It's not the horrendousness of the API... He was complaining that apple's api is only available from objective-c code, whereas microsoft allow you to code to their api in c, c++ and more...
Second, software virtually never comes with a guarantee, you can shout as loud as you like about some commercial company selling you garbage, but you already waived your right to any comeback when you agreed to their terms. If software makers were accountable for bugs, Microsoft would have gone bankrupt years ago.
They're not talking about desktop software, but open source as a whole...
If you deploy an Apache webserver, you're using open source... If you buy a wireless access point running an embedded linux, you're using open source. If you buy a firewall based on linux, you're using open source.
It really is everywhere, wether you realise it or not... I'm surprised it's not 90% of more already, it's many years since i went to a business that wasn't running at least some kind of open source based application somewhere... Even Microsoft run wireless access points based on an embedded version of linux.
There's always someone to blame, wether the code is open or proprietary... And in both situations, blaming them usually won't do any good, because they are simply under no obligation to care about your problems.
As for users doing testing, open source has that option but you don't need to make use of it... Noone is forcing you to run beta versions. Commercial vendors do the same too.
And with active development, sure open source projects die, and commercial products cease to be developed too... There are thousands of dead commercial applications out there and just as many dead open source projects. The difference is where you go when the software you use dies...
If a proprietary app dies, you probably wont be able to buy more licenses if you need to deploy it on more systems, not a problem with open source.
If a proprietary app dies it will be completely unmaintained, and noone is going to fix anything for you, you will have to migrate away from it. If an open source app dies, and it's sufficiently important to you, there's always the possibility of employing or contracting programmers to maintain it for you.
If you do decide to migrate away, a proprietary app is far more likely to be storing it's data in proprietary formats, whereas an open source app will more likely keep it's data in open formats which are much easier to convert to a format supported by whatever you migrate to.
Obviously it's not a great situation to be in, to have a piece of software you use cease to be maintained, but the problem is much worse if that software was closed source and proprietary.
Yes, absolutely... When something is safety critical, i would like the code to be completely open to third party inspection... I would also much prefer that a single rock solid piece of code is developed, rather than multiple companies reinventing the wheel... It makes no sense for Boeing and Airbus to develop their own autopilot code completely independently, instead they should pool their resources to ensure there is as much safety checking as possible.
They are going to develop the code, and thoroughly audit it anyway... Having third parties looking at it is a good thing, even if they don't accept external code submissions. And ensuring that the standards are as high as possible, and that everyone is up to the same high standards is important too. Much better than some smaller manufacturer in a country with less strict safety rules, who is far more likely to take shortcuts when writing their own code.
On either system you can set things up entirely through the gui, modern redhat comes with graphical tools for configuration etc... The difference is that you're typical unix admin is a step up from the typical windows admin, so you have a greater chance of a unix admin being competent enough to do things by hand if they want/need to. And then there's the fact that doing things by hand gives you more power and control, and more highly skilled people are more likely to want and appreciate a higher level of control. Aside from that, doing things by hand is often a lot easier on linux than it is on windows, text based configuration files are much easier to manipulate using standard text editing tools (of which there are many).
Why should you pay for the software *and* the support? Why should you be forced to obtain support from the same company who wrote the software?
You want supported open source software? Give IBM, Sun, HP, SGI, Novell, RedHat, or many other companies a call. And since theres multiple vendors, there's competition, and competition is good for the consumer.
You're reasoning is also a very common fallacy... Software almost always comes with NO GUARANTEES, wether you pay for it or not (read the license agreements)... And the supplier is under NO OBLIGATION WHATSOEVER to fix any bugs.
With proprietary code, you have a single vendor who *could* fix the bug for you, but they are under no obligation to do so or to charge you a fair price for doing so. They can turn around and refuse to fix the bug, or charge you a billion dollars for doing it, leaving you with no other option.
By contrast, open source code can be fixed by multiple parties, assuming you bought a supported package of open source software from eg RedHat you can go there first... Failing that, you can *always* pay third party programmers to fix the bugs for you.
While your requirements are completely valid, the logical course of action with such requirements is open source.
No they can't... The old platform ceases to be supported, security fixes wont be released, bugs wont be fixed, new apps wont run, new hardware wont have drivers, eventually it just becomes unworkable to keep using an old platform and you have no choice but to upgrade.
The Amiga used to be about the best platform available, fast and multitasking... but you wouldn't use one now, all the software is very old as is the hardware, and a fast os can only compensate for slow hardware to a certain extent.
