Wait a few years? It's here - the AT&T/Verizon telco choice. If you get to choose which one serves you, you're one of the less than 1%ers. The rest of us don't have a choice.
Cell phones are also shrinking down, we now have AT&T(/Cingular), Verizon, T-Mobil and Sprint/Nextel.
Radio is down to 4 companies, effectively, and one of those, ClearChannel, also owns 33% of XM (a main reason I'm not an XM subscriber).
So next you're going to promote recording all meetings and archiving tapes?
Seriously, what ever gave anyone the idea that email "must" be saved? This move to require archiving IMs is equally intrusive and ridiculous, because next they'll require saving all voice conversations... after all, A/V IM is here already, and it's not too great a leap to go from archiving IM to archiving A/V IM to archiving phone calls....
I'd mod you up, just for disparaging Armageddon. That piece of tripe was terrible. I thought deep Deep Impact was much better, although too much mushy story in there, so still only a C+ at best.
OS/2 most certainly did suffer that problem and was quite infamous for it (30 seconds Googling for "Single Input Queue problem" should tell you that).
MS's OSes still suffer that problem, especially if you use some of the legacy APIs (ie, old apps).
No, they don't.
OK, I'll admit I was wrong on OS/2's single input queue. They fixed it in the versions I ran though with a work around. So it seemed like I didn't have that problem.
However, you're wrong as well. Windows NT certainly did have a single input queue, albeit an "OS only" queue, it still can lock up the entire system.
No, it didn't. OS/2 and Windows NT came from _completely_ different code bases. They shared an API - you could run text mode OS/2 apps in early releases of NT - but that's the closest they came to having "common components"[0].
Again incorrect - there was support for HPFS (which MS owned rights to), the OS/2 1.x API, and even OS/2's PM API.
You are thinking of NT's "API personalities", which were designed in from that start as modular chunks that could be activated and deactivated at will (win32 is implemented the same way). They were a result of its microkernel-ish architecture, not of any "OS/2 heritage".
The only "personalities" I recall being talked about at that time were IBM's OS personalities running on top of their microkernel architecture. IIRC, they demoed a single box running NT, OS/2 and RISC 6000 OS personalities on different monitors.
As for the "modular" APIs, I believe all but POSIX is completely gone now. If they were truly modular, why did no one other than MS ever release a module? (I can't think of one, Cygwin is an installed program after all)
You're correct in that OS/2 2.x was radically different from NT (VMS) in design. It was far superior, and by 2.3 had a GUI and performance I have yet to see matched by anyone, including Mac OSX.
It was most certainly not superior (and OS X is hardly a performance benchmark to strive for, it's probably the slowest mainstream OS around). OS/2 had a monolithic kernel, wasn't portable, was single user (so nothing even as basic as filesystem permissions, let alone the pervasive security model of NT), didn't support multiple processors (although that was hacked into some specialised builds), still had significant subsystems that were still 16 bit, even into the mid-90s, had poor memory management (no dynamically-sizing disk cache, for example), etc, etc.
It had a hybrid kernel, not quite monolithic, not quite micro. The HPFS filesystem was a stripped down system for a single box. It was the only common 16 bit piece still running in OS/2 2.3 onward. It could be replaced with HPFS386, which was 32 bit and solved most of your other observations (you'd need some of the additional server modules to run true multi-user on a box)
And exactly what "pervasive security model" permeates NT? The one that allows random code to run and modify files everywhere? That's some security.
And then let's get to where it was superior - the threading model. A truly pre-emptive threading model versus the time-slice. MS still has issues with the time slice, even though it's down to 16ms in XP.
Windows NT was superior to OS/2 in pretty much every conceivable way. Which it should have been, given that - as the original poster noted - it was built to replace OS/2.
Hence Microsoft stating that NT should be rebooted at least once a week (Exchange Servers) and once a month (others) vs OS/2 systems that just ran and ran... (running Sendmail, I might add)
It required less (OS/2 was tolerable on a machine with 6MB RAM, NT required at least 10MB) - but that's hardly surprising since it wasn't _doing_ anywhere near as much.
No wonder your experience was horrible. I wouldn't dream of running it on less than
According to the show - the critical age is somewhere between 3-4. After that, grammar/sentence structure is no longer possible, while word acquisition is. That appears to stop around 8-12, although the number of cases is (fortunately) exceedingly small and that data therefore highly unreliable.
You're not serious? Read some History. Before the IBM/Microsoft breakup over OS/2, what became Windows NT 3.1 was supposed to be OS/2 3.0.......
No released version of OS/2 bears any architectural resemblance to Windows NT. None. Five minutes listing the key features and looking at the block diagram of each should tell anyone with a passing acquaintance to OS design that.
