I'm glad that part of my post irritated you. You see, we're supposed to be the Pros from Dover - that's why guys in our industry get the big bucks.
I've (unfortunately) seen Windows-based solutions pitched as (and made to appear as) vastly simpler; so much so that your average Windows administrator makes about 70% as much as a UNIX administrator (in this area of the Midwest, although my research leads me to believe that this is not a localized phenomenon).
Why is this? It is because a far larger percentage of PHB's can see how much easier and less demanding Windows is to administer. They therefore pay less; and become angry when they discover that their Windows sysadmins aren't the nearly omniscient Gurus they've come to expect from IT. Then, the pay scale drops even further. Now, guys like me work in UNIX shops and other PHB's want to know why they should pay for what I can bring to the table when there are plenty of other sysadmins who will work for less. Never mind the old bit "you gets what you pays for", the bottom line is that I'm still expected to be able to configure and maintain a more complex (more flexible, IMHO more stable) system while accepting the lower pay due a Windows administrator (who's only recourse when things break is often to get Microsoft support to acknowledge that something's broke - at least, that gets him off the hook).
Bottom line - Windows administrators get less respect because they do less. UNIX administrators have to do considerably more studying, understand the "under the hood" aspects of their OS in greater depth than their Windows counterparts, get more involved in the design, configuration, testing and implementation of software on their systems; yet we're collectively lumped into the "Computer System Administrator" bin as often as not, with disastrous results.
We see the Windows administrators going home earlier, because most of the larger shops still acknowledge the CW that Windows is for clients, UNIX is for servers; and the servers are considerably more important (machine for machine) than the clients. We watch these guys get the same pay and perks as we do, but they have an easier job not because their job is easier, but because the nuts 'n' bolts stuff is running under *NIX.
Yes, I know that there are numerous exceptions, enough to make my stated case not an overwhelming truth but rather a simple majority of cases. Part of the "dumbing down" of IT. Incidentally, most modern *NIX operating systems can be patched instead of upgraded; you ain't gettin' your old WinNT box up to XP functionality without reinstalling. Am I disdainful of your position because you want to get the same pay and title as I but work less? Yes I am! Let's put all of the background services (DNS, NTP, NFS/file shares, print services, etc.) on Windows instead of *NIX and compare 1)Administrator effort and 2)Downtime. There's a reason the majority of businesses stick with *NIX for core infrastructure - you gets what you pays for.
Read more carefully. The ships entire command and control structure were relegated to Windows NT at that time. After testing, the Navy fell back to existing solutions; whether this remains the case I don't know, but I do remember the test I described as being the first test of an "all Windows NT" solution. The article I cited was not specific to the test I remember (it was a long time ago, after all), but was intended to refute the AC's assertion that NT wasn't in use.
Besides, VOIP is not that radically different from the way that modern digital telephone networks operate anyhow; the underlying network layer is different, but conceptually the general idea is pretty similar. Thus, digital encryption of voice telephony is already possible (and available). If (when) VOIP matures, all the same technologies will be possible; and many which are currently technically infeasible will become common, even trivial. The biggest impact VOIP will have is that there will no longer be a unique wire associated with your telephone's network address (erm, telephone number); rather, it'll be a MAC address (I don't recall if VOIP relies on TCP or UDP primarily, but MAC or IPv6 will almost certainly end up as the layer of choice, I think).
Your comments have improved my understanding of the concepts in question. Also, the way you presented them (non-argumentative) was a true departure from the/. norm - most refreshing and welcome!
Although I still find it disturbing that all cosmologists and theoretical physicists agree that (with regard to understanding) once you pass the event horizon, all bets are off.
For like eight years I've been Administrator to my company's server. It's a small company with a few programmers... I work as a programmer full time but I also install and manage all the software in the server which hosts POP3 and SMPT server, a Samba file server, DNS, Apache, NFS and some web apps like Mediawiki, Mantis, dotProject and myPHPAdmin. Through this years we've had to reinstall the server a number of times... a couple of times it was to do a major kernel upgrade because we've decided to upgrade the motherboard and CPU, I think another time it was because of a hard disk failure and the last one was because upon enabling DMA for the HDD it got SERIOUSLY corrupted (a problem it seems many people with certain VIA chipsets had).
For over ten years (not like ten years, but over ten years) I've been a System Administrator for various companies, both big and small. I've worked with POP3/SMTP, Samba, DNS, NFS, NIS/NIS+, et. al. Through the years, I've had to upgrade or perform initial installations on a number of servers, for a wide variety of reasons. Only once was I obliged to upgrade a *NIX OS because it had been EOL'ed, and that was Solaris2.6 -> Solaris 8 (but that wasn't open source at the time, still isn't really open source).
