But that's not what he said. There's a big difference between promoting a product as non-Microsoft and educating people on why Microsoft is bad for them.
[Ironic]
That ridiculous web page is an *excellent* example of the problem GP was talking about.
If the portion of OSS community who cared about it expended as much effort on making their software better (or event just finding out what end users wanted) as they did proclaiming "Micro$oft is t3h sux0rs", they'd probably *have* decent competitors to Exchange, Office and AD already.
The reason for this is that 95% of the businesses which are Microsoft shops have a sales, marketing and middle management that is MS Exchange addicted and is living under the false impression that it is good process and business practice to drag any person from any other part of the business into a meeting on a whim based on his schedule in Exchange.
I know a hell of a lot of Engineers who are similarly "addicted".
Of course, this is largely because it allows them to game the system and avoid management harassment by marking their calendars "busy" or "in a meeting" when they want to be left alone...
With that said, quite frankly I find the hostility against the sort of scheduling/messaging integration Exchange & friends offers to be completely irrational. It's an excellent example of the kind of automation and communications streamling computers are *supposed* to be providing for us. You should not be blaming the system because certain users are narcissists, or because it makes it easier for your boss to figure out whether or not you're actually working. When one person can find out if persons A, B and C are free at time X just by having a quick look in their schedules and then schedule a co-ordinated meeting in a couple of minutes, that is a *massive* improvement over having to communicate and co-ordinate a suitable time over the course of half a day (or more) while people are in and out of the office.
This isn't organising meetings between management and/or sales staff, either, its between technical staff with multiple projects who are teamed together on some but not others. The functionality provided by Exchange & co. allows them to significantly manage their time better.
In order to convert even a part of a Microsoft addicted business [...]
In order to "convert" a "Microsoft addicted business" you need to offer them a product that provides the same functionality at a dramatically lower cost, or provides dramatically better functionality. Until then, you will not be able to sell them on a conversion - and rightfully so, as well. Why would any business want to take a step backwards in efficiency ?
There's a reason why Exchange is so popular. It's because it does an important job better than anything else. Very few other products - and *nothing* in the OSS world - even comes close.
But not before that. And Microsoft knows this and does their best to provide "solutions" which allow you not to compartamentalise your business.
You seem to have that arse-about-face. Integrated scheduling/communications makes it *easier* to "compartmentalise your business" because it makes inter-department communications and co-ordination so much better/easier.
Keep in mind that Vista is designed as a new generation of operating system, taking advantage of it (or even using it) will require pretty advanced and expensive hardware...
A several-years-old standard desktop PC is "advanced and expensive hardware" ?
It never would because in unregulated capitalism, as soon as someone gained enough market share to set whatever price they want, someone will be there that's willing to provide the services for cheaper and still make a good profit. This is a basic fact of supply and demand.
This hinges on the assumption that the barriers to entry are low enough for new competitors to enter the market *and survive long enough to become profitable*.
That assumption (and it is a best-case assumption) is, at best, shaky.
The worst-case assumption is even worse - that a new competitor can enter the market and survive, even when the established companies are both wearing temporary losses undercutting it on price *and* (where applicable) using their existing relationships with suppliers to restrict the supply of raw materials. To say a new provider in the market could survive in such a scenario in any less than extraordinary circumstances is delusional.
I just don't believe this to be the case. As Murray Rothbard said, "The very term "public utility" is an absurd one. Every good is useful "to the public," and almost every good may be considered "necessary."
I can't agree. Some "utilities" are *clearly* more "essential" than others. Water, electricity, (basic) communications (eg: to emergency services), sanitation, police protection, fire protection, basic medical care, etc. I have never heard any argument against these as "essential services" that isn't venturing well and truly into absurd libertarian/anarchistic/rich white men fantasies.
This is not to say that every single aspect of such services need be government owned and controlled. For example, power distribution and power generation are two clearly separate aspects of "electricity", but the latter is trivially compatible with a "free market", whereas the former is not. Thus, IMHO, power *distribution* should be government owned and controller, while power *generation* should be privatised. This gives the maximum benefit to everyone - generation companies can market themselves to consumers as being cheap/green/clean/whatever, allowing the consumer to decide where their money should go (and trivially switch between them), but no company is endowed with the natural monopoly that results from owning such a massive piece of infrastructure as electricity grids.
