The problem is that patents are potentially harmful to all software industry. Companis get patenst to be in a position to trade if attacked. But that only work if the one attacking you have any need for your patents, if you are attacked by patent trolls, there is no remedy but paying their extortion money. The bigger and richer your company is the more likely it is you get attacked. All software industry would be much better off without patents,
Yes, I have to agree, I work for the third largest IT company in the world and they send as many people as they can to attend Android courses. It is a very long time since I saw something this big. Given that there have been very few devices available to consumers I would say 3% market share is amazing.
I'm not so sure I would prefer iPhone hardware over e.g. HTC Hero. After all, in iPhone you can't even replace the battery yourself, and that in a phone that have even worse battery life. It is ridiculous. The Hero also feels more solid than the iPhone, I would say the Android, in HTC Hero, have got the Porche the auther of the original article is waiting for. Unfortunately it is a GSM phone so it will not do much good on the US market.
Uptime for more than a few months probably doesn't matter in most cases.You probably need to reboot your computer in that timeframe anyway e.g. to install kernel patches to keep up with security updates. If you need that kind of availability you should plan your installation in another way and add redundancy.
The important thing to most people is not if the computer is still running, but if the service it provides is available, and windows Vista most certainly can run for a few months, even old win NT3.51 could do that.
The advantage of Linux here is not that it can stay up much longer, but rather that it doesnt cost an arm and a leg in licence costs to set up some kind of fail-over system to provide uninterupted services if you need to reboot or get an hardware failure. Unfortunately the hardware still costs money though./uno
You are right and you are wrong. Yes, it's true that Vista, XP or even Windows 2k are rock solid, but only as long as you don't add third party hardware driveres of dubious quality. Unfortunately many hardware venders don't spend as much effort as they should to develop good drivers. Just using the drivers that comes with windows leaves you with a rather small set of supported hardware, so people install whatever drivers that comes with the hardware they buy, and as a result they get BSOD if they are unlucky, and then they blame Microsoft.
Actually, content laws have a lot to do with freedom of speach. More and more information is distributed on digital media, and if we build systems, technical or legal, to control such media they can be used for other things than making sure that recordcompanies get their "fair" share.
My guess is that even the most frequent travellers will use their laptop more outside an aeroplane than in one, and if you are not in an airplane the larger Lenovo screen is most likely to be preferred by most people. Another absolute showstopper for the Mac is the lack of ports. Dongles breaks, can be misplaced having a real ethernet port is a big bonus. So is having a DVD player, not only for entertainment but it makes it simple to reinstall the OS in case of a virus atack or some other mishap. If I was to give some important presentation away from home I would certainly carry backup media both of my presentation, and the software needed to give it, including the OS. Naturally you can use an external DVD player for the Mac but then the advantage of being small and light is lost.
As for the unlit keyboard, X300 has Think Lihgt a little diode lamp at the top of the screen that gives enough light to make it possible to handle in the dark
The problem is not lawyers in this case, the problem is software patents. Novels, short stories and mathematics, or even legal texts are covered by copyright, why need software be special. Especially mathematics and legal texts have very big likeness to software in that they attempt express something in a non ambigous way. Layers doesn't need to pay licence fees to use e.g a insanity defence in a murder case. Why should software developers be patent suits if they don't pay licences for things that are equally obvious within their field. It is true that many software patents will not hold up in court, but the cost of going to court is so high both in terms of money and in time that very few developers can afford it.
I think you are right about the MacOS. Just like in the case of windows where users doensn't switch to other OS:es because they don't like change, Linux will not change enough to meet the needs of ordinary office/home users because the people developing GUIs for Linux don't like change. They will continue to develop very good interfaces to Unix and Unixlike systems as long as the result still looks like an interface to Unix. But what most non IT people need and want is an interface to make their work easier.
Unfortunately, Unix is irrelevant to most non IT users, sure they want the stability and security Linux/Unix can offer, but most of them are not interested about what they find in/etc,/proc,/lib,/bin,/var,/dev,.. Still these directories seem to be absolutely necessary to everey Linux/Unix GUI develeoper, as apart from MacOS-X developers at Apple, it seams to be totally impossible for a Unix developr to hide them from ordinary users that just gets confused by them.