In short, the poster I replied to is wrong - NT does not have the Single Input Queue problem that OS/2 did (and probably still does).
I think you have that wrong - OS/2 didn't suffer the problem. MS's OSes still suffer that problem, especially if you use some of the legacy APIs (ie, old apps).
On OS/2, you're also not 100% correct. OS/2 1.x, which were released and in widespread usage in the banking industry, especially in ATMs, actually did have some common components with NT 3.1. NT 3.5 separated this some, and it was left as a user installed support module in NT 4.0 before disappearing in Win 2K.
You're correct in that OS/2 2.x was radically different from NT (VMS) in design. It was far superior, and by 2.3 had a GUI and performance I have yet to see matched by anyone, including Mac OSX. (disclaimer, I never used BeOS, which was about the only OS at the time that was reported to exceed OS/2 in performance and some really liked their GUI). If you think OS/2 didn't perform, your hardware wasn't good enough, as it required the same hardware as NT 3.5 to run well.
Can't you just use Alt+Tab? Only to get to a Command prompt, provided you have one running. If you don't, you're screwed, even Task Manager won't come up. It's one of the "features" of the single GDI thread MS insists on sticking to.
The GGGP however, is incorrect about OS/2. OS/2 didn't suffer from this problem in the same way, since each application ran on its own thread(s). A properly designed PM app actually had separate threads for input and output, which helped even more. There was no single system wide thread for input/output. Even the Win32s apps weren't locked in the same way they were on MS OSes, hence the "Better windows than Windows" statement.
I'd personally love to see IBM offer the PM and filesystem components of OS/2 running on a Linux kernel. The PM interface would solve one of the major issues with KDE/Gnome, and their HPFS386 file system (since 2MB cache in HPFS is just too small these days) is an incredible performer for workstations. The licensing/patent deal on that should have expired as well, so there's no more $80 per copy payment to MS. (If you ever wondered why OS/2 wasn't sold for less to compete with MS, there's one major part of it)
There's a show on Discovery or TLC about feral children, young children that were severely neglected. The show focuses on the language and brain development with age, as several of these children had human contact as they grew from infant to 6 or more years of age. (The oldest was around 12 when she was found).
Despite the show being deeply disturbing, it is also fascinating from a scientific viewpoint regarding social development, language development, and brain function/development.
As it relates to the P/GP, humans apparently learn both to work together and be competitive from their environment. It is based on feedback loops. There was one boy, found at age 3 or 4 in the Ukraine, that lived with a pack of dogs. He'd become adopted because he offered food to a stray, and over time the pack grew, or so the theory goes. He developed the behavioral aspects of a dog for both the competitive and cooperative aspects.
Think about this - their claim to woe and "damage" is that revenue is dropping.
First, recall Revenue != Profits. Most companies that go under do so not for lack of revenue but lack of profits.
My point was that their costs have dropped significantly, they've only experienced a 4% drop in revenue, yet the revenue drop is what they're citing as proof that online sales/piracy is hurting them deeply. Revenue is irrelevant when the model changes. You could have a 50% drop in revenue yet still increase profit if your item costs went from 70% of the revenue stream down to It should also be noted that previous numbers were most likely inflated due to the fact that people had to buy 10-20 songs in a package when they only wanted 1 or 2.
No, this doesn't mean revenues were inflated, it measn they used to pay $9-$18 per song, now they pay $1 per song (and still you bitch, moan, and whine?). If they make up for this by buying 9x to 18x more songs, revenues stay the same. Since most costs associated with the industry aren't variable w/ the number of units sold, only then do profits stay the same.
It certainly does mean revenue numbers were inflated. When you package things so that you have to buy 9X what you need your revenue was just inflated nine fold. This appears when someone else comes along and sells just the amount you need, and takes business away from the 9X packaging. That's one thing that's happening in the music industry, and it's high time it happened. I remember the death of $0.50 - $0.99 45s when all of a sudden I had to buy $16+ CDs, or, amazingly, $7 LPs which used to be $12-16 prior to the introduction of CDs.
Here's another thing for you to ponder: CDs cost between $0.80 and $1.50 for the entire manufacture/packaging/shipping process and sell for an MSRP of around $15-18. LPs cost around $3 for the same, and run about $7-10. Care to explain?
You know - here's a thought - is the music industry really in any trouble whatsoever? Think about this - their claim to woe and "damage" is that revenue is dropping. Isn't that what's supposed to happen when your costs drop? Note with online transactions they're not incurring any manufacturing, packaging, shipping, nor physical theft/damage losses, all which raise the price of physical media vs online media, not to mention dead inventory.