Similarly, new competitors into the market are not faced with the effectively insurmountable barrier to entry of having to a) establish a widespread backbone grid, b) convince people to switch to their grid and c) connect to their houses. On the consumer side, they are not faced with having to run multiple physical power links onto their property and suffer power outages when any transfers are made.
While it is true to say "every good is useful to the public" (I would argue it should be "every good is useful to *someone*"), it does *not* follow that "every good is essential to the public (ie: *everyone*)" or "no goods are essential to the public".
Any designation of a few industries as "public utilities" is completely arbitrary and unjustified."
Of course it is arbitrary, but to say it is unjustified is ludicrous. You cannot, for example, sustain functional high-density urban living without proper sanitation and clean water. Both of these services require massive levels of infrastructure which is extremely invasive (and expensive) for any potential competitors to duplicate, and consumers to choose between. Fire protection is another example of a service that is essential to high-density population.
I contend that having the Government own (and maintain) this infrastructure, charging fees to all who use it, is more efficient than having the Government apply (potentially discriminatory) regulation to make access to privately-owned equivalents "fair" or having that infrastructure duplicated multiple times.
(However, one thing I am completely opposed to is collaboration between the public and private sector, especially on large
Not necessarily. From what I've read, one form of socialism is where control may be indirectly exercised on behalf of the people by the State, which is basically what we have now.
It seems to me this is semantics, particularly in the theoretical context this discussion appears to be concerned with.
Anyone who claims that such and such a bad thing will happen with unrestrained capitalism basically has nothing to go on, so they are really just fearmongering.
Well, not really. They can look at the results of "restrained capitalism" and extrapolate from there based on what companies try to do before getting beaten back by regulations and laws.
As I've said elsewhere, I can't see any possible result from "unrestrained capitalism" except cartels and monopolies (and, ultimately, just a single entity controlling "everything").
However, if they do, they are now responsible for 100% of the illegal stuff that goes across their wires. Someone downloads kiddy porn? Arrest the CEO. Gambling? Arrest them. Drug purchases? Seize their assets. DDOS attack orginating or terminating on your network? You pay for lost time, damages, penalties.
Your logic is faulty.
You are saying that because they can control one aspect of traffic (its "speed") then they can by definition control all other aspects (eg: its "content"). You are trying to conflate two completely different things.
It's like saying if a private company can set speed limits on a toll road they are liable if someone uses that road to transport a body aftering committing murder.
It's as if a shipping company opened up every single package and was going to bill a higher price per pound for magazines than for books w/o pictures. If they were to see the magazines were kiddie porn, but since that was a higher profit shippment they put it back in the box and collect their fee. They'd sure as heck be criminally liable then.
Right. Except these ISPs aren't trying to prioritise traffic based on its content, but based on its source and destination.
A correct version of your analogy would have shipping companies billing different rates for the same weight package depending upon who was shipping it.
I'm afraid that you, good sir, are the one who does not understand how free markets work. The reason our "capitalism" has collapsed into cartels and monopolies is that we have seen a vast influx of socialism into our modern system.
Eh ? It's hard to see how unregulated capitalism would ever end up anywhere else *except* "cartels and monopolies". The only variable is how long it takes, which should be inversely proportional to the cost of market entry (the more expensive it is to enter the market, the sonner "cartels and monopolies" will be controlling it).
The "free market" works well for some things and badly for others. One of the things it works badly for is infrastructure/essential service type utilities, mainly because the costs - for both sellers to enter the market and buyers to move between competitors - is so high.
Infrastructure and essential services should be Government owned and controlled (ie: by everyone). Everything else can be free market.
I'm just saying that "telling" developers to do the rigth thing isn't quite enough.
Well, there's rather a limit as to what else can be done.
That kind of thing needs to be either enforced or the right way nees to be clear in the API.
Enforced how ? How is the Win32 API not unlear ?