Most people who actually need to see and edit the contents of thees directoris, usually opens a terminal and edits them in vi. So why bother ordinary users, if they really need to change things here they need nice GUIs to do it. Seeing these files in their file browser will not help them, or anybody else.
Well, keeping the cost down is quite an important reason to many people. Another reason is that it consumes less memory, so the user can use more of it for his own purposes. The fact that Asus manages to sell their product with Linux preinstalled tells us that Linux is good enough for most potential users.
My guess is that very few people will reinstall it with windows. The basis for the idea that windows is easy to use very much comes from that it is preinstalled on almost every PC sold. My experience with windows is that it is quite hard to install compared to most Linux distros, even if you have the right drivers, so I guess the option to install windows on it only will be open to computer literate users.
There also will be very little incentive for the user to do so, the eee pc already have all the applications needed for most people in a palmtop device preinstalled. If you install XP there won't be much room for other advance windows applications.
Yes, I can understand that you are annoyed if you have a lot of old C++ code, but Java is one of the most tought languages to todays students on universites around the world. These students are usually young people that may have other ideas on how to use their g-phones than those of us that started to use a computer back in days when C++ emerged on the scene.
It is also much faster to develop an application in Java than in C++, unless of course you have a lot of old C++ that you could make use of. This means that it will be easy to interest new developers, and cheep to test new ideas on what a phone could be used for, so I think Java was quite a good choice from google.
Actually pirate bay facilitates download of software that is perfectly legal to download, i.e. Linux distributions and other large free software packages. They don't go after postal service because they may carry illegally copied CD:s or even worse bombs, they don't go after telphone service provideers even though somebody may use copyrighted material for music on their answering machine, or use it to plan bank robberies.
What if you want to distribute your own music to promote yourself as an artist without going through RIA channels. This is another perfectly legal thing torrents and Pirate Bay could be used for. In a sense you could claim that RIA is trying to close down alternative ways of publishing music. In the long run, artists finding their customers directly through internet, could be a much bigger threat to RIA, than ordinary consumers pirating a CD or two once in a while.
Firtst of all, at least in KDE3, the menu have four entries "move here", "copy here", "link here" and "cancel". From a concistency point of view, you would find that no other menus have a "Cancel" button so why do this menu need it. The reason is of course that the simple action of moving something have become too complicated.
Second, on my hard dreve, and most other hardrives of people running Linux or some othe kind of unixlike system, the number of links are only around 0.01% of all files on the hard drive, and most of these files are created by some kind of script. So the chance that this menu item should be used is very small.
Third, at least in all konquerer there was no other way to make a link, in the unusual event that you actually would want one. This means that a blind person having menus read to him as he navigates the menus of konquerer with a screen reader probably never will find it, as he most likely will use cut and paste to move or copy files.
Fourth, it doesn't fit in the desktop methaphore. There is no such thing as copying things or making links on your real physical desktop.
Fifth, Mac, Gnome, Windows and CDE users seam to be able to live without it. This means that more than 90% of all new KDE users wil havet to relearn when they switch to KDE. There is nothing wrong in having this menu as an option, but this fact alone is enought to not make it the default behavior.
So, why have a menu where 50% of the items ("link here" and "cancel") are almost never used. Just try to sort a few hundred images from your digital camera with this popping up all the time. It's really annoying, and given that almost everything else is configurable in KDE this should be too, or they could do it the windows way, where you have the best of both worlds, a menu if you do right button drag, and no menu if you do left button drag. For poeple that only have one mouse button modifier keys could be used to bring up the menu when needed.
This is really a usability problem, as it breaks the users flow of operation.
Yes, MySQL has ACID, triggers, and stored procedures, but it still lacks full support for referential integrety. Not to mention that you have to go through hoopsl and loops to turn on whatever referential integrety they have. By default it tries to be MySQL 4 compatible where referential checking was wery lax. MySQL also lacks in more normal SQL language. e.g. it lacks "EXCEPT". Try doing relational division without it.
