And where is the key stored? My iphone has a password, and when i boot it up it starts connecting to my mail servers in the background before i unlock the screen so it clearly has whatever keys it needs to decrypt the data accessible on the device... I would be much happier if it was unable to boot at all until a key had been entered, such that the key was never stored on the device.
Consumer grade tech is much cheaper and far more readily available... While the devices may break more often, replacements and spare parts are easily acquired when necessary. The cheaper cost and more rapid replacement also makes it easier to upgrade to newer devices in the future, rather than being stuck with ancient legacy devices that cannot easily adapt to changing needs and end up being resented by the people who use them. Also, employees are far more likely to already be familiar with consumer devices, and thus require less training on their use.
And while some of these specialised devices may be "highly refined", this is often not the case... Many may do their existing job well, but are difficult to adapt for changing requirements and will serve new purposes very badly or not at all. Similarly because such devices are highly specialized there are very few suppliers of them and the market is very hard to break into, so you often end up with devices suffering from many serious flaws and no willingness from the manufacturers to fix them because they have you locked in anyway.
Consider the parallel with desktop computers... Mainframes and highend unix systems with dumb terminals were reliable, powerful, easy to centrally manage and yet they got replaced by cheap individual non redundant desktop computers running extremely fragile and easily damaged software...
A browser monoculture based on webkit is at least better than a monoculture based on a closed source rendering engine... Just how bad it is, really comes down to who controls it and how much input other people have into it.
Of course without intervention pretty much everything will end up heading towards a monoculture... Linux for instance has pretty much killed the varied proprietary unixes that existed just as x86 has killed the risc processors they ran on.
So if a monoculture is inevitable, then minimising the damage by keeping it open is the best you can hope for.
Modern servers, especially the densely populated ones are far noisier than these older servers, which generally have large but slower spinning fans to cool them...
I had a quad processor sparc32 (a sparcstation 20) running gentoo linux with no problems a couple of years back, it was definitely at least a 2.4 if not 2.6 kernel. I still have the box, and the drives with the linux install on them somewhere.
Those powermac G4 actually support 2gb of ram (4x 512mb), the official spec says 1.5 because that's all OS9 could use and in those days not many people would actually have that much.
Work to "secure" a windows environment is often wasted, since you will still have serious design flaws rendering all your hard work pointless...
As for surface pro, windows tablets have been around for years and you can already run existing software on them... They have always failed while the ipad succeeded, and the surface pro only changes one of those reasons while leaving the others...
1, The OS is not touch friendly, well windows 8 goes a long way towards fixing this but it still has its quirks... 2, The apps are not touch friendly - installing your own existing apps isn't gonna be popular if they are unusable, you will end up rewriting them anyway at least to add a new touch friendly ui. 3, The hardware is bulkier than an ipad or android device, with inferior battery life, surface pro still has this problem. For something your meant to carry around in one hand, bulky is not good. 4, The hardware costs more - surface pro hasn't solved this either.
A lot of closed source software is extremely fragile, primarily because it was only ever designed to build in one place so you often end up with something that is an absolute nightmare to build
It's hard to make money from LibreOffice, because it's competing against a monopoly market leader... If the market was more open and competitive it would be much easier, primarily by offering support, training and customisation services around it.
Ohh also, look for contracting positions, you already have a company so your half way there... Lots of companies don't want to take on permanent staff at the moment because they are harder to get rid of than contractors. They see uncertain times ahead, and want to be able to shed staff if necessary.
By creating your own company you have demonstrated a possible conflict of interest...
Being a graduate doesn't mean a whole lot, if you are a recent graduate then chances are you have no real world experience, and thus no real world references.
Being a graduate shows you can read books and understand the theory and/or cram for exams, but doesn't show if you have a natural interest or aptitude for the subject at hand. Personally for an IT position i'd rather hire a geek who is self taught than some random joe who thought "people make a lot of money in computing, i will study a course", ie people for whom it's a genuine interest and not just a job.