Our first Linux was Slackware I think... we decided to switch to Red Hat, but since I live in Argentina and this was before the days of Broadband we couldn't get a Red Hat CD... so we bought a Mandrake distro CD instead. For the next upgrade I meant to download Mandrake, but though I don't remember exactly why (I think it was because they were limiting the support options for non-paying customers or something like that) I decided to switch over to Red Hat... when the time came to upgrade the Red Hat (we had bought an AMD Athlon and new motherboard), the project had split in Fedora and Red Hat... so I switched over to Debian. Since the joke of town is that Debian doesn't update their distribution enough, I guess I'd switch to Ubuntu if I had to reinstall the server tomorrow.
In this of this switches I got a lot of new stuff I didn't have before: easier configuration... new kernel... support for new hardware... new software. But each time I had to learn a whole lot of new things. Example: I started with a Sendmail configured by hand... then the m4 configurations became the norm but I didn't know what to do with those... and now I'm running with Exim4, since it's the standard in Debian and was way easier to configure to my liking. The same with printing: I had lpq, but then I got CUPS with the Red-Hat and it was the standard for the distribution, thus easier to configure, had more support, was nicer... and now I don't even remember what I run my printers with (I think I still have CUPS... but perhaps I was offered a newer one at some point...). Besides changing programs each distribution came with different paths, different startup scripts, different location for the logs, different update mechanisms, different security measures and different best practices.
I also got my first taste of Linux via Slackware. I also have RedHat experience, because for some time RedHat was the darling of businesses implementing Linux. I also used Mandrake (a RedHat derivitave) just to see what they meant by "more user friendly". Yes, I also remember having to relearn things when Sendmail went to using the M4 preprocessor. Yes, I remember all the old 'lp' commands which CUPS came along, forcing another relearn. Then again, we all know that WindowsXP is virtually identical to Windows 3.1, right? Oh, except . . . for virtually everything. If I really have the urge, I can still locate older versions of most distributions through what would be considered proper channels. Uh, where (except for a garage sale or e-Bay) am I going to find Windows 3.1? Windows 3.11? Windows 95? Windows NT? Boy, I hope everybody likes Vista, 'cuz XP's days are numbered a
(Reference to Babylon 5, when Bester explains to Mr. Garabaldi that he can't act against the Psi Corps because he's been "Asimoved" - instilled with a rule which won't permit him to directly act against Psi Corps or any of its personnel)
I can see it now - "Firing solution discarded: potential damage to Microsoft-owned assets. Fire mission against enemy aborted (OKAY) (CANCEL)"
Anybody remember the first tests of the Aegis-class cruiser? The first US Navy vessel controlled exclusively by a Microsoft product (Windows NT)? That's the ship which identified Catalina Island as a "fast moving target" and proceeded to lock her guns on the vessel assigned to monitor operations (the USS Forrestal, if memory serves). They had to shut down the entire ship and have her towed back to port. Yeesh! Daystrom's M5 duotronic unit, all over again (at least this one had an "off" switch)!
And what will TCM do if a unit suffers battlefield damage and a solution improvised to bring it back up? For battlefield conditions, a ruggedized, highly-available system is required. How do we reconcile that with TCM, a single point-of-failure if ever I saw one!
Out of the box, XP doesn't let limited users burn CD's/DVD's - I never gave fixing this particular failing any thought, having convinced myself that my SO is pretty savvy regarding computers (well, savvy as users go anyway).
I was once taught that the more powerful a gravity field, the slower time progresses within that field - a consequence of Einstein's General Theory of Relativity. If that is true, then during the collapse of a star towards becoming a singularity, we would expect the passage of time to be slowed more and more severely as the mass collapses and produces a denser and denser gravity well. Wouldn't this result in a real-universe equivalent of Zeno's paradox - i.e., wouldn't the collapsing mass always be moving closer to becoming a black hole but never arriving at that point due to the increasing time dilation?
I've been wondering about that for some number of years now - but the fact that cosmologists have generally accepted the existence of singularities as all but proven fact and have even had many observations which supported this belief has always prevented me from thinking too hard about it (after all, why pit my amateur understanding of cosmology and relativity against that of experts?).