For those of you that don't know what relational divison is, consider an example with the following relations: COURSE(*CourseID, CourseName), STUDENT(*StudentID,StudentName), STUDIES(*CourseID,*StudentID). Now make a query that gives all the students that studies the exact same combination of courses as sdudnet X. I wouldn't even think of doing that in MySQL 5.
Another problem with MySQL, is speed. MySQL have long had a bait and switch tactics where they have told you that they have the fastest database in the world, but didn't tell you that you had to turn off most normal database features to get it. They also forget to tell you that it scales very porly on multiprocessor systems compared to e.g. Postgresql.
No, lets hope not, at least not if they don't do Postgresql as well. Sun allready invest heavily in Postgresql that is a much better product, both with respect to speed, feetures and standards compliance. As of version 8.3 (now in pre release) it have one of the fastest full text engines in the market and XML support, something that perhaps could prove useful in combination with an office suite.
Now, I don't think they consider either Postgresql or MySQL, as they both are database servers, while the database in OpenOffice is supposed to work with out a lot of server configurations in the same way as Microsoft Access works in MS Office.
First of all, it probably will help the FOSSies. Second it will hurt Microsoft.One or two people running Linux will not hurt them, but when a couple of millions do, then there will be a market for Linux oriented products, more hardware venders will ship Linux drivers, more software will be ported to Linux from other platforms and as more people get aware of Linux some Linux software will be ported to Windows.
The main reason for chosing windows is that all your friends have it, all your documents are created by programs that only run on windows, you can go blindfolded into a store and buy a WiFi card and it will work, and not the least almost all PC:s ship with windows included. Some people may not even know that there is choise, and the people who do, may think that this choice is only for geeks. Nobody runs Windows because it is such good software. They run it because it is good enough, and get the work done.
Most Linux distros will run just fine, on these low price boxes, in fact it will look much better than any windows+hardware combination Microsoft can offer. The lower the threashhold there is to switch to something else than windows, the bigger the chance that poeple will switch, especially if the alternative is free. Free as in free of licensing costs. Free as in free of speach. Free as in free of problems with copyprotection, activation keys and licence management.
In the end this will make it impossible for Microsoft to sell Windows Vista Ultimate, or its successor for $49 and they will have to switch to a more service based business where they give away their OS completely for free. In that situation they will have nothing to gain from not being interoperable with other systems. I'm not saying Microsoft won't be able to make money from this, but most Linux companies will have a head start in this new market. Above all, they will be in a position where they don't have a near monopoly advantage. So yes, a new ultra low price PC market will hurt Microsoft in the long run.
You also have to consider that the number of poor people in the world that doesn't allready have a PC, but can afford one if it is priced low enough, most likely is much bigger than rich people that allready have a powerful PC that they replace every three years or so. This means that there will be much more room for expansion in hte low price market than in the high price market segment.
OS-X is very nice for the desktop, but I would stay away from it on the server side, not because there is all that much technically wrong with it, but because Apple don't seam to get the server market.
By hard earned experience with Apple server products I have learned that you can't trust them to support their products over long times. The all of a sudden discontinues products without any resonable migraton paths to the successr, if there even is a successor. E.g. they dicontiued A/UX and replaced with an Apple version of AIX that they then dropped totally in just a couple of years.
When they distribute updates they have more than once totally destroyed, customized settings, and the open source software that comes with the server version of MacOS-X is often incomplete or lacking in functionality compared to the same software on Linux or Solaris.
Chosing between Windows and Mac, I would choose Mac any day. MacOS-X is at least simple to use.
Nobody is surprised that Vista isn't a success (perhaps with the exception of one or two Microsoft employees). XP did/do the job for most people. So, why upgrade? The only time Vista is worth to consider is if you buy a new machine. But even then, Vista makes you machine more expensive, both in terms of hardware and software. Then there is the question if it will work well with your old existing network of XP or even win2k boxes.
Microsoft had the same problem to get people to upgrade from win2k to XP, but XP didn't look like such a total failure. The reason for that was that there were a lot of win9x users that left that platform for XP. Unlike the win2k users these customers actually got good value for their money, so it was not so hard to make them upgrade.