It should also be noted that previous numbers were most likely inflated due to the fact that people had to buy 10-20 songs in a package when they only wanted 1 or 2.
So, the real question would be if their legitimate profits are sufferring. Not their net profit, as I'm sure the RIAA requires a pretty penny to leash and feed, but actual profit from their "product".
For all of MAD's suckyness (Microsoft Active Directory...MAD delivers functionality that OS X can't even dream of.
Example: You can send your admin-monkey to the server, with a few manual procedure steps,...from using a DOS command shell.
This configuration change will go out on the network with the next reboot. And poof! 500 nosy, troublesome Users are now a bit less able to shoot themselves in the foot, or work mischief on your systems. First, I love the MAD acronym.:)
Secondly, here's a question for you: does OSX even require this functionality, or is it merely a consequence of the MS world-view that this functionality seems to be required?
Let's look at your example, and let me admit I've not used OSX in an enterprise setting, but I have used Solaris, HPUX, Linux, and AIX in enterprise settings and all are *nix variants like OSX. First, you have to image all your drives - that's standard across all systems. Next, you have central servers with user profile information on them on one variant or another (again, standard in this scenario). With the *nix variants, the user home directories can be NFS mounted, with every machine giving you the same view instantly, with the same performance as you'd experience on any other machine. Unless ADS has changed, I believe a new profile is downloaded/updated on every login/logoff, and is slower than molasses if your system is configured with or default/user stores large files on the desktop or in the profile. Also, should I want to change run perms, I change it on the server(s) and voila - INSTANT changes in what users can do - no logoff/login cycle required. Now, OSX being a *nix variant, can probably be setup exactly the same way (The only uncertainty remains because I haven't done it nor experienced it first hand).
Other examples: disable the IE address bar. (and prevent Trojans from hooking it). Disable the Tools menu so users can't mung with the security settings in IE. Disable control panels. Enforce a password-protected screen saver across the enterprise. Take the File-Open menu away from MS Excel. Whatever. I assure you, as draconian and capricious as these sound - some of them are ABSOLUTELY NECESSARY to operate computers in a secure environment. These are all windows centric issues, however, given the above configuration with NFS mounts, applications can be controlled in exactly the same way. General users don't have access to system settings in network *nix environments without a super user password. Your strawmen don't exist in a standard *nix rollout.
...But the effort required to "roll your own" system to manage client configuration on this scale, with this ease of use, would be on a pretty much unimaginable level. I'm afraid you're mistaken on that. See below:
I am an unashamed Mac fanboy. The bane of my life is when I have to go into work, and fix broken Domain Policies or MAD server. I have 4 Macs at home, and I try to manage them somewhat like an enterprise - and I'm telling you - the tools just are not there. There *is* a usable infrastructure, but you'd need to pump tens of thousands of man-hours from a very skilled scripting guru to pull off the equivalent thing on a Mac. I think you'll find that with a half-way knowledgeable *nix system administrator that more than half your issues will go away, and a decent one will make 95% of your issues evaporate. You're still seeing the world through MS glasses. You expect systems to work like MS systems. MS systems are standalone systems that sort of communicate (badly) through networks, since networks were after thoughts. *nix systems were designed with networks (well, multiple terminals/multi-user systems which were much easier to convert to network centric systems) in mind and thus have a completely different view of networked systems.
Jokes aside, ADS did create a directory services structure for windows. It still sucks rocks, and is trying to still come up to NDS (which is pretty much a goner these days).
I still remember Exchange pre-ADS. It was fast. Try sending a 250K user list in Exchange w/ ADS.... you'll be waiting hours, and that's only if you're lucky and you've done some serious optimizations in your mailing list structures.
ADS sucks in just about every way possible, it's another standard MS clunky POS. BTW, Exchange sucks rocks too, there's just not anything better yet in the categories it addresses, but that doesn't mean Exchange is "good". Think of it as you've got a yugo, a fiat, and a hyundai and those are your only choices. They're all bad unreliable POSes, but the hyundai will be your best bet out of those three.
30 years ago, a simple +-*/ calculator was easily twice the size of today's standard calculators. While today's calculators could be made smaller, in general it's cheaper and UI design issues that stabilize the size. After all, using 50+ keys on a calculator requires a certain size in design, and ruggedness requires a certain thickness unless you start going with more exotic (expensive) materials. After all - would you (be able to) use a scientific calculator the size of a postage stamp?
Your sassy AC comments aside, there's a lot more to it than what I posted. But without that core tenet, I really do feel that you're only lipsticking the pig.
The entire point is to remove special interests, and allow anyone to effectively run. Yes, this also removes one of the prime components that keeps the 2 party system going - that of effectively raising campaign funds and promoting a single person.