The closest they could come to "enforcing" it would be making the default user non-Admin, which is just going to break a lot of software (and blame will subsequently be placed on Microsoft) and just end up with users manually adding themselves to the Administrator's group anyway. That's why there's all these ugly hacks in Vista to "fake" admin-level privileges to broken applications, without actually requiring them.
Other systems manage it. Are Windows developers just especially stubborn and rebelious?
No, their mistakes are just a lot more noticable because 95% of the world's computers show them off and because the end users are more tolerant.
It is too bad Microsoft never had to the courage to just start a new API from scatch.
.NET.
Instead they chose to make almost everything backwards compatable. And now they have a monster API set that few people know how to use correctly.
By far the most common problems with applications not working with non-Admin users have nothing whatsoever to do with any possible API issues and everything to do with laziness, stupidity and ignorance. Things like trying to store runtime data in the application's directory, trying to open system files read/write, trying to write to system areas of the Registry, trying to store user data in the application directory, etc, etc. In fact, I don't think I've *ever* seen an app that needlessly required admin-level privileges, that wasn't requiring them because of some obviously boneheaded decision the developer had made. Probably 95% of "needs admin to run" problems are either an application trying to write data to its own directory, or trying to write to system areas of the Registry.
If a user can fix it by twiddling file/registry permissions and without access to the source, it's got nothing to do with the API and everything to do with the developer.
MS will never switch to Unix style permissions and filesystem setup because that would be a tacit admission that they've got it wrong for the last 13 years.
Actually it's because it would be a dramatic step backwards in capabilities.
If MS had just bitten the bullet and made the original NT unix-like (or even VMS-like) they'd have spared themselves years of grief, bugs, exploits and hassle.
Internally, NT and VMS are so similar that Microsoft was sued by DEC because of it.
But no , they knew better.... yeah right.
They certainly knew well enough not to repeat the mistakes of unix's primitive and coarse security model.
In a system where the file extension is the only thing that determines whether or not a file is executable, it's kind of idiotic that they would ever disable it.
Firstly, it's not the only thing.
Secondly, they're turned off because most users don't know what they mean.
Um, those benchmarks are only for statistical computations. I don't know about you but most computer users aren't performing statistical analysis.
What makes you think the same trend doesn't hold ? What makes you think there's something special or unique about what this benchmark is showing ?
Ask anyone who uses a Mac on a regular basis and they'll tell you it hums along nicely.
I use Macs on a regular basis. On anything less than a top-end dual processor PowerMac G5 (haven't had a chance to try the Core Duos yet), the interactive responsiveness of OS X is awful with more than a trivial load (with the exception of the things that are almost completely handled within the video card).
My 1Ghz, 768M RAM G4 iBook has worse interactive responsiveness than my 550Mhz, 384M RAM P3 Compaq Armada running Windows 2003 (at least until the Armada starts swapping). Running Linux on the iBook makes it *noticably* faster (but nowhere near as nice to use, so I leave OS X on it).
OS X does a lot of things well, but performance - *particularly* interactive responsiveness - is not one of its strong points.
This is the result from developing for an OS that changes its interfaces every few years.
Which "interfaces" in Windows are you thinking of that change "every few years" ?
Look at Unix/Unix like OSes. A port to the next generation or a different incarnation is often a recompile and nothing else.
As it is on Windows. Hell, even having to recompile on Windows at all is unusual.
Why? Because there is a stable API!
Which part of Windows's APIs haven't been stable ?
Nobody uses platform specific stuff, unless there is no choice.
Right. That explains why there's so much "unix" software out there that frequently only works on Linux without modification. Or why there's all those wonderful autoconf hacks for different platforms.
Effect: Far less bugs, far less security critical stuff, because the software is older and well tested! The "cutting edge" (not saying it is, but MS certainly _think_ they are there) has no businedd being used in office and the like environments.
"Unix" software is just as buggy as Windows software.
Honestly, I think all the vendors complaining and all the customers suffering get exactly what they deserve for their short-sightedness.
You mean their short-sightedness for not writing modular code ?
What I know is that people will be forced to use it, and that new Microsoft apps will require Vista features to work properly.
And when you say "forced" you mean "go out and buy themselves". And when you say "new Microsoft apps" you mean "new Microsoft apps released 5+ years later".