Another factor is that the competition is much harder now than when they released XP. Apple is starting to get back in the game, and Linux looks better and better and evolving fast.
Google is in no way dependent of Microsoft, If Microsoft was gone tomorrow people would still find a way to use the web and Google services. If they are dependent of anything it would be good network services. Actually, if Microsoft was gone, it would leave google more room to expand in areas like e-mail handling and office application services.
Accessibility is important as it opens new markets for Linux. In many countries this kind of features are required for government use. I would also imagine that not caring for people with disabilities could leave corporate users in some countries open to lawsuits or additional cost when integrating employees that happen to get blind or otherwise disabled during their employment.
I agree with you real usability tests are far better than this, but they are also much more resource intensive to perform. This is why they are not performed as often as one would wish in free software projects.
Just consider a simple thing like the use of a file manager. For one thing, you could get information like, users in general use move file twice as often as they use copy file and 999 time more often than they make symbolik links in konqueror. This could be used in e.g. KDE filemanagers to determine the best order on the "copy here", "move here", "link here", "cancel" popup menu that appears when you drop files over a drop target. (figures just fictional). Or it may turn out that the best thing would be to remove the menu altogheter if one item turn out to be much more used than the others.
You could get information on what directories users opens and edit files in using gui editors.
This could result in new desktop shortcuts or, that some seldom used directories should act like hidden files in the GUI. My suspicion is that/etc/,/dev,/proc/lib/bin/sys are rarely used from the GUI, as people who know what to do with them most likely edit them from a terminal using vi or some other cli text editor.
Having information like this may help the Gnome team to not simplify/remove things that people actually usem and the KDE team to simplify/remove things that people almost never use.
To get the most out of this kind of data, I would try to make statistics of the most probable user action given that a certain action just have been performed. And the frequency of each action. Combine this information with the file types if any involved and user serveys and you would probably get a quite good picture on how the GUI is used.
This will give Microsoft valuable information that will help them develop better software in the future. The quesiton is how many users are prepared to trust them enough to allow this. But as it is comleatly volentary I see no problem in doing so, other than that they may get some biased results as people that value privacy and security most likely will not participate. However, personally, I wourld like to be paid a little more than just with some free software to do this kind of testing for Microsoft.
I would actually like to see open source projects such as Gnome and KDE to do a similar thing (As long as it is volentary). Knowing how the users use their is essential to create good usable systems, and very few usability tests are performed on software in the FOSS world, so something like this would probably be even more beneficial to these kind of projects. I would also think that more people would be prepared to volontear this kind information to e.g. the KDE or Gnome team than they would to big evil Microsoft. After all very few Gnome or KDE users would worry about that a big evil company might discover unlicenced software on their systems.
This statement have high probablility to become just as famous as "There will be a total market for five computers in the world" and "nobody will ever need more than 640k memory".
Over this time new algorithms are likely to turn up, and many of the encryption technologies we today consider safe will turn into the digital eqivalent of a simple, easily broken padlock.
Slashdot Mirror
The problem is that patents are potentially harmful to all software industry. Companis get patenst to be in a position to trade if attacked. But that only work if the one attacking you have any need for your patents, if you are attacked by patent trolls, there is no remedy but paying their extortion money. The bigger and richer your company is the more likely it is you get attacked. All software industry would be much better off without patents,
Yes, I have to agree, I work for the third largest IT company in the world and they send as many people as they can to attend Android courses. It is a very long time since I saw something this big. Given that there have been very few devices available to consumers I would say 3% market share is amazing.
No, Hero is not in the same leage, its an iPhone killer.
Multitasking, multitouch, Adobe flash, very much improved GUI by HTC, and you can change battery and SIM card without tools, ...
I'm not so sure I would prefer iPhone hardware over e.g. HTC Hero. After all, in iPhone you can't even replace the battery yourself, and that in a phone that have even worse battery life. It is ridiculous. The Hero also feels more solid than the iPhone, I would say the Android, in HTC Hero, have got the Porche the auther of the original article is waiting for. Unfortunately it is a GSM phone so it will not do much good on the US market.