If this assertion proves to be tenable, what effect will it have on collateral theories in cosmology (for example, estimates on the total mass of the Universe, which in turn affects our understanding of whether or not space is curved and if so, positively or negatively)? Much of our current understanding of the cosmos is based directly on the correctness of the Theory of Relativity, but this finding (if confirmed) would appear to falsify at least some of relativity's conclusions. Does this tenative finding square with string theory? How much of Einstein's Theory of General Relativity will need to be updated to accomodate these findings?
Then again, the New Scientist isn't exactly the most unimpeachable scientific news source. Perhaps I won't trash Al's most famous (and best supported) theory just yet.
After all, under Windows my SO has the (now demonstrated) ability to bring my network to its knees. To give her a limited account would be to render the machine almost unusable to her (as there are too many things which she does which require Administrator level access).
Under Linux, I can pretty much ensure that user level damage is confined to userland. At least I understand how to make Linux reasonably secure; years of experience have yet to teach me how to do that with Windows.
On a machine which I maintain for my SO and children, M$ XP Pro is installed. The default browser is FireFox, which I have managed to convince my SO and children to use.
My daughter (with a limited user account, no less) viewed a malicious advertising banner while logged into MySpace.com. I'm quite sure she clicked "yes" to running a WMF exploit.
She has a limited account. End of story, you say? Nope, read on . ..
My wife logged in a couple days later. A popup baloon warned her that the machine was infested and she should "click here to fix the problem". Well, she installed AntiVirusGolden v3.3 (from her not-so-limited user account). Who can blame her? I wouldn't have fallen for it (I already had CA's EZ-Antivirus installed and more or less trusted it), but it looked like a valid course of action to her, so the next thing I knew there were nearly a dozen payloads whanging around the rusty innards of my SO's computer - some acquired on the spot, others dropped there during the following week, I'm sure.
That machine now runs Linux (like the rest of my home network). I'd like to thank the wonderful malware authors at AntivirusGolden for giving me the leverage I needed to convince my SO to give up on Windows and use a somewhat more securable OS.
Oh, but I'll continue to use Firefox, now that I've closed that horrible WMF exploit that it has! You'd think the Firefox development team would know better than to trust end-users with the option to execute WMF's. Hmmph!
*(The above is intentionally sardonic; but the basic facts are true)*
Size of an INT, Unsigned INT, DOUBLE and Unsigned DOUBLE are all compiler dependant (to a lesser extent, hardware/architecture dependant). 8-bit INT is merely the most common situation, especially with x86 hardware being so prevalent.
Test it - write a short C program, build on VAX, UNIX, CRAY and Intel hardware. Use UNIX cc, Borland C++, etc. Sooner or later, you'll get a surprise. Don't believe me? Keep writing code which assumes these limits, you're in for an unpleasant surprise.
My Linux box has an ATI TVWonder Pro card in it - I leave the desktop resolution at 800x600 and do all my computing from my couch. Granted, it took me a while to get used to the lack of screen real estate, but once I did it was great! Instead of having to shell out for a remote control for my TV card, I just use a wireless KB/Mouse combo (Microsoft's bottom-end -1000 series desktop - works great in Linux, okay in Windows).
but it needs to be done in a way that won't 1) swamp the internet backbone with huge quantities of digitized voice telephony, and 2) give ISP's a good excuse to insist on multi-tiering the internet.
That said, can video telephony and the kind of communication we've seen portrayed on Star Trek et. al. be far behind?
Rootkits, draconian DRM built into blu-ray, a track record of considering all their customers to be pirates and thieves . ..
And let's not forget that while Sony says they have no intention of using this technology, their track record suggests that this is not a trustworthy statement. They didn't invent and patent this technology just to add to their patent portfolio.
Come to think of it, I've written similar routines for code I've used and distributed. Would that constitute "prior art"?
Slashdot Mirror
I've (unfortunately) seen Windows-based solutions pitched as (and made to appear as) vastly simpler; so much so that your average Windows administrator makes about 70% as much as a UNIX administrator (in this area of the Midwest, although my research leads me to believe that this is not a localized phenomenon).
Why is this? It is because a far larger percentage of PHB's can see how much easier and less demanding Windows is to administer. They therefore pay less; and become angry when they discover that their Windows sysadmins aren't the nearly omniscient Gurus they've come to expect from IT. Then, the pay scale drops even further. Now, guys like me work in UNIX shops and other PHB's want to know why they should pay for what I can bring to the table when there are plenty of other sysadmins who will work for less. Never mind the old bit "you gets what you pays for", the bottom line is that I'm still expected to be able to configure and maintain a more complex (more flexible, IMHO more stable) system while accepting the lower pay due a Windows administrator (who's only recourse when things break is often to get Microsoft support to acknowledge that something's broke - at least, that gets him off the hook).