I'm not aware of that example (wrong country, wrong time), but Mathematics education is having similar problems to English for similar reasons - maths teachers that barely know the basics (have trouble manipulating fractions, for example), curriculums and philosophies that place little emphasis on actually getting the right answer (or even close) and an over-reliance on calculators at the expense of mental calculations (and the number manipulation tricks that make them easier).
The end result being that it's not at all unusual to see people under the age of 25 or so regularly using calculators to do things like multiply by 0 or 1 or subtract 10%. Ask them to do something like divide one fraction into another and you're lucky if they don't pass out on the spot.
I am a teacher, and my grammar skills are just fine.
Then you should be well aware of all the problems I mentioned. Mathematics is another field in similar peril, for similar reasons.
I am not a teacher, however, I have several very close friends and relatives who are, or used to be. Do not mistake of thinking this as an issue I am viewing from afar, or with only an "outsiders" knowledge.
Your essay is poorly orgnanized and wanders from the point.
Your assumption is flawed. I was writing a post to Slashdot, not an essay. My post was compiled in a series of one or two minute gaps over a 30 - 45 minute timeframe, of _course_ it's poorly organised and unfocused.
You are exactly right 5 quad port cards and and onboard. I considered a managed switch but this way was actually much cheaper.
Are you sure ? Sometimes it's hard to see the big picture cost-wise. Has the cost of your ongoing troubles in terms of frustration and downtime added up to the value of a VLAN-capable switch yet ?
Students tune out grammatical rules. What matters is that they learn how good writing sounds. Once you get the cadence and rhythm of proper English, you are able to generate it much more easily.
This is an incredibly important point I overlooked, that cannot be emphasised enough. While the saying "in English, it is the exception that is the rule" certainly has some validity, if you have been exposed to sufficient "good" writing, you instinctively know it is "good". Once you're at this point, it becomes easy to identify bad spelling and/or grammar, even if you're not entirely sure of which technical rules it's breaking (and, hence, how to make it "sound right").
The downside is that incorrect English then sounds like nails on a chalk board (e.g. "more easy" versus "easier").
I'm glad there are other people out there who mentally (and physically, sometimes) wince whenever they read something like that:).
Heck, Microsoft has started talking (so far just talking, as near as I can tell, but it's a start) about userland drivers, and you *know* they're not going to a microkernel.
No, they've already *got* one.
(To be fair NT hasn't been especially microkernel-ish in implementation since 3.51, but I'm sure the microkernel-based architecture and design hasn't changed substantially since then, it's just had a lot of subsystems pushed into kernel mode, which are going to be moved back out into user mode).
The result is that if you have reasonably common hardware the kernel is getting much more stable but for things like my non PCI sparc(compile problem with some options) or my 21 ethernet port firewall (needs special options to boot or it crashes) it has gotten more buggy.
Assuming you mean 21 physical ethernet ports, goddamn that sounds an awful way to be doing things (5 quad port cards and an onboard NIC ?). Have you considered investing in a switch that does 802.1q vlan tagging and cutting down the number of physical connections to, say, two or three ?
Slashdot Mirror
[Ironic]
That ridiculous web page is an *excellent* example of the problem GP was talking about.
If the portion of OSS community who cared about it expended as much effort on making their software better (or event just finding out what end users wanted) as they did proclaiming "Micro$oft is t3h sux0rs", they'd probably *have* decent competitors to Exchange, Office and AD already.
I know a hell of a lot of Engineers who are similarly "addicted".
Of course, this is largely because it allows them to game the system and avoid management harassment by marking their calendars "busy" or "in a meeting" when they want to be left alone...
With that said, quite frankly I find the hostility against the sort of scheduling/messaging integration Exchange & friends offers to be completely irrational. It's an excellent example of the kind of automation and communications streamling computers are *supposed* to be providing for us. You should not be blaming the system because certain users are narcissists, or because it makes it easier for your boss to figure out whether or not you're actually working. When one person can find out if persons A, B and C are free at time X just by having a quick look in their schedules and then schedule a co-ordinated meeting in a couple of minutes, that is a *massive* improvement over having to communicate and co-ordinate a suitable time over the course of half a day (or more) while people are in and out of the office.