I don't know why you think anyone gets a message out at all in the current setup. Take Al Gore or Bob Dole and contrast the "marketed packaged presidential candidate" with the post campaigning version. Both of their post campaign versions would have kicked butt compared to the stiff cardboard their campaigns made them into.
If you want to fix politics, eliminate campaign funding. Completely. No commercials. Only a slice of time on PBS (that's a purpose that PBS was originally designed for - a medium to reach everyone) with other stations also giving time as needed, such as for debates. (It's in their charters)
If you don't eliminate campaign finance, you continue to support the buying of politicians by business interests, and it shouldn't surprise you that those politicians place the interests of business ahead of their constituents.
OK - that's easy - run a supply chain planning full system, end to end. The model alone for some companies were over 20GB, while the super scaled down dev model usually came in around 1GB. And that's in-memory. Besides that, you'll generally be running at least 1 full fledged DB, several servers (appservers, webservers, various daemons, etc) and potentially an entire SOA on a single machine. Another one would be developing management and maintenance software that also integrates with existing systems such as Tivoli and HPOpenView, including some resources to monitor (usually a full instance of whatever you're trying to manage and monitor.
Ever want to see your memory usage skyrocket, try either of those scenarios. Is a laptop the best venue for development of that sort? No. However, when you need to be portable to show things to potential clients, it's pretty much the only way.
I should also mention that this type of code only runs on server OSes, because it requires over 2GB of RAM, which no client OS supports - that would be MS Windows, BTW, all other current OSes in this class support 4GB or more. MS OSes only supply a max of 3.5GB in user space unless you're running 64bit MS OSes. I'm not sure what 64bit XP is limited to. I'm sure MS is still smarting from the 20bit kernel -> 32bit counter for memory paging that guaranteed BSODs on NT 4.0 prior to SP1.
Samsung - 2 identical N400 phones - one has 3 bars, the other 5 when held next to each other. Battery on phone 1 lasts maybe 1.5 days. Batter on phone 2 lasts 7 days with some usage.
LG - my current fancy-shmancy phone - standby time of at least a week. Can be on calls for hours. I get three bars where the best Samsung got only 1 bar, and 5 bars where the Samsung got 2. The service provider, however, I think blows. Calls drop, connections are sporadic, especially within network. This would be one of the ones claiming fewest dropped calls and a proposer of the tiered internet legislation. They're already tiering their cell service. In-network (free) calls have worse sound quality and are prone to drops, horrible quality, or failure to connect while calling someone out of network (toll) are crystal clear and immediately connect and almost never drop.
No SIM card possible on either one I used to have.
Overall, LG, Samsung, and Nokia are all far superior to Motorolla. If it wasn't for the Razr family, I don't think anyone would willingly buy a Motorolla phone at if you'd seen the competition.
I strongly suspect you're right. It's going to be one of those things that people try out, and go "Hey, I really like this. It does x,y, and z, and I can dump 2 other devices and just carry one, and I get bonus feature v to boot!"
I know when I saw the presentation, I went "Now there's what I've been looking for in a phone. I personally hate cell phone interfaces. I'm sure I'm not alone. LG so far has the least painful interface, Motorolla should get an F for interface design (Whoever thought that having separate entries for each of 4 phone/fax numbers for a single person was a good idea should have to navigate phones using that system for the rest of their lives) Using a cellphone for anything other than a phone (with the occasional camera shot) is so painful as to be useless.
Enter the iPhone. At the very least, it will spark a much needed overhaul of interface design. At worst (for the competitors) it will dominate the market. After all, how many $300+ iPods were sold? Now you get a super duper cell phone to go along with it plus a host of other easy to use features (easy compared to current cell phones) in a relatively slick and sleek package that will interface seamlessly with your computer.
The core of IT is coding. Without it, there's no IT.
The claimed shortage is in software engineers/programmers of various types, which are also in general the higher paid tasks. I don't mean to diminish the other IT areas, but I don't believe there's much of a shortage in those areas with a couple of very specific exceptions.
I've done everything from System/Email/DB (sort of) administration to network/systems/integration architecture, in addition to software. I am, or was at one time, an expert in several of those fields. Unfortunately with the speed of change in technology, some of those have fallen by the wayside, obsolete and largely useless knowledge unless you wish to support some legacy crap in a dark corner of some large company.
What is interesting is how some of the mistakes of those legacy systems are repeated over and over in "new" systems, including being designed right into new software written in newer languages. Basically we either haven't learned from our earlier mistakes or have failed to solve the original problem.
Slashdot Mirror
Wait a few years? It's here - the AT&T/Verizon telco choice. If you get to choose which one serves you, you're one of the less than 1%ers. The rest of us don't have a choice.