Exchange is pretty bad if you're ever forced to use it... It's really terrible at handling different timezones, sometimes gets confused by DST, sometimes mail gets completely lost in the system, and then shows up months later... It also keeps all its mail spools in a jet-derived database, a pretty half assed database engine that's prone to corruption and has a rather small maximum size limit. also because of this, you need to buy expensive third party backup software to back up the mail spools (hidden costs), which is absolutely essential because no company would allow their mail server to be not backed up. It also works really poorly with any client other than the corresponding version of outlook so it's largely impractical if you have users using other mail clients, you can access the mail via imap and address book with ldap, but there's seemingly no way to pull the calendar data and other stuff out... The web interface is also pretty crap, and works poorly with any browsers than the latest version of msie, and doesnt seem to work at all with safari.
On the other hand, you have... postpath - a drop in exchange replacement (learned of it through a banner ad on slashdot), you can use the same crappy outlook client if you want, but it has a better web interface, stores its mail in a more standard format, scales better and will hopefully support standards for calendaring etc soon.. Lotus notes - seems to be used at some large companies, not used it much myself openxchange - meant to be good zimbra - meant to be better
I only really use email myself, but having been forced to use exchange before it couldn't possibly be very difficult to find something superior.
While it would be good for it to be unix based, then every major os would be unix based (what else is there, vms is pretty much a small niche, and then theres embedded os's but many are unix based). They simply won't do that, because it would be too compatible with other unixes, and therefore much easier to replace (good for consumer, bad for the microsoft lock-people-in business model).
The original NT kernel was quite a good design, made by people from DEC... It's all the microsoft stuff they shoehorned into it where all the problems stem from... and then their attempts to replace flawed implentations with new, while keeping the old flawed ones there for backwards compatibility (eg lanman encryption, all the library functions with numbers after them etc)...
Unlikely, while a unix based app would make it easier to port apps *to* the new windows, it would also make it easier to port apps away from it to linux... Microsoft don't want to make it easier for third parties to choose what's best for them, they want to take away that choice so that the users have to run microsoft.
Some information doesn't age... Why should just because something is made in an old piece of software, it not be readable anymore... I have ASCII text files from long before win95 existed, and they still work perfectly. And think of all the great literary works that were written on paper, they are still readable today.
If you're developing windows apps, you will be forced to pay the $150 to buy win7 in order to test compatibility of your apps anyway... It might make a lot of sense to make cross platform apps in any case, if you make the apps in a cross platform way it's easy to make them work on linux and osx too, bringing you more potential users, and easier to make them support new systems in the future. Considering that both OSX and Linux are increasing in popularity, i would want to at have apps at least ready to roll, so that anyone looking to migrate doesn't immediately have to drop my apps.
As a business, there will come a point when it's simply unfeasible to stay on an old version of windows... It will be unsupported, it won't work with new hardware, it won't run new software, it won't get security patches and nor will any of the apps it runs. So you end up with the choice: Upgrade windows, and endure the hassle and cost Replace windows, and endure the hassle and cost
Microsoft only need to make sure their software is proprietary and nonstandard enough that the hassle and cost of migrating is higher than the hassle and cost of upgrading to the next version of windows. By keeping you locked in with proprietary formats etc, they can afford to make their os more onerous without risking losing customers.
Well, there's always vmware and other full machine emulation apps on which you can install an old os. Although vmware incurs a performance hit, chances are it will still be faster than the hardware that was natively running those older os's.
Users of older OS's that are no longer generally supported by their manufacturer, and are rapidly being phased out. Those users really need to upgrade, it's microsoft's upgrade treadmill, and those machines are probably still pretty serviceable, but the software is unsupported by it's original vendor, and cannot be supported by anyone else. The same applies to people running SunOS 4, MacOS 9 and other such deprecated systems.
The Microsoft path is to upgrade these old machines... If you don't like the option Microsoft gives you, there are various Linux distributions designed for older hardware too.
If the vendor who supplied the software no longer supports it, what makes you think anyone else should cater to it? You're just stifling progress...
Also users of those older windows versions have the choice to install a third party browser, since Microsoft has abandoned their platforms for new browser releases.
And the Amiga is a historic curiosity, I would never consider using it for any serious work, nor would I try to inconvenience anyone else by demanding they support it.