And if you do have a genuine interest, demonstrate it... Participate in communities, publish code you've written, even write a blog... Even if noone normally reads it prospective employers might, and if the writing there is intelligent then it can only help your case.
It also suggests you are less likely to work the bottom rung positions, which while the pay might be shit they get you the first step towards experience.
If you're talking pre-OSX mac, then you have to compare it to what else was around at the time... Windows 3.x and 9x were also extremely well known for being slow and crashing constantly, AmigaOS was massively faster but still very prone to crashing.
If you wanted a decent reliable system you had to fork out for a highend unix box, which would be reliable and fast but too expensive for most people.
So you get your application software from a program provided by the OS vendor which automatically downloads, installs and updates it?
Wasn't so long ago that people were claiming this distribution model was scary for users, and that users would demand to drive to a physical store and buy software in a box containing physical media...
Linux distros were always way ahead in this respect, but they failed to promote the benefits to users... Just look at all the lowsy distributions that came preinstalled on netbooks.
The walled garden only applies to iOS... OSX although it now provides the convenience of an app store (just as most linux distros provide a repository), you are not forced to use it. OSX is still far more open than windows in many many ways.
You need to get people interested young, before their ability to learn is damaged...
More specifically, start them off on something they are free to experiment with and cannot break and let their own natural curiosity encourage them to learn. The worst thing you can do is expose youngsters to windows especially on shared machines... It's a fragile environment which is easily damaged, and which actively discourages users from poking into its internals. And even worse in a shared environment, where kids will be punished if they damage anything, which only teaches them not to be curious.
There are a shortage of competent workers, not a shortage of workers in general. Graduates are not necessarily competent, and almost always lack experience. The biggest problem is that many people simply follow what they've been told, and don't think for themselves.
MS is in a large part responsible for the laughable way in which "computing" is taught in schools... Not only have they pushed very hard for schools to simply teach how to use MS products rather than about computing in general, but they are also directly responsible for scaring millions of people off from learning about how their machines work...
When you bought early home computers like a C64 you were actually encouraged to learn about it, and if you crashed it the absolute worst case was hitting reset because the core system was held in ROM and couldn't be damaged by the user. MS on the other hand have always pushed an extremely fragile system which is very easy to damage, and then covered it with all manner of scary warnings "this directory contains system files, don't look here!" etc.
The first perceptions at a young age have a significant affect on later life... If your first exposure to computing is a system you are free to experiment with, knowing nothing bad will happen to you then a childs natural curiosity will drive them to learn about it. On the other hand if you are faced by a fragile system covered in scary warnings, and even worse someone (eg parents, teachers) who will become angry if you break it then you become afraid and won't experiment, won't learn or think for yourself and will very rigidly follow instructions for fear of what might happen if you don't.
If you want to teach kids about computing, give them systems they can't break (or at the very least a trivial way to restore them to how they were) and let them explore. Teachers should be there to guide them and point things out, but not to dictate a specific path the kids must take. Something that would work well is games which include sourcecode (eg the old BASIC games on C64 and similar), let them play for a while and then show them how the change the rules...
Logic says PowerPC is easier to support because there is far less hardware to worry about... There are only a small number of powerbook models, with known wireless, ethernet, video chipsets etc. There are millions of x86 machines with an even wider range of possible peripherals, including often oem versions where the hardware is the same but small firmware differences cause compatibility problems etc, and problems where certain combinations of hardware simply don't work together.
You could run Simcity 2000 for Acorn, which would be a native ARM binary and thus not need cpu emulation (although you would need to virtualize the os its expecting).... On a modern ARM cpu with no processor emulation overhead it should run extremely fast.
Unless those computers never interact with the outside world at all (eg removable media etc), they are still at risk of being compromised and having all your data destroyed...
Faster depends how much change has to be counted out (either by the customer or the clerk, sometimes both), how fast the card processing terminal and the network it uses is and of course the individual customer. Cash transactions can be painfully slow.