This isn't organising meetings between management and/or sales staff, either, its between technical staff with multiple projects who are teamed together on some but not others. The functionality provided by Exchange & co. allows them to significantly manage their time better.
In order to convert even a part of a Microsoft addicted business [...]
In order to "convert" a "Microsoft addicted business" you need to offer them a product that provides the same functionality at a dramatically lower cost, or provides dramatically better functionality. Until then, you will not be able to sell them on a conversion - and rightfully so, as well. Why would any business want to take a step backwards in efficiency ?
There's a reason why Exchange is so popular. It's because it does an important job better than anything else. Very few other products - and *nothing* in the OSS world - even comes close.
But not before that. And Microsoft knows this and does their best to provide "solutions" which allow you not to compartamentalise your business.
You seem to have that arse-about-face. Integrated scheduling/communications makes it *easier* to "compartmentalise your business" because it makes inter-department communications and co-ordination so much better/easier.
A several-years-old standard desktop PC is "advanced and expensive hardware" ?
This hinges on the assumption that the barriers to entry are low enough for new competitors to enter the market *and survive long enough to become profitable*.
That assumption (and it is a best-case assumption) is, at best, shaky.
The worst-case assumption is even worse - that a new competitor can enter the market and survive, even when the established companies are both wearing temporary losses undercutting it on price *and* (where applicable) using their existing relationships with suppliers to restrict the supply of raw materials. To say a new provider in the market could survive in such a scenario in any less than extraordinary circumstances is delusional.
I just don't believe this to be the case. As Murray Rothbard said, "The very term "public utility" is an absurd one. Every good is useful "to the public," and almost every good may be considered "necessary."
I can't agree. Some "utilities" are *clearly* more "essential" than others. Water, electricity, (basic) communications (eg: to emergency services), sanitation, police protection, fire protection, basic medical care, etc. I have never heard any argument against these as "essential services" that isn't venturing well and truly into absurd libertarian/anarchistic/rich white men fantasies.
This is not to say that every single aspect of such services need be government owned and controlled. For example, power distribution and power generation are two clearly separate aspects of "electricity", but the latter is trivially compatible with a "free market", whereas the former is not. Thus, IMHO, power *distribution* should be government owned and controller, while power *generation* should be privatised. This gives the maximum benefit to everyone - generation companies can market themselves to consumers as being cheap/green/clean/whatever, allowing the consumer to decide where their money should go (and trivially switch between them), but no company is endowed with the natural monopoly that results from owning such a massive piece of infrastructure as electricity grids.
Similarly, new competitors into the market are not faced with the effectively insurmountable barrier to entry of having to a) establish a widespread backbone grid, b) convince people to switch to their grid and c) connect to their houses. On the consumer side, they are not faced with having to run multiple physical power links onto their property and suffer power outages when any transfers are made.
While it is true to say "every good is useful to the public" (I would argue it should be "every good is useful to *someone*"), it does *not* follow that "every good is essential to the public (ie: *everyone*)" or "no goods are essential to the public".
Any designation of a few industries as "public utilities" is completely arbitrary and unjustified."
Of course it is arbitrary, but to say it is unjustified is ludicrous. You cannot, for example, sustain functional high-density urban living without proper sanitation and clean water. Both of these services require massive levels of infrastructure which is extremely invasive (and expensive) for any potential competitors to duplicate, and consumers to choose between. Fire protection is another example of a service that is essential to high-density population.
I contend that having the Government own (and maintain) this infrastructure, charging fees to all who use it, is more efficient than having the Government apply (potentially discriminatory) regulation to make access to privately-owned equivalents "fair" or having that infrastructure duplicated multiple times.
(However, one thing I am completely opposed to is collaboration between the public and private sector, especially on large
It seems to me this is semantics, particularly in the theoretical context this discussion appears to be concerned with.
Well, not really. They can look at the results of "restrained capitalism" and extrapolate from there based on what companies try to do before getting beaten back by regulations and laws.
As I've said elsewhere, I can't see any possible result from "unrestrained capitalism" except cartels and monopolies (and, ultimately, just a single entity controlling "everything").