Cell phones are also shrinking down, we now have AT&T(/Cingular), Verizon, T-Mobil and Sprint/Nextel.
Radio is down to 4 companies, effectively, and one of those, ClearChannel, also owns 33% of XM (a main reason I'm not an XM subscriber).
I have seen Night of the Comet, it's bad, but funny bad and at least watchable. Both those other movies aren't.
Oh good god. Was that the one with the 3 waves of ICBMs (US and USSR) attempting to blow up the asteroid?
:)
If so, I'd agree with you that a movie has to be pretty bad to top that one. Here's one that easily bests it at being horrible: Reptilicus.
So next you're going to promote recording all meetings and archiving tapes?
Seriously, what ever gave anyone the idea that email "must" be saved? This move to require archiving IMs is equally intrusive and ridiculous, because next they'll require saving all voice conversations... after all, A/V IM is here already, and it's not too great a leap to go from archiving IM to archiving A/V IM to archiving phone calls....
I'd mod you up, just for disparaging Armageddon. That piece of tripe was terrible. I thought deep Deep Impact was much better, although too much mushy story in there, so still only a C+ at best.
I wish I had mod points. Excellent joke!
OS/2 most certainly did suffer that problem and was quite infamous for it (30 seconds Googling for "Single Input Queue problem" should tell you that).
MS's OSes still suffer that problem, especially if you use some of the legacy APIs (ie, old apps).
No, they don't.
OK, I'll admit I was wrong on OS/2's single input queue. They fixed it in the versions I ran though with a work around. So it seemed like I didn't have that problem.
However, you're wrong as well. Windows NT certainly did have a single input queue, albeit an "OS only" queue, it still can lock up the entire system.
No, it didn't. OS/2 and Windows NT came from _completely_ different code bases. They shared an API - you could run text mode OS/2 apps in early releases of NT - but that's the closest they came to having "common components"[0].
Again incorrect - there was support for HPFS (which MS owned rights to), the OS/2 1.x API, and even OS/2's PM API.
You are thinking of NT's "API personalities", which were designed in from that start as modular chunks that could be activated and deactivated at will (win32 is implemented the same way). They were a result of its microkernel-ish architecture, not of any "OS/2 heritage".
The only "personalities" I recall being talked about at that time were IBM's OS personalities running on top of their microkernel architecture. IIRC, they demoed a single box running NT, OS/2 and RISC 6000 OS personalities on different monitors.
As for the "modular" APIs, I believe all but POSIX is completely gone now. If they were truly modular, why did no one other than MS ever release a module? (I can't think of one, Cygwin is an installed program after all)
You're correct in that OS/2 2.x was radically different from NT (VMS) in design. It was far superior, and by 2.3 had a GUI and performance I have yet to see matched by anyone, including Mac OSX.
It was most certainly not superior (and OS X is hardly a performance benchmark to strive for, it's probably the slowest mainstream OS around). OS/2 had a monolithic kernel, wasn't portable, was single user (so nothing even as basic as filesystem permissions, let alone the pervasive security model of NT), didn't support multiple processors (although that was hacked into some specialised builds), still had significant subsystems that were still 16 bit, even into the mid-90s, had poor memory management (no dynamically-sizing disk cache, for example), etc, etc.
It had a hybrid kernel, not quite monolithic, not quite micro. The HPFS filesystem was a stripped down system for a single box. It was the only common 16 bit piece still running in OS/2 2.3 onward. It could be replaced with HPFS386, which was 32 bit and solved most of your other observations (you'd need some of the additional server modules to run true multi-user on a box)
And exactly what "pervasive security model" permeates NT? The one that allows random code to run and modify files everywhere? That's some security.
And then let's get to where it was superior - the threading model. A truly pre-emptive threading model versus the time-slice. MS still has issues with the time slice, even though it's down to 16ms in XP.
Windows NT was superior to OS/2 in pretty much every conceivable way. Which it should have been, given that - as the original poster noted - it was built to replace OS/2.
Hence Microsoft stating that NT should be rebooted at least once a week (Exchange Servers) and once a month (others) vs OS/2 systems that just ran and ran... (running Sendmail, I might add)
It required less (OS/2 was tolerable on a machine with 6MB RAM, NT required at least 10MB) - but that's hardly surprising since it wasn't _doing_ anywhere near as much.
No wonder your experience was horrible. I wouldn't dream of running it on less than
According to the show - the critical age is somewhere between 3-4. After that, grammar/sentence structure is no longer possible, while word acquisition is. That appears to stop around 8-12, although the number of cases is (fortunately) exceedingly small and that data therefore highly unreliable.