Then there is the risk of the cash being lost or stolen, the hassle of getting it to the bank, the additional risk of errors (miscounting, over changing customers, or under changing and being caught out), the risk of fake currency, the cost of getting change (yes banks impose a surcharge when they deliver big bags of small change to retailers).
Also while you actually get paid immediately, that cash isn't earning you anything while its sat in the store, it's actually just a liability because the more cash you have in the store - the more attractive a target you are for theft. You need to get the cash to the bank, and depending on the size of your operation the cash collection may not occur daily, or you may have to take the cash to the bank yourself which could also be dangerous.
Personally i avoid paying with cash simply because it costs the same. If paying by cash was sufficiently cheaper that it was beneficial to do so i would. Other people however have more trouble managing their money, and might be severely inconvenienced by having to pay cash.
Maintaining a monopoly is easy, you don't have to compete like everyone else does, you just have to be "not horrendously bad" and you will retain customers. Any competitors have a serious uphill struggle, and have to be significantly superior in order for anyone to even consider them against a dominant monopolist.
Continued growth is also easy when you have a monopoly in a growing market, you will just grow along with it - again, your competitors have to be hugely superior before anyone will even consider them while you just have to be mediocre.
Similarly expanding into related markets is easy when you dominate one, you can leverage that market to force yourself into others - even if your products are inferior to the competition. You know that the potential clients will be buying your products in the market you dominate, so you use that leverage to get yourself into others be it through freebies, or scuppering the competition due to incompatibility or exclusivity deals.
MS only ever advocate open standards and interoperability in markets where they are doing badly, or where forced to do so by external forces...
In markets they dominate, they always try to do the exact opposite.
Come back when they start advocating ODF and CalDAV etc.
They may not be threatening mono, but while java code is cross platform by default (and by accident).net code often only works on mono if specifically written to be cross platform, which is just another way to keep smaller platforms down.
And where is the key stored?
My iphone has a password, and when i boot it up it starts connecting to my mail servers in the background before i unlock the screen so it clearly has whatever keys it needs to decrypt the data accessible on the device...
I would be much happier if it was unable to boot at all until a key had been entered, such that the key was never stored on the device.
Consumer grade tech is much cheaper and far more readily available... While the devices may break more often, replacements and spare parts are easily acquired when necessary.
The cheaper cost and more rapid replacement also makes it easier to upgrade to newer devices in the future, rather than being stuck with ancient legacy devices that cannot easily adapt to changing needs and end up being resented by the people who use them.
Also, employees are far more likely to already be familiar with consumer devices, and thus require less training on their use.
And while some of these specialised devices may be "highly refined", this is often not the case... Many may do their existing job well, but are difficult to adapt for changing requirements and will serve new purposes very badly or not at all. Similarly because such devices are highly specialized there are very few suppliers of them and the market is very hard to break into, so you often end up with devices suffering from many serious flaws and no willingness from the manufacturers to fix them because they have you locked in anyway.
Consider the parallel with desktop computers... Mainframes and highend unix systems with dumb terminals were reliable, powerful, easy to centrally manage and yet they got replaced by cheap individual non redundant desktop computers running extremely fragile and easily damaged software...
A browser monoculture based on webkit is at least better than a monoculture based on a closed source rendering engine...
Just how bad it is, really comes down to who controls it and how much input other people have into it.
Of course without intervention pretty much everything will end up heading towards a monoculture... Linux for instance has pretty much killed the varied proprietary unixes that existed just as x86 has killed the risc processors they ran on.
So if a monoculture is inevitable, then minimising the damage by keeping it open is the best you can hope for.
And how many of those who downloaded 3.4.1, also had 3.4.0 before? Even MS makes minor updates available for free...
The lobbyists, being based in the US are not doing anything illegal. It is the MEPs who would be breaking the law by accepting the money.
Modern servers, especially the densely populated ones are far noisier than these older servers, which generally have large but slower spinning fans to cool them...