Uptime for more than a few months probably doesn't matter in most cases.You probably need to reboot your computer in that timeframe anyway e.g. to install kernel patches to keep up with security updates. If you need that kind of availability you should plan your installation in another way and add redundancy.
The important thing to most people is not if the computer is still running, but if the service it provides is available, and windows Vista most certainly can run for a few months, even old win NT3.51 could do that.
The advantage of Linux here is not that it can stay up much longer, but rather that it doesnt cost an arm and a leg in licence costs to set up some kind of fail-over system to provide uninterupted services if you need to reboot or get an hardware failure. Unfortunately the hardware still costs money though. /uno
You are right and you are wrong. Yes, it's true that Vista, XP or even Windows 2k are rock solid, but only as long as you don't add third party hardware driveres of dubious quality. Unfortunately many hardware venders don't spend as much effort as they should to develop good drivers. Just using the drivers that comes with windows leaves you with a rather small set of supported hardware, so people install whatever drivers that comes with the hardware they buy, and as a result they get BSOD if they are unlucky, and then they blame Microsoft.
Actually, content laws have a lot to do with freedom of speach. More and more information is distributed on digital media, and if we build systems, technical or legal, to control such media they can be used for other things than making sure that recordcompanies get their "fair" share.
My guess is that even the most frequent travellers will use their laptop more outside an aeroplane than in one, and if you are not in an airplane the larger Lenovo screen is most likely to be preferred by most people. Another absolute showstopper for the Mac is the lack of ports. Dongles breaks, can be misplaced having a real ethernet port is a big bonus. So is having a DVD player, not only for entertainment but it makes it simple to reinstall the OS in case of a virus atack or some other mishap. If I was to give some important presentation away from home I would certainly carry backup media both of my presentation, and the software needed to give it, including the OS. Naturally you can use an external DVD player for the Mac but then the advantage of being small and light is lost.
As for the unlit keyboard, X300 has Think Lihgt a little diode lamp at the top of the screen that gives enough light to make it possible to handle in the dark
The problem is not lawyers in this case, the problem is software patents. Novels, short stories and mathematics, or even legal texts are covered by copyright, why need software be special. Especially mathematics and legal texts have very big likeness to software in that they attempt express something in a non ambigous way. Layers doesn't need to pay
licence fees to use e.g a insanity defence in a murder case. Why should software developers be patent suits if they don't pay licences for things that are equally obvious within their field. It is true that many software patents will not hold up in court, but the cost of going to court is so high both in terms of money and in time that very few developers can afford it.
I think you are right about the MacOS. Just like in the case of windows where users doensn't switch to other OS:es because they don't like change, Linux will not change enough to meet the needs of ordinary office/home users because the people developing GUIs for Linux don't like change. They will continue to develop very good interfaces to Unix and Unixlike systems as long as the result still looks like an interface to Unix. But what most non IT people need and want is an interface to make their work easier.
/etc, /proc, /lib, /bin, /var, /dev,.. Still these directories seem to be absolutely necessary to everey Linux/Unix GUI develeoper, as apart from MacOS-X developers at Apple, it seams to be totally impossible for a Unix developr to hide them from ordinary users that just gets confused by them.
Unfortunately, Unix is irrelevant to most non IT users, sure they want the stability and security Linux/Unix can offer, but most of them are not interested about what they find in
Most people who actually need to see and edit the contents of thees directoris, usually opens a terminal and edits them in vi. So why bother ordinary users, if they really need to change things here they need nice GUIs to do it. Seeing these files in their file browser will not help them, or anybody else.
Well, keeping the cost down is quite an important reason to many people. Another reason is that it consumes less memory, so the user can use more of it for his own purposes. The fact that Asus manages to sell their product with Linux preinstalled tells us that Linux is good enough for most potential users.
My guess is that very few people will reinstall it with windows. The basis for the idea that windows is easy to use very much comes from that it is preinstalled on almost every PC sold. My experience with windows is that it is quite hard to install compared to most Linux distros, even if you have the right drivers, so I guess the option to install windows on it only will be open to computer literate users.