It's uninstallable. Just not by Mom and Pop, or even anything under an expert guru.
Live update is a pita. Requires a manual cleansing of the registry and file system. bah!
You're not serious? Read some History. Before the IBM/Microsoft breakup over OS/2, what became Windows NT 3.1 was supposed to be OS/2 3.0.......
No released version of OS/2 bears any architectural resemblance to Windows NT. None. Five minutes listing the key features and looking at the block diagram of each should tell anyone with a passing acquaintance to OS design that.
In short, the poster I replied to is wrong - NT does not have the Single Input Queue problem that OS/2 did (and probably still does).
I think you have that wrong - OS/2 didn't suffer the problem. MS's OSes still suffer that problem, especially if you use some of the legacy APIs (ie, old apps).On OS/2, you're also not 100% correct. OS/2 1.x, which were released and in widespread usage in the banking industry, especially in ATMs, actually did have some common components with NT 3.1. NT 3.5 separated this some, and it was left as a user installed support module in NT 4.0 before disappearing in Win 2K.
You're correct in that OS/2 2.x was radically different from NT (VMS) in design. It was far superior, and by 2.3 had a GUI and performance I have yet to see matched by anyone, including Mac OSX. (disclaimer, I never used BeOS, which was about the only OS at the time that was reported to exceed OS/2 in performance and some really liked their GUI). If you think OS/2 didn't perform, your hardware wasn't good enough, as it required the same hardware as NT 3.5 to run well.
The GGGP however, is incorrect about OS/2. OS/2 didn't suffer from this problem in the same way, since each application ran on its own thread(s). A properly designed PM app actually had separate threads for input and output, which helped even more. There was no single system wide thread for input/output. Even the Win32s apps weren't locked in the same way they were on MS OSes, hence the "Better windows than Windows" statement.
I'd personally love to see IBM offer the PM and filesystem components of OS/2 running on a Linux kernel. The PM interface would solve one of the major issues with KDE/Gnome, and their HPFS386 file system (since 2MB cache in HPFS is just too small these days) is an incredible performer for workstations. The licensing/patent deal on that should have expired as well, so there's no more $80 per copy payment to MS. (If you ever wondered why OS/2 wasn't sold for less to compete with MS, there's one major part of it)
There's a show on Discovery or TLC about feral children, young children that were severely neglected. The show focuses on the language and brain development with age, as several of these children had human contact as they grew from infant to 6 or more years of age. (The oldest was around 12 when she was found).
Despite the show being deeply disturbing, it is also fascinating from a scientific viewpoint regarding social development, language development, and brain function/development.
As it relates to the P/GP, humans apparently learn both to work together and be competitive from their environment. It is based on feedback loops. There was one boy, found at age 3 or 4 in the Ukraine, that lived with a pack of dogs. He'd become adopted because he offered food to a stray, and over time the pack grew, or so the theory goes. He developed the behavioral aspects of a dog for both the competitive and cooperative aspects.
My point was that their costs have dropped significantly, they've only experienced a 4% drop in revenue, yet the revenue drop is what they're citing as proof that online sales/piracy is hurting them deeply. Revenue is irrelevant when the model changes. You could have a 50% drop in revenue yet still increase profit if your item costs went from 70% of the revenue stream down to It should also be noted that previous numbers were most likely inflated due to the fact that people had to buy 10-20 songs in a package when they only wanted 1 or 2.First, recall Revenue != Profits. Most companies that go under do so not for lack of revenue but lack of profits.
It certainly does mean revenue numbers were inflated. When you package things so that you have to buy 9X what you need your revenue was just inflated nine fold. This appears when someone else comes along and sells just the amount you need, and takes business away from the 9X packaging. That's one thing that's happening in the music industry, and it's high time it happened. I remember the death of $0.50 - $0.99 45s when all of a sudden I had to buy $16+ CDs, or, amazingly, $7 LPs which used to be $12-16 prior to the introduction of CDs.No, this doesn't mean revenues were inflated, it measn they used to pay $9-$18 per song, now they pay $1 per song (and still you bitch, moan, and whine?). If they make up for this by buying 9x to 18x more songs, revenues stay the same. Since most costs associated with the industry aren't variable w/ the number of units sold, only then do profits stay the same.
Here's another thing for you to ponder: CDs cost between $0.80 and $1.50 for the entire manufacture/packaging/shipping process and sell for an MSRP of around $15-18. LPs cost around $3 for the same, and run about $7-10. Care to explain?
You know - here's a thought - is the music industry really in any trouble whatsoever? Think about this - their claim to woe and "damage" is that revenue is dropping. Isn't that what's supposed to happen when your costs drop? Note with online transactions they're not incurring any manufacturing, packaging, shipping, nor physical theft/damage losses, all which raise the price of physical media vs online media, not to mention dead inventory.