Your logic is faulty.
You are saying that because they can control one aspect of traffic (its "speed") then they can by definition control all other aspects (eg: its "content"). You are trying to conflate two completely different things.
It's like saying if a private company can set speed limits on a toll road they are liable if someone uses that road to transport a body aftering committing murder.
It's as if a shipping company opened up every single package and was going to bill a higher price per pound for magazines than for books w/o pictures. If they were to see the magazines were kiddie porn, but since that was a higher profit shippment they put it back in the box and collect their fee. They'd sure as heck be criminally liable then.
Right. Except these ISPs aren't trying to prioritise traffic based on its content, but based on its source and destination.
A correct version of your analogy would have shipping companies billing different rates for the same weight package depending upon who was shipping it.
Have you _ever_ seen any examples of "unrestrained capitalism" ?
Uh, isn't the point of socialism to place the resources "into the hands" of *everyone* ?
Eh ? It's hard to see how unregulated capitalism would ever end up anywhere else *except* "cartels and monopolies". The only variable is how long it takes, which should be inversely proportional to the cost of market entry (the more expensive it is to enter the market, the sonner "cartels and monopolies" will be controlling it).
The "free market" works well for some things and badly for others. One of the things it works badly for is infrastructure/essential service type utilities, mainly because the costs - for both sellers to enter the market and buyers to move between competitors - is so high.
Infrastructure and essential services should be Government owned and controlled (ie: by everyone). Everything else can be free market.
Well, there's rather a limit as to what else can be done.
That kind of thing needs to be either enforced or the right way nees to be clear in the API.
Enforced how ? How is the Win32 API not unlear ?
The closest they could come to "enforcing" it would be making the default user non-Admin, which is just going to break a lot of software (and blame will subsequently be placed on Microsoft) and just end up with users manually adding themselves to the Administrator's group anyway. That's why there's all these ugly hacks in Vista to "fake" admin-level privileges to broken applications, without actually requiring them.
Other systems manage it. Are Windows developers just especially stubborn and rebelious?
No, their mistakes are just a lot more noticable because 95% of the world's computers show them off and because the end users are more tolerant.
It is too bad Microsoft never had to the courage to just start a new API from scatch.
.NET.
Instead they chose to make almost everything backwards compatable. And now they have a monster API set that few people know how to use correctly.
By far the most common problems with applications not working with non-Admin users have nothing whatsoever to do with any possible API issues and everything to do with laziness, stupidity and ignorance. Things like trying to store runtime data in the application's directory, trying to open system files read/write, trying to write to system areas of the Registry, trying to store user data in the application directory, etc, etc. In fact, I don't think I've *ever* seen an app that needlessly required admin-level privileges, that wasn't requiring them because of some obviously boneheaded decision the developer had made. Probably 95% of "needs admin to run" problems are either an application trying to write data to its own directory, or trying to write to system areas of the Registry.
If a user can fix it by twiddling file/registry permissions and without access to the source, it's got nothing to do with the API and everything to do with the developer.
Actually it's because it would be a dramatic step backwards in capabilities.
If MS had just bitten the bullet and made the original NT unix-like (or even VMS-like) they'd have spared themselves years of grief, bugs, exploits and hassle.
Internally, NT and VMS are so similar that Microsoft was sued by DEC because of it.
But no , they knew better.... yeah right.
They certainly knew well enough not to repeat the mistakes of unix's primitive and coarse security model.
Firstly, it's not the only thing.
Secondly, they're turned off because most users don't know what they mean.
What makes you think the same trend doesn't hold ? What makes you think there's something special or unique about what this benchmark is showing ?
Ask anyone who uses a Mac on a regular basis and they'll tell you it hums along nicely.
I use Macs on a regular basis. On anything less than a top-end dual processor PowerMac G5 (haven't had a chance to try the Core Duos yet), the interactive responsiveness of OS X is awful with more than a trivial load (with the exception of the things that are almost completely handled within the video card).
My 1Ghz, 768M RAM G4 iBook has worse interactive responsiveness than my 550Mhz, 384M RAM P3 Compaq Armada running Windows 2003 (at least until the Armada starts swapping). Running Linux on the iBook makes it *noticably* faster (but nowhere near as nice to use, so I leave OS X on it).