It's not great, it's CSS support while better than 6 (its hard to be worse) is still pretty poor compared to everything else... It's tabbed browsing is fairly lacklustre and it seems to have been rushed, because it's even more unstable than 6 was. It also seems to be even worse with memory than firefox 2 which is much maligned for it's memory leaks.
While playing music is the basic core, it doesn't really matter where that music comes from... A DJ typically does other things too, like produce his own remixes of songs, sometimes doing so live. Technology changes, and the way things are done change as a result. Sometimes the job title changes to reflect the new technology, sometimes it doesn't.
And there's no reason he can't keep a garage full of physical media, while carrying it around on a hard drive.
It makes transportation a lot easier. It discourages theft (its easy to keep track of a single HD, and keep it locked down. its hard to keep track of one in a thousand cds) It lessens the impact of theft/damage/etc, if the HD gets destroyed you still have the original media, and you could easily have 2 or more copies of the HD.
The point is that anyone can, so people will and post binaries online...
If you need to modify the code to make it compile, you decrease the number of people willing/capable of doing so depending on the complexity of the changes. If only a single company can recompile the code, they won't bother doing so unless they see a sufficiently large potential profit from doing so. All of which negatively affects the quantity and diversity of available applications.
I have a massive set of apps for my N800, most of them i didn't compile myself, but use binaries other people had compiled. A lot of those apps were compiled simply because it was easy, people wouldn't have been willing to put significant effort into it.
Lots of things don't display the same in IE6... You're talking about a browser that's over 7 years old now, and woefully behind everything else out there. Slashdot doesn't look the same or load very quickly on my Amiga either.
XP dies in June. Unless they extend it every last one of those boxes is going out the door with Intel's MobiLinux or a distro that supports that platform. Let's talk about BMW, where Linux comes standard with many models. They did extend it, there was a slashdot story about it very recently... Amusingly, this means that users of these small laptops will end up with a system microsoft claim is inferior, and which they're trying hard to make sure new apps don't support... Conversely, apps will continue to support XP in order to run on these popular small laptops, leaving microsoft with even more fragmentation.
It's already annoying enough when "windows mobile" os so crippled and incompatible with other versions of "windows", sharing little more than the name. By contrast, i can recompile apps from my desktop linux system to run on my linux based nokia n800 tablet.
Slashdot Mirror
It's not the horrendousness of the API...
He was complaining that apple's api is only available from objective-c code, whereas microsoft allow you to code to their api in c, c++ and more...
This is a common misconception...
First, you can buy open source from IBM...
Second, software virtually never comes with a guarantee, you can shout as loud as you like about some commercial company selling you garbage, but you already waived your right to any comeback when you agreed to their terms. If software makers were accountable for bugs, Microsoft would have gone bankrupt years ago.
They're not talking about desktop software, but open source as a whole...
If you deploy an Apache webserver, you're using open source...
If you buy a wireless access point running an embedded linux, you're using open source.
If you buy a firewall based on linux, you're using open source.
It really is everywhere, wether you realise it or not... I'm surprised it's not 90% of more already, it's many years since i went to a business that wasn't running at least some kind of open source based application somewhere... Even Microsoft run wireless access points based on an embedded version of linux.
There's always someone to blame, wether the code is open or proprietary...
And in both situations, blaming them usually won't do any good, because they are simply under no obligation to care about your problems.
As for users doing testing, open source has that option but you don't need to make use of it... Noone is forcing you to run beta versions. Commercial vendors do the same too.
And with active development, sure open source projects die, and commercial products cease to be developed too... There are thousands of dead commercial applications out there and just as many dead open source projects. The difference is where you go when the software you use dies...
If a proprietary app dies, you probably wont be able to buy more licenses if you need to deploy it on more systems, not a problem with open source.
If a proprietary app dies it will be completely unmaintained, and noone is going to fix anything for you, you will have to migrate away from it. If an open source app dies, and it's sufficiently important to you, there's always the possibility of employing or contracting programmers to maintain it for you.
If you do decide to migrate away, a proprietary app is far more likely to be storing it's data in proprietary formats, whereas an open source app will more likely keep it's data in open formats which are much easier to convert to a format supported by whatever you migrate to.
Obviously it's not a great situation to be in, to have a piece of software you use cease to be maintained, but the problem is much worse if that software was closed source and proprietary.
Yes, absolutely...
When something is safety critical, i would like the code to be completely open to third party inspection...
I would also much prefer that a single rock solid piece of code is developed, rather than multiple companies reinventing the wheel... It makes no sense for Boeing and Airbus to develop their own autopilot code completely independently, instead they should pool their resources to ensure there is as much safety checking as possible.