It should also be noted that previous numbers were most likely inflated due to the fact that people had to buy 10-20 songs in a package when they only wanted 1 or 2.
So, the real question would be if their legitimate profits are sufferring. Not their net profit, as I'm sure the RIAA requires a pretty penny to leash and feed, but actual profit from their "product".
Example:
You can send your admin-monkey to the server, with a few manual procedure steps,
This configuration change will go out on the network with the next reboot. And poof! 500 nosy, troublesome Users are now a bit less able to shoot themselves in the foot, or work mischief on your systems. First, I love the MAD acronym.
Secondly, here's a question for you: does OSX even require this functionality, or is it merely a consequence of the MS world-view that this functionality seems to be required?
Let's look at your example, and let me admit I've not used OSX in an enterprise setting, but I have used Solaris, HPUX, Linux, and AIX in enterprise settings and all are *nix variants like OSX. First, you have to image all your drives - that's standard across all systems. Next, you have central servers with user profile information on them on one variant or another (again, standard in this scenario). With the *nix variants, the user home directories can be NFS mounted, with every machine giving you the same view instantly, with the same performance as you'd experience on any other machine. Unless ADS has changed, I believe a new profile is downloaded/updated on every login/logoff, and is slower than molasses if your system is configured with or default/user stores large files on the desktop or in the profile. Also, should I want to change run perms, I change it on the server(s) and voila - INSTANT changes in what users can do - no logoff/login cycle required. Now, OSX being a *nix variant, can probably be setup exactly the same way (The only uncertainty remains because I haven't done it nor experienced it first hand). Other examples: disable the IE address bar. (and prevent Trojans from hooking it). Disable the Tools menu so users can't mung with the security settings in IE. Disable control panels. Enforce a password-protected screen saver across the enterprise. Take the File-Open menu away from MS Excel. Whatever. I assure you, as draconian and capricious as these sound - some of them are ABSOLUTELY NECESSARY to operate computers in a secure environment. These are all windows centric issues, however, given the above configuration with NFS mounts, applications can be controlled in exactly the same way. General users don't have access to system settings in network *nix environments without a super user password. Your strawmen don't exist in a standard *nix rollout.
...But the effort required to "roll your own" system to manage client configuration on this scale, with this ease of use, would be on a pretty much unimaginable level. I'm afraid you're mistaken on that. See below: I am an unashamed Mac fanboy. The bane of my life is when I have to go into work, and fix broken Domain Policies or MAD server. I have 4 Macs at home, and I try to manage them somewhat like an enterprise - and I'm telling you - the tools just are not there. There *is* a usable infrastructure, but you'd need to pump tens of thousands of man-hours from a very skilled scripting guru to pull off the equivalent thing on a Mac. I think you'll find that with a half-way knowledgeable *nix system administrator that more than half your issues will go away, and a decent one will make 95% of your issues evaporate. You're still seeing the world through MS glasses. You expect systems to work like MS systems. MS systems are standalone systems that sort of communicate (badly) through networks, since networks were after thoughts. *nix systems were designed with networks (well, multiple terminals/multi-user systems which were much easier to convert to network centric systems) in mind and thus have a completely different view of networked systems.AD, kind of short for ADHD...
Jokes aside, ADS did create a directory services structure for windows. It still sucks rocks, and is trying to still come up to NDS (which is pretty much a goner these days).
I still remember Exchange pre-ADS. It was fast. Try sending a 250K user list in Exchange w/ ADS.... you'll be waiting hours, and that's only if you're lucky and you've done some serious optimizations in your mailing list structures.
ADS sucks in just about every way possible, it's another standard MS clunky POS. BTW, Exchange sucks rocks too, there's just not anything better yet in the categories it addresses, but that doesn't mean Exchange is "good". Think of it as you've got a yugo, a fiat, and a hyundai and those are your only choices. They're all bad unreliable POSes, but the hyundai will be your best bet out of those three.
30 years ago, a simple +-*/ calculator was easily twice the size of today's standard calculators. While today's calculators could be made smaller, in general it's cheaper and UI design issues that stabilize the size. After all, using 50+ keys on a calculator requires a certain size in design, and ruggedness requires a certain thickness unless you start going with more exotic (expensive) materials. After all - would you (be able to) use a scientific calculator the size of a postage stamp?
Your sassy AC comments aside, there's a lot more to it than what I posted. But without that core tenet, I really do feel that you're only lipsticking the pig.
The entire point is to remove special interests, and allow anyone to effectively run. Yes, this also removes one of the prime components that keeps the 2 party system going - that of effectively raising campaign funds and promoting a single person.