There also will be very little incentive for the user to do so, the eee pc already have all the applications needed for most people in a palmtop device preinstalled. If you install XP there won't be much room for other advance windows applications.
Yes, I can understand that you are annoyed if you have a lot of old C++ code, but Java is one of the most tought languages to todays students on universites around the world. These students are usually young people that may have other ideas on how to use their g-phones than those of us that started to use a computer back in days when C++ emerged on the scene.
It is also much faster to develop an application in Java than in C++, unless of course you have a lot of old C++ that you could make use of. This means that it will be easy to interest new developers, and cheep to test new ideas on what a phone could be used for, so I think Java was quite a good choice from google.
Actually pirate bay facilitates download of software that is perfectly legal to download, i.e. Linux distributions and other large free software packages. They don't go after postal service because they may carry illegally copied CD:s or even worse bombs, they don't go after telphone service provideers even though somebody may use copyrighted material for music on their answering machine, or use it to plan bank robberies.
What if you want to distribute your own music to promote yourself as an artist without going through RIA channels. This is another perfectly legal thing torrents and Pirate Bay could be used for. In a sense you could claim that RIA is trying to close down alternative ways of publishing music. In the long run, artists finding their customers directly through internet, could be a much bigger threat to RIA, than ordinary consumers pirating a CD or two once in a while.
Firtst of all, at least in KDE3, the menu have four entries "move here", "copy here", "link here" and "cancel".
From a concistency point of view, you would find that no other menus have a "Cancel" button so why do this menu need it.
The reason is of course that the simple action of moving something have become too complicated.
Second, on my hard dreve, and most other hardrives of people running Linux or some othe kind of unixlike system, the number of links are only around 0.01% of all files on the hard drive, and most of these files are created by some kind of script. So the chance that this menu item should be used is very small.
Third, at least in all konquerer there was no other way to make a link, in the unusual event that you actually would want one. This means that a blind person having menus read to him as he navigates the menus of konquerer with a screen reader probably never will find it, as he most likely will use cut and paste to move or copy files.
Fourth, it doesn't fit in the desktop methaphore. There is no such thing as copying things or making links on your real physical desktop.
Fifth, Mac, Gnome, Windows and CDE users seam to be able to live without it. This means that more than 90% of all new KDE users wil havet to relearn when they switch to KDE. There is nothing wrong in having this menu as an option, but this fact alone is enought to not make it the default behavior.
So, why have a menu where 50% of the items ("link here" and "cancel") are almost never used. Just try to sort a few hundred images from your digital camera with this popping up all the time. It's really annoying, and given that almost everything else is configurable in KDE this should be too, or they could do it the windows way, where you have the best of both worlds, a menu if you do right button drag, and no menu if you do left button drag. For poeple that only have one mouse button modifier keys could be used to bring up the menu when needed.
This is really a usability problem, as it breaks the users flow of operation.
Yes, MySQL has ACID, triggers, and stored procedures, but it still lacks full support for referential integrety.
Not to mention that you have to go through hoopsl and loops to turn on whatever referential integrety they have. By default it tries to be MySQL 4 compatible where referential checking was wery lax. MySQL also lacks in more normal SQL language. e.g. it lacks "EXCEPT". Try doing relational division without it.
For those of you that don't know what relational divison is, consider an example with the following relations: COURSE(*CourseID, CourseName), STUDENT(*StudentID,StudentName), STUDIES(*CourseID,*StudentID). Now make a query that gives all the students that studies the exact same combination of courses as sdudnet X. I wouldn't even think of doing that in MySQL 5.
Another problem with MySQL, is speed. MySQL have long had a bait and switch tactics where they have told you that they have the fastest database in the world, but didn't tell you that you had to turn off most normal database features to get it. They also forget to tell you that it scales very porly on multiprocessor systems compared to e.g. Postgresql.
No, lets hope not, at least not if they don't do Postgresql as well. Sun allready invest heavily in Postgresql that is a much better product, both with respect to speed, feetures and standards compliance. As of version 8.3 (now in pre release) it have one of the fastest full text engines in the market and XML support, something that perhaps could prove useful in combination with an office suite.
Now, I don't think they consider either Postgresql or