I had a quad processor sparc32 (a sparcstation 20) running gentoo linux with no problems a couple of years back, it was definitely at least a 2.4 if not 2.6 kernel. I still have the box, and the drives with the linux install on them somewhere.
Those powermac G4 actually support 2gb of ram (4x 512mb), the official spec says 1.5 because that's all OS9 could use and in those days not many people would actually have that much.
Work to "secure" a windows environment is often wasted, since you will still have serious design flaws rendering all your hard work pointless...
As for surface pro, windows tablets have been around for years and you can already run existing software on them... They have always failed while the ipad succeeded, and the surface pro only changes one of those reasons while leaving the others...
1, The OS is not touch friendly, well windows 8 goes a long way towards fixing this but it still has its quirks...
2, The apps are not touch friendly - installing your own existing apps isn't gonna be popular if they are unusable, you will end up rewriting them anyway at least to add a new touch friendly ui.
3, The hardware is bulkier than an ipad or android device, with inferior battery life, surface pro still has this problem. For something your meant to carry around in one hand, bulky is not good.
4, The hardware costs more - surface pro hasn't solved this either.
A lot of closed source software is extremely fragile, primarily because it was only ever designed to build in one place so you often end up with something that is an absolute nightmare to build
It's hard to make money from LibreOffice, because it's competing against a monopoly market leader...
If the market was more open and competitive it would be much easier, primarily by offering support, training and customisation services around it.
Ohh also, look for contracting positions, you already have a company so your half way there... Lots of companies don't want to take on permanent staff at the moment because they are harder to get rid of than contractors. They see uncertain times ahead, and want to be able to shed staff if necessary.
By creating your own company you have demonstrated a possible conflict of interest...
Being a graduate doesn't mean a whole lot, if you are a recent graduate then chances are you have no real world experience, and thus no real world references.
Being a graduate shows you can read books and understand the theory and/or cram for exams, but doesn't show if you have a natural interest or aptitude for the subject at hand. Personally for an IT position i'd rather hire a geek who is self taught than some random joe who thought "people make a lot of money in computing, i will study a course", ie people for whom it's a genuine interest and not just a job.
And if you do have a genuine interest, demonstrate it... Participate in communities, publish code you've written, even write a blog... Even if noone normally reads it prospective employers might, and if the writing there is intelligent then it can only help your case.
It also suggests you are less likely to work the bottom rung positions, which while the pay might be shit they get you the first step towards experience.
If you're talking pre-OSX mac, then you have to compare it to what else was around at the time...
Windows 3.x and 9x were also extremely well known for being slow and crashing constantly, AmigaOS was massively faster but still very prone to crashing.
If you wanted a decent reliable system you had to fork out for a highend unix box, which would be reliable and fast but too expensive for most people.
So you get your application software from a program provided by the OS vendor which automatically downloads, installs and updates it?
Wasn't so long ago that people were claiming this distribution model was scary for users, and that users would demand to drive to a physical store and buy software in a box containing physical media...
Linux distros were always way ahead in this respect, but they failed to promote the benefits to users... Just look at all the lowsy distributions that came preinstalled on netbooks.
The walled garden only applies to iOS... OSX although it now provides the convenience of an app store (just as most linux distros provide a repository), you are not forced to use it. OSX is still far more open than windows in many many ways.
You need to get people interested young, before their ability to learn is damaged...
More specifically, start them off on something they are free to experiment with and cannot break and let their own natural curiosity encourage them to learn. The worst thing you can do is expose youngsters to windows especially on shared machines... It's a fragile environment which is easily damaged, and which actively discourages users from poking into its internals. And even worse in a shared environment, where kids will be punished if they damage anything, which only teaches them not to be curious.
There are a shortage of competent workers, not a shortage of workers in general.
Graduates are not necessarily competent, and almost always lack experience. The biggest problem is that many people simply follow what they've been told, and don't think for themselves.
MS is in a large part responsible for the laughable way in which "computing" is taught in schools...