OS X does a lot of things well, but performance - *particularly* interactive responsiveness - is not one of its strong points.
Which "interfaces" in Windows are you thinking of that change "every few years" ?
Look at Unix/Unix like OSes. A port to the next generation or a different incarnation is often a recompile and nothing else.
As it is on Windows. Hell, even having to recompile on Windows at all is unusual.
Why? Because there is a stable API!
Which part of Windows's APIs haven't been stable ?
Nobody uses platform specific stuff, unless there is no choice.
Right. That explains why there's so much "unix" software out there that frequently only works on Linux without modification. Or why there's all those wonderful autoconf hacks for different platforms.
Effect: Far less bugs, far less security critical stuff, because the software is older and well tested! The "cutting edge" (not saying it is, but MS certainly _think_ they are there) has no businedd being used in office and the like environments.
"Unix" software is just as buggy as Windows software.
Honestly, I think all the vendors complaining and all the customers suffering get exactly what they deserve for their short-sightedness.
You mean their short-sightedness for not writing modular code ?
And when you say "forced" you mean "go out and buy themselves". And when you say "new Microsoft apps" you mean "new Microsoft apps released 5+ years later".
I'm not aware of that example (wrong country, wrong time), but Mathematics education is having similar problems to English for similar reasons - maths teachers that barely know the basics (have trouble manipulating fractions, for example), curriculums and philosophies that place little emphasis on actually getting the right answer (or even close) and an over-reliance on calculators at the expense of mental calculations (and the number manipulation tricks that make them easier).
The end result being that it's not at all unusual to see people under the age of 25 or so regularly using calculators to do things like multiply by 0 or 1 or subtract 10%. Ask them to do something like divide one fraction into another and you're lucky if they don't pass out on the spot.
Then you should be well aware of all the problems I mentioned. Mathematics is another field in similar peril, for similar reasons.
I am not a teacher, however, I have several very close friends and relatives who are, or used to be. Do not mistake of thinking this as an issue I am viewing from afar, or with only an "outsiders" knowledge.
Your essay is poorly orgnanized and wanders from the point.
Your assumption is flawed. I was writing a post to Slashdot, not an essay. My post was compiled in a series of one or two minute gaps over a 30 - 45 minute timeframe, of _course_ it's poorly organised and unfocused.
Do not confuse rant with reason.
I don't see how I did.
Are you sure ? Sometimes it's hard to see the big picture cost-wise. Has the cost of your ongoing troubles in terms of frustration and downtime added up to the value of a VLAN-capable switch yet ?
Ugh. Damn, that's embarassing, particularly since its/it's is one of the larger annoyances.
(In my defence, the first sentence I wrote there used "it's" correctly - I still should have checked it more thoroughly, however.)
This is an incredibly important point I overlooked, that cannot be emphasised enough. While the saying "in English, it is the exception that is the rule" certainly has some validity, if you have been exposed to sufficient "good" writing, you instinctively know it is "good". Once you're at this point, it becomes easy to identify bad spelling and/or grammar, even if you're not entirely sure of which technical rules it's breaking (and, hence, how to make it "sound right").
The downside is that incorrect English then sounds like nails on a chalk board (e.g. "more easy" versus "easier").
I'm glad there are other people out there who mentally (and physically, sometimes) wince whenever they read something like that :).
No, they've already *got* one.
(To be fair NT hasn't been especially microkernel-ish in implementation since 3.51, but I'm sure the microkernel-based architecture and design hasn't changed substantially since then, it's just had a lot of subsystems pushed into kernel mode, which are going to be moved back out into user mode).
Darwin(/OS X) is about as much of a microkernel as Windows NT.
Which is to say, not much - while it might be by design and architecture, it certainly isn't by implementation.
And they took about a decade to do it...
Assuming you mean 21 physical ethernet ports, goddamn that sounds an awful way to be doing things (5 quad port cards and an onboard NIC ?). Have you considered investing in a switch that does 802.1q vlan tagging and cutting down the number of physical connections to, say, two or three ?