be stuck with 1998 technologies. I like knowing that I can use certain advanced features, but in more than one case I've had to pull a cool CSS trick out of a page after finding out that IE doesn't support it (or doesn't support it properly). Often the work-around is to re-implement it using JavaScript, which isn't worth the effort IMO.
Just look at the new features we've gained over the last five years (that are actually being used). Then look at the five years prior to that (1994 to 1999) and you'll notice that things have stagnated around the time IE became stagnant. Granted we're still seeing many technologies becoming more widely used (eg, CSS), but only because the market generally has to catch up. By the time we get anything new (at the mercy of Microsoft most likely), we'll see actual development stagnate for a while, while we wait for the new tech to mature, and for there to be enough users with a newer browser...
At least that's my opinion, based on observation... however I do feel that HTML/CSS is currently pretty darned flexible if you learn the right tricks; it could be worse (it's not like we're stuck with plain ASCII text).
In summary, I fully agree with you and wanted to comment on a few points, and to point out that I don't feel that it's lack of intelligence, but rather specific knowledge that many users don't care to know (nor should they have to, ideally). Computers are tools to most people...
I do believe you used intelligence unintentionally to mean knowledge, as the other points in your post seem to agree.
Oh, and I hate PDF. More specifically, the idea (portability) is great, they print nice, but I cannot stand the Acrobat viewer (I wonder if there's another Windows viewer out there somewhere...)
(I really should get to bed soon; I noticed I'm rambling quite a bit here...)
I tested and can only send up to 5 MB. I tried a 9 and then a 7 MB attachment, but each time it told me I had exceeded the 10 MB limit - which obviously wasn't the case...
Email attachments are encoded (usually base64), which adds quite a bit of overhead. Someone else said it was around 30%, but I believe it may be more than that for binary attachments...
So your 7MB message, when encoded, likely does in fact exceed 10MB...
Isn't redirecting information to /dev/null technically destroying it?
/dev/null is therefore destruction of data you own.
/dev/null, and when asked for some specific portion of data you mention your policy, you shouldn't have anything to worry about. It's only when you normally *do* keep data, but the requested data is suddenly "not available", that you begin to look suspicious.
I think the point was that there's a difference between destroying data before it's requested, and destroying it when/after it's requested.
Document/data retention policies are written for just this purpose. If the FBI (or whoever) asks for data, and you (as a business policy) don't normally keep that data (whether at all, or beyond some set period of time), then you are generally okay. If OTOH the data is destroyed (or comes up "missing") when it's requested, you may be in for some legal trouble.
There is no law that I'm aware of requiring one to keep data/logs, except in specific cases. However, it should be provable that the data is missing in accordance to an already established policy if you want anyone to believe that the data wasn't destroyed simply because it was requested...
If an application generates some logging info, then you have it in your possesion, sending it to
And the point is to be consistant. If you always direct your log information to
Granted it's somewhat suspicious in itself when one doesn't keep any logs that are normally kept (for statistics/accounting purposes, etc), but that alone (if consistant) isn't grounds for accusation AFAIK...
Either way by making yourself insecure like this you are just adding to the many problems of the Internet.
Nail on head. Please refer to the SMTP protocol and open relays as a prime example.
I think if one is running a network of any kind -- free or otherwise -- one should be responsible for it. If you're not prepared to be responsible, don't offer the service.
I don't think it's very responsible to (as a bad analogy, as most are) leave a loaded weapon out in the open in the hopes that if someone used it for an illegal purpose that you'd be immune to any consequences.
It's one thing to have an open network due to ignorance, and another to just say "hey, if someone does something bad with it, I'm not responsible". Knowingly allowing people to use your resources for "bad" things (illegal or otherwise) is asking for trouble IMO.
Whether it's illegal or not isn't the issue here. I'm not on any kind of moral high-horse, but I don't see this as any differnt from running an open SMTP relay, or leaving a gun out in the open... in either case -- whether by ignorance or using it as an excuse to escape liability -- it's wrong. In my opinion anyway...
Personally, popup ads will tend to mean I'll never buy something from the company, both on principle (stop shitting over my desktop popup advertisers!) and because I worry about the potential ethics of said company (if they are prepared to kick my door in and force me to watch ads, will they run off with my wallet also?). However from time to time, I'll click out of curiosity on a tasteful non intrusive banner, specifically the google style text ones that arent in the face and are informative.
I couldn't agree more with this. I haven't seen a popup ad or animated Flash ad in quite a while (switched to Mozilla since 0.95, Flash/Java disabled) but one company I refuse to do business with is X10. Sure their little cameras look cool, but before I switched from IE, their ads were (at the time) so freakin' annoying and pervasive that I just can't see giving them any sort of money.