They are going to develop the code, and thoroughly audit it anyway... Having third parties looking at it is a good thing, even if they don't accept external code submissions. And ensuring that the standards are as high as possible, and that everyone is up to the same high standards is important too.
Much better than some smaller manufacturer in a country with less strict safety rules, who is far more likely to take shortcuts when writing their own code.
On either system you can set things up entirely through the gui, modern redhat comes with graphical tools for configuration etc...
The difference is that you're typical unix admin is a step up from the typical windows admin, so you have a greater chance of a unix admin being competent enough to do things by hand if they want/need to.
And then there's the fact that doing things by hand gives you more power and control, and more highly skilled people are more likely to want and appreciate a higher level of control.
Aside from that, doing things by hand is often a lot easier on linux than it is on windows, text based configuration files are much easier to manipulate using standard text editing tools (of which there are many).
Why should you pay for the software *and* the support?
Why should you be forced to obtain support from the same company who wrote the software?
You want supported open source software? Give IBM, Sun, HP, SGI, Novell, RedHat, or many other companies a call.
And since theres multiple vendors, there's competition, and competition is good for the consumer.
You're reasoning is also a very common fallacy...
Software almost always comes with NO GUARANTEES, wether you pay for it or not (read the license agreements)... And the supplier is under NO OBLIGATION WHATSOEVER to fix any bugs.
With proprietary code, you have a single vendor who *could* fix the bug for you, but they are under no obligation to do so or to charge you a fair price for doing so. They can turn around and refuse to fix the bug, or charge you a billion dollars for doing it, leaving you with no other option.
By contrast, open source code can be fixed by multiple parties, assuming you bought a supported package of open source software from eg RedHat you can go there first... Failing that, you can *always* pay third party programmers to fix the bugs for you.
While your requirements are completely valid, the logical course of action with such requirements is open source.
No they can't...
The old platform ceases to be supported, security fixes wont be released, bugs wont be fixed, new apps wont run, new hardware wont have drivers, eventually it just becomes unworkable to keep using an old platform and you have no choice but to upgrade.
The Amiga used to be about the best platform available, fast and multitasking... but you wouldn't use one now, all the software is very old as is the hardware, and a fast os can only compensate for slow hardware to a certain extent.
Exchange is pretty bad if you're ever forced to use it...
It's really terrible at handling different timezones, sometimes gets confused by DST, sometimes mail gets completely lost in the system, and then shows up months later... It also keeps all its mail spools in a jet-derived database, a pretty half assed database engine that's prone to corruption and has a rather small maximum size limit. also because of this, you need to buy expensive third party backup software to back up the mail spools (hidden costs), which is absolutely essential because no company would allow their mail server to be not backed up. It also works really poorly with any client other than the corresponding version of outlook so it's largely impractical if you have users using other mail clients, you can access the mail via imap and address book with ldap, but there's seemingly no way to pull the calendar data and other stuff out...
The web interface is also pretty crap, and works poorly with any browsers than the latest version of msie, and doesnt seem to work at all with safari.
On the other hand, you have...
postpath - a drop in exchange replacement (learned of it through a banner ad on slashdot), you can use the same crappy outlook client if you want, but it has a better web interface, stores its mail in a more standard format, scales better and will hopefully support standards for calendaring etc soon..
Lotus notes - seems to be used at some large companies, not used it much myself
openxchange - meant to be good
zimbra - meant to be better
I only really use email myself, but having been forced to use exchange before it couldn't possibly be very difficult to find something superior.
While it would be good for it to be unix based, then every major os would be unix based (what else is there, vms is pretty much a small niche, and then theres embedded os's but many are unix based).
They simply won't do that, because it would be too compatible with other unixes, and therefore much easier to replace (good for consumer, bad for the microsoft lock-people-in business model).
The original NT kernel was quite a good design, made by people from DEC...
It's all the microsoft stuff they shoehorned into it where all the problems stem from... and then their attempts to replace flawed implentations with new, while keeping the old flawed ones there for backwards compatibility (eg lanman encryption, all the library functions with numbers after them etc)...
I'm sure win2k supported drives larger than 120gb... There were at least large scsi disk arrays much bigger than that when win2k came out.
Unlikely, while a unix based app would make it easier to port apps *to* the new windows, it would also make it easier to port apps away from it to linux...
Microsoft don't want to make it easier for third parties to choose what's best for them, they want to take away that choice so that the users have to run microsoft.