Not only have they pushed very hard for schools to simply teach how to use MS products rather than about computing in general, but they are also directly responsible for scaring millions of people off from learning about how their machines work...
When you bought early home computers like a C64 you were actually encouraged to learn about it, and if you crashed it the absolute worst case was hitting reset because the core system was held in ROM and couldn't be damaged by the user.
MS on the other hand have always pushed an extremely fragile system which is very easy to damage, and then covered it with all manner of scary warnings "this directory contains system files, don't look here!" etc.
The first perceptions at a young age have a significant affect on later life... If your first exposure to computing is a system you are free to experiment with, knowing nothing bad will happen to you then a childs natural curiosity will drive them to learn about it. On the other hand if you are faced by a fragile system covered in scary warnings, and even worse someone (eg parents, teachers) who will become angry if you break it then you become afraid and won't experiment, won't learn or think for yourself and will very rigidly follow instructions for fear of what might happen if you don't.
If you want to teach kids about computing, give them systems they can't break (or at the very least a trivial way to restore them to how they were) and let them explore. Teachers should be there to guide them and point things out, but not to dictate a specific path the kids must take.
Something that would work well is games which include sourcecode (eg the old BASIC games on C64 and similar), let them play for a while and then show them how the change the rules...
Logic says PowerPC is easier to support because there is far less hardware to worry about...
There are only a small number of powerbook models, with known wireless, ethernet, video chipsets etc. There are millions of x86 machines with an even wider range of possible peripherals, including often oem versions where the hardware is the same but small firmware differences cause compatibility problems etc, and problems where certain combinations of hardware simply don't work together.
You could run Simcity 2000 for Acorn, which would be a native ARM binary and thus not need cpu emulation (although you would need to virtualize the os its expecting).... On a modern ARM cpu with no processor emulation overhead it should run extremely fast.
Unless those computers never interact with the outside world at all (eg removable media etc), they are still at risk of being compromised and having all your data destroyed...
Faster depends how much change has to be counted out (either by the customer or the clerk, sometimes both), how fast the card processing terminal and the network it uses is and of course the individual customer. Cash transactions can be painfully slow.
Then there is the risk of the cash being lost or stolen, the hassle of getting it to the bank, the additional risk of errors (miscounting, over changing customers, or under changing and being caught out), the risk of fake currency, the cost of getting change (yes banks impose a surcharge when they deliver big bags of small change to retailers).
Also while you actually get paid immediately, that cash isn't earning you anything while its sat in the store, it's actually just a liability because the more cash you have in the store - the more attractive a target you are for theft. You need to get the cash to the bank, and depending on the size of your operation the cash collection may not occur daily, or you may have to take the cash to the bank yourself which could also be dangerous.
Personally i avoid paying with cash simply because it costs the same. If paying by cash was sufficiently cheaper that it was beneficial to do so i would. Other people however have more trouble managing their money, and might be severely inconvenienced by having to pay cash.
Maintaining a monopoly is easy, you don't have to compete like everyone else does, you just have to be "not horrendously bad" and you will retain customers. Any competitors have a serious uphill struggle, and have to be significantly superior in order for anyone to even consider them against a dominant monopolist.
Continued growth is also easy when you have a monopoly in a growing market, you will just grow along with it - again, your competitors have to be hugely superior before anyone will even consider them while you just have to be mediocre.
Similarly expanding into related markets is easy when you dominate one, you can leverage that market to force yourself into others - even if your products are inferior to the competition. You know that the potential clients will be buying your products in the market you dominate, so you use that leverage to get yourself into others be it through freebies, or scuppering the competition due to incompatibility or exclusivity deals.
MS only ever advocate open standards and interoperability in markets where they are doing badly, or where forced to do so by external forces...
In markets they dominate, they always try to do the exact opposite.
Come back when they start advocating ODF and CalDAV etc.
They may not be threatening mono, but while java code is cross platform by default (and by accident) .net code often only works on mono if specifically written to be cross platform, which is just another way to keep smaller platforms down.