Google has it right. Their ads are text based, and they are picky about what kind of ads they will even show. I've been on both sides, and as an advertiser I once received a notice that the word "FREE" with all caps (as in, "FREE download") was potentially problematic. If they're that picky about that, I'm sure the ads I see go through the same type of checking before they're shown to me. In fact I pay more attention to Google ads (always clearly marked as such) far more than any other type of ad, based solely on their reputation with me personally.
Targetting people who you know don't want your ads (getting around popup blockers or spam filters, or ignoring the Do-Not-Call lists, etc) is simply not a good marketing move IMO. You can't sell to people who have (in some way) stated that they have no interest in your ads...
However, I have no problem with "normal" ads. Ads that aren't annoying (many TV/radio commercials are as bad as popup windows/Flash ads), that simply state what you're offering and possibly why I might want/need your service, are the types of ads I would respond to. Google's ads are almost always in this category, as are some (but not many) radio/TV ads, and even some normal static/animated banner ads on the 'Net. Bogus error messages or bogus contests ("You are visitor xxx..." or "Punch the Monkey") are even worse IMO...
The point is, if you have to try that hard to get me to notice your product, your product likely isn't good enough on its own to warrant a look. Other times (X10) the product itself may be worth the monetary cost, but I refuse to do business with a company who thinks this type of behavior is acceptable. Supporting such a company is (directly or indirectly) supporting these annoying advertising techniques, thus making them appear effective, and contributing to the problem.
Hell, I forgot the topic at hand here, but once I get started on advertising ('Net and elsewhere) I tend to rant...
GOTO's make spaghetti code. It is very hard to trace through, especially if the code is uncommented. If you end up with a bug, you will have a very hard time trying to trace through your GOTO's to find it.
I disagree. While GOTOs are often abused, they do have their place. I use them in C, infrequently (maybe 2 total in my main C project).
But the key is the uncommented bit: why does GOTO necessarily mean it's uncommented? The two GOTOs I use in my main project are well commented, and are only used in case of an audio card/driver failure. It's easy to follow the program flow, and obvious even if you don't read the comments (which in my project is often more than the code itself).
There's nothing inherent about the GOTO feature that is bad -- rather, it's an easily abused feature. That doesn't mean there aren't good uses for it. I can think of good uses for GOTO in BASIC, C, and other languages. I can also think of many abuses of other features of hundreds of languages...
I think that a company trying this would shoot itself in the foot. Every element would need to be mined from the US, all of the "Intellectual Property" would need to be US-based, and I mean every single little piece.
...but I don't believe that Toyota tried to imply that every component was USA-born either, they were mostly stressing the fact that they created a lot of jobs for US workers since at the time, many people wouldn't buy foreign vehicles for that very reason.
It wasn't all that long ago that Toyota had commercials showing the fact that most Toyotas sold in the US are actually built in the US. Of course much of the steel came from Japan, but many American car companies had by that point also started using Japanese steel; it was cheaper and higher quality in many cases...
Of course these days people don't seem to care about that, so the campaign hasn't been necessary. People buy cars made in Japan, Korea, and elsewhere if they find a good deal (and I think that's a good thing, really).
I don't like my browser or mail client doing things I'm not explicitly aware of. I cannot use Windows with a clear conscience because of IE's and Outlook's persistent security failures.
Just because one uses Windows does not mean they are limited to IE and OE. I personally run XP on my main desktop machine, but I use Mozilla. I've successfully converted more than a handfull of people to Mozilla or Firefox or Opera -- and all Windows users (of course there are the stubborn people like my dad who will inevitably end up using IE). There's nothing that says you ever have to use IE or Outlook Express just because they are preinstalled; on a new install, I fire up IE one time, to go to Mozilla.org. That, and Windows Update...
I don't so much find Windows to be inferior. It's just that Linux and the canon of open source software built upon it make so much more sense financially, socially, and from an engineering standpoint.
I agree to an extent on the engineering standpoint. On a server I'd never ever use Windows (is Hotmail still on BSD? Or Microsoft's own DNS servers?) but for me, Windows does have its place.
My stance is this: use the right tool for the job. Sure, I hope to be able to afford a Mac one day soon, but at the moment Windows runs my desktop (while Linux and/or FreeBSD runs my other four home systems and my servers). Windows is, for me personally, the best tool for the job on the desktop. On my servers, Linux or FreeBSD (depending on the task) are the best tool.
I have no interest in Linux over-taking the desktop. It does a great job for the things I use it for, but Windows does (and I'm sure the Mac does) a fine job on a desktop machine. Security problems, sure, but I patch my WinXP box and I patch my Linux boxes and both are, with proper patching/upgrading, pretty damned stable.
I guess the main problem I had with your post was the implication that one who uses Windows necessarily has to use IE or Outlook, which is simply not true. There are many alternatives to those particular applications, and (at least since 2000/XP) the OS itself is pretty solid...