I don't know why you think anyone gets a message out at all in the current setup. Take Al Gore or Bob Dole and contrast the "marketed packaged presidential candidate" with the post campaigning version. Both of their post campaign versions would have kicked butt compared to the stiff cardboard their campaigns made them into.
If you want to fix politics, eliminate campaign funding. Completely. No commercials. Only a slice of time on PBS (that's a purpose that PBS was originally designed for - a medium to reach everyone) with other stations also giving time as needed, such as for debates. (It's in their charters)
If you don't eliminate campaign finance, you continue to support the buying of politicians by business interests, and it shouldn't surprise you that those politicians place the interests of business ahead of their constituents.
OK - that's easy - run a supply chain planning full system, end to end. The model alone for some companies were over 20GB, while the super scaled down dev model usually came in around 1GB. And that's in-memory. Besides that, you'll generally be running at least 1 full fledged DB, several servers (appservers, webservers, various daemons, etc) and potentially an entire SOA on a single machine. Another one would be developing management and maintenance software that also integrates with existing systems such as Tivoli and HPOpenView, including some resources to monitor (usually a full instance of whatever you're trying to manage and monitor.
Ever want to see your memory usage skyrocket, try either of those scenarios. Is a laptop the best venue for development of that sort? No. However, when you need to be portable to show things to potential clients, it's pretty much the only way.
I should also mention that this type of code only runs on server OSes, because it requires over 2GB of RAM, which no client OS supports - that would be MS Windows, BTW, all other current OSes in this class support 4GB or more. MS OSes only supply a max of 3.5GB in user space unless you're running 64bit MS OSes. I'm not sure what 64bit XP is limited to. I'm sure MS is still smarting from the 20bit kernel -> 32bit counter for memory paging that guaranteed BSODs on NT 4.0 prior to SP1.
Samsung - 2 identical N400 phones - one has 3 bars, the other 5 when held next to each other. Battery on phone 1 lasts maybe 1.5 days. Batter on phone 2 lasts 7 days with some usage.
LG - my current fancy-shmancy phone - standby time of at least a week. Can be on calls for hours. I get three bars where the best Samsung got only 1 bar, and 5 bars where the Samsung got 2. The service provider, however, I think blows. Calls drop, connections are sporadic, especially within network. This would be one of the ones claiming fewest dropped calls and a proposer of the tiered internet legislation. They're already tiering their cell service. In-network (free) calls have worse sound quality and are prone to drops, horrible quality, or failure to connect while calling someone out of network (toll) are crystal clear and immediately connect and almost never drop.
They're actually making me miss Sprint.
Interesting that 3 Motorolla sales folks did not know that.
No SIM card possible on either one I used to have.
Overall, LG, Samsung, and Nokia are all far superior to Motorolla. If it wasn't for the Razr family, I don't think anyone would willingly buy a Motorolla phone at if you'd seen the competition.
I strongly suspect you're right. It's going to be one of those things that people try out, and go "Hey, I really like this. It does x,y, and z, and I can dump 2 other devices and just carry one, and I get bonus feature v to boot!"
I know when I saw the presentation, I went "Now there's what I've been looking for in a phone. I personally hate cell phone interfaces. I'm sure I'm not alone. LG so far has the least painful interface, Motorolla should get an F for interface design (Whoever thought that having separate entries for each of 4 phone/fax numbers for a single person was a good idea should have to navigate phones using that system for the rest of their lives) Using a cellphone for anything other than a phone (with the occasional camera shot) is so painful as to be useless.
Enter the iPhone. At the very least, it will spark a much needed overhaul of interface design. At worst (for the competitors) it will dominate the market. After all, how many $300+ iPods were sold? Now you get a super duper cell phone to go along with it plus a host of other easy to use features (easy compared to current cell phones) in a relatively slick and sleek package that will interface seamlessly with your computer.
The core of IT is coding. Without it, there's no IT.
The claimed shortage is in software engineers/programmers of various types, which are also in general the higher paid tasks. I don't mean to diminish the other IT areas, but I don't believe there's much of a shortage in those areas with a couple of very specific exceptions.
I've done everything from System/Email/DB (sort of) administration to network/systems/integration architecture, in addition to software. I am, or was at one time, an expert in several of those fields. Unfortunately with the speed of change in technology, some of those have fallen by the wayside, obsolete and largely useless knowledge unless you wish to support some legacy crap in a dark corner of some large company.
What is interesting is how some of the mistakes of those legacy systems are repeated over and over in "new" systems, including being designed right into new software written in newer languages. Basically we either haven't learned from our earlier mistakes or have failed to solve the original problem.