Bottom line - Windows administrators get less respect because they do less. UNIX administrators have to do considerably more studying, understand the "under the hood" aspects of their OS in greater depth than their Windows counterparts, get more involved in the design, configuration, testing and implementation of software on their systems; yet we're collectively lumped into the "Computer System Administrator" bin as often as not, with disastrous results.
We see the Windows administrators going home earlier, because most of the larger shops still acknowledge the CW that Windows is for clients, UNIX is for servers; and the servers are considerably more important (machine for machine) than the clients. We watch these guys get the same pay and perks as we do, but they have an easier job not because their job is easier, but because the nuts 'n' bolts stuff is running under *NIX. Yes, I know that there are numerous exceptions, enough to make my stated case not an overwhelming truth but rather a simple majority of cases. Part of the "dumbing down" of IT. Incidentally, most modern *NIX operating systems can be patched instead of upgraded; you ain't gettin' your old WinNT box up to XP functionality without reinstalling. Am I disdainful of your position because you want to get the same pay and title as I but work less? Yes I am! Let's put all of the background services (DNS, NTP, NFS/file shares, print services, etc.) on Windows instead of *NIX and compare 1)Administrator effort and 2)Downtime. There's a reason the majority of businesses stick with *NIX for core infrastructure - you gets what you pays for.
Read more carefully. The ships entire command and control structure were relegated to Windows NT at that time. After testing, the Navy fell back to existing solutions; whether this remains the case I don't know, but I do remember the test I described as being the first test of an "all Windows NT" solution. The article I cited was not specific to the test I remember (it was a long time ago, after all), but was intended to refute the AC's assertion that NT wasn't in use.
Besides, VOIP is not that radically different from the way that modern digital telephone networks operate anyhow; the underlying network layer is different, but conceptually the general idea is pretty similar. Thus, digital encryption of voice telephony is already possible (and available). If (when) VOIP matures, all the same technologies will be possible; and many which are currently technically infeasible will become common, even trivial. The biggest impact VOIP will have is that there will no longer be a unique wire associated with your telephone's network address (erm, telephone number); rather, it'll be a MAC address (I don't recall if VOIP relies on TCP or UDP primarily, but MAC or IPv6 will almost certainly end up as the layer of choice, I think).
Although I still find it disturbing that all cosmologists and theoretical physicists agree that (with regard to understanding) once you pass the event horizon, all bets are off.
Stick to what you know - you'll look a whole lot smarter!
For over ten years (not like ten years, but over ten years) I've been a System Administrator for various companies, both big and small. I've worked with POP3/SMTP, Samba, DNS, NFS, NIS/NIS+, et. al. Through the years, I've had to upgrade or perform initial installations on a number of servers, for a wide variety of reasons. Only once was I obliged to upgrade a *NIX OS because it had been EOL'ed, and that was Solaris2.6 -> Solaris 8 (but that wasn't open source at the time, still isn't really open source).
I also got my first taste of Linux via Slackware. I also have RedHat experience, because for some time RedHat was the darling of businesses implementing Linux. I also used Mandrake (a RedHat derivitave) just to see what they meant by "more user friendly". Yes, I also remember having to relearn things when Sendmail went to using the M4 preprocessor. Yes, I remember all the old 'lp' commands which CUPS came along, forcing another relearn. Then again, we all know that WindowsXP is virtually identical to Windows 3.1, right? Oh, except . . . for virtually everything. If I really have the urge, I can still locate older versions of most distributions through what would be considered proper channels. Uh, where (except for a garage sale or e-Bay) am I going to find Windows 3.1? Windows 3.11? Windows 95? Windows NT? Boy, I hope everybody likes Vista, 'cuz XP's days are numbered a
I can see it now - "Firing solution discarded: potential damage to Microsoft-owned assets. Fire mission against enemy aborted (OKAY) (CANCEL)"
Anybody remember the first tests of the Aegis-class cruiser? The first US Navy vessel controlled exclusively by a Microsoft product (Windows NT)? That's the ship which identified Catalina Island as a "fast moving target" and proceeded to lock her guns on the vessel assigned to monitor operations (the USS Forrestal, if memory serves). They had to shut down the entire ship and have her towed back to port. Yeesh! Daystrom's M5 duotronic unit, all over again (at least this one had an "off" switch)!