Some information doesn't age... Why should just because something is made in an old piece of software, it not be readable anymore...
I have ASCII text files from long before win95 existed, and they still work perfectly.
And think of all the great literary works that were written on paper, they are still readable today.
If you're developing windows apps, you will be forced to pay the $150 to buy win7 in order to test compatibility of your apps anyway...
It might make a lot of sense to make cross platform apps in any case, if you make the apps in a cross platform way it's easy to make them work on linux and osx too, bringing you more potential users, and easier to make them support new systems in the future. Considering that both OSX and Linux are increasing in popularity, i would want to at have apps at least ready to roll, so that anyone looking to migrate doesn't immediately have to drop my apps.
As a business, there will come a point when it's simply unfeasible to stay on an old version of windows...
It will be unsupported, it won't work with new hardware, it won't run new software, it won't get security patches and nor will any of the apps it runs.
So you end up with the choice:
Upgrade windows, and endure the hassle and cost
Replace windows, and endure the hassle and cost
Microsoft only need to make sure their software is proprietary and nonstandard enough that the hassle and cost of migrating is higher than the hassle and cost of upgrading to the next version of windows. By keeping you locked in with proprietary formats etc, they can afford to make their os more onerous without risking losing customers.
Forcing their customers to deal with inferior garbage rather than improving their products is, and always has been, the microsoft way.
Well, there's always vmware and other full machine emulation apps on which you can install an old os. Although vmware incurs a performance hit, chances are it will still be faster than the hardware that was natively running those older os's.
Users of older OS's that are no longer generally supported by their manufacturer, and are rapidly being phased out.
Those users really need to upgrade, it's microsoft's upgrade treadmill, and those machines are probably still pretty serviceable, but the software is unsupported by it's original vendor, and cannot be supported by anyone else.
The same applies to people running SunOS 4, MacOS 9 and other such deprecated systems.
The Microsoft path is to upgrade these old machines... If you don't like the option Microsoft gives you, there are various Linux distributions designed for older hardware too.
If the vendor who supplied the software no longer supports it, what makes you think anyone else should cater to it? You're just stifling progress...
Also users of those older windows versions have the choice to install a third party browser, since Microsoft has abandoned their platforms for new browser releases.
And the Amiga is a historic curiosity, I would never consider using it for any serious work, nor would I try to inconvenience anyone else by demanding they support it.
It's not great, it's CSS support while better than 6 (its hard to be worse) is still pretty poor compared to everything else... It's tabbed browsing is fairly lacklustre and it seems to have been rushed, because it's even more unstable than 6 was. It also seems to be even worse with memory than firefox 2 which is much maligned for it's memory leaks.
While playing music is the basic core, it doesn't really matter where that music comes from...
A DJ typically does other things too, like produce his own remixes of songs, sometimes doing so live.
Technology changes, and the way things are done change as a result. Sometimes the job title changes to reflect the new technology, sometimes it doesn't.
And there's no reason he can't keep a garage full of physical media, while carrying it around on a hard drive.
It makes transportation a lot easier.
It discourages theft (its easy to keep track of a single HD, and keep it locked down. its hard to keep track of one in a thousand cds)
It lessens the impact of theft/damage/etc, if the HD gets destroyed you still have the original media, and you could easily have 2 or more copies of the HD.
The point is that anyone can, so people will and post binaries online...
If you need to modify the code to make it compile, you decrease the number of people willing/capable of doing so depending on the complexity of the changes.
If only a single company can recompile the code, they won't bother doing so unless they see a sufficiently large potential profit from doing so.
All of which negatively affects the quantity and diversity of available applications.
I have a massive set of apps for my N800, most of them i didn't compile myself, but use binaries other people had compiled. A lot of those apps were compiled simply because it was easy, people wouldn't have been willing to put significant effort into it.
Lots of things don't display the same in IE6...
You're talking about a browser that's over 7 years old now, and woefully behind everything else out there.
Slashdot doesn't look the same or load very quickly on my Amiga either.
Amusingly, this means that users of these small laptops will end up with a system microsoft claim is inferior, and which they're trying hard to make sure new apps don't support... Conversely, apps will continue to support XP in order to run on these popular small laptops, leaving microsoft with even more fragmentation.
It's already annoying enough when "windows mobile" os so crippled and incompatible with other versions of "windows", sharing little more than the name. By contrast, i can recompile apps from my desktop linux system to run on my linux based nokia n800 tablet.