And what will TCM do if a unit suffers battlefield damage and a solution improvised to bring it back up? For battlefield conditions, a ruggedized, highly-available system is required. How do we reconcile that with TCM, a single point-of-failure if ever I saw one!
Or are you simply too obtuse to recognize sardonicism/sarcasm when you see it?
Out of the box, XP doesn't let limited users burn CD's/DVD's - I never gave fixing this particular failing any thought, having convinced myself that my SO is pretty savvy regarding computers (well, savvy as users go anyway).
I've been wondering about that for some number of years now - but the fact that cosmologists have generally accepted the existence of singularities as all but proven fact and have even had many observations which supported this belief has always prevented me from thinking too hard about it (after all, why pit my amateur understanding of cosmology and relativity against that of experts?).
If this assertion proves to be tenable, what effect will it have on collateral theories in cosmology (for example, estimates on the total mass of the Universe, which in turn affects our understanding of whether or not space is curved and if so, positively or negatively)? Much of our current understanding of the cosmos is based directly on the correctness of the Theory of Relativity, but this finding (if confirmed) would appear to falsify at least some of relativity's conclusions. Does this tenative finding square with string theory? How much of Einstein's Theory of General Relativity will need to be updated to accomodate these findings?
Then again, the New Scientist isn't exactly the most unimpeachable scientific news source. Perhaps I won't trash Al's most famous (and best supported) theory just yet.
'Nuff said?
Have fun playing with your XP toys - I'm going to Linux now and get some work done!
Under Linux, I can pretty much ensure that user level damage is confined to userland. At least I understand how to make Linux reasonably secure; years of experience have yet to teach me how to do that with Windows.
Read my previous post again; this time, assume I'm sneering when I speak. It'll make more sense.
My daughter (with a limited user account, no less) viewed a malicious advertising banner while logged into MySpace.com. I'm quite sure she clicked "yes" to running a WMF exploit.
She has a limited account. End of story, you say? Nope, read on . . .
My wife logged in a couple days later. A popup baloon warned her that the machine was infested and she should "click here to fix the problem". Well, she installed AntiVirusGolden v3.3 (from her not-so-limited user account). Who can blame her? I wouldn't have fallen for it (I already had CA's EZ-Antivirus installed and more or less trusted it), but it looked like a valid course of action to her, so the next thing I knew there were nearly a dozen payloads whanging around the rusty innards of my SO's computer - some acquired on the spot, others dropped there during the following week, I'm sure.
That machine now runs Linux (like the rest of my home network). I'd like to thank the wonderful malware authors at AntivirusGolden for giving me the leverage I needed to convince my SO to give up on Windows and use a somewhat more securable OS.
Oh, but I'll continue to use Firefox, now that I've closed that horrible WMF exploit that it has! You'd think the Firefox development team would know better than to trust end-users with the option to execute WMF's. Hmmph!
*(The above is intentionally sardonic; but the basic facts are true)*
All the neat toys they let me play with.
Consider it just another SNAFU.
Test it - write a short C program, build on VAX, UNIX, CRAY and Intel hardware. Use UNIX cc, Borland C++, etc. Sooner or later, you'll get a surprise. Don't believe me? Keep writing code which assumes these limits, you're in for an unpleasant surprise.
Deja foobar - the feeling of having made the same mistake before.
My Linux box has an ATI TVWonder Pro card in it - I leave the desktop resolution at 800x600 and do all my computing from my couch. Granted, it took me a while to get used to the lack of screen real estate, but once I did it was great! Instead of having to shell out for a remote control for my TV card, I just use a wireless KB/Mouse combo (Microsoft's bottom-end -1000 series desktop - works great in Linux, okay in Windows).
Give 'em hell, Mr. Hogan!
That said, can video telephony and the kind of communication we've seen portrayed on Star Trek et. al. be far behind?
Lucky bastard. My ISP didn't grease me at all, and they didn't even kiss me when they were through.
Windows NT was just Windows 95 with networking slapped on.
Or else, why do you think NT shot up to SP4 so fast? SP5 and SP6 took longer to fix routine bugs.
NT - Never Tested.
NT - No, Thanks.
NT - No throughput.
NT - Not Today!
NT - Now Thrashing.
And let's not forget that while Sony says they have no intention of using this technology, their track record suggests that this is not a trustworthy statement. They didn't invent and patent this technology just to add to their patent portfolio.
Come to think of it, I've written similar routines for code I've used and distributed. Would that constitute "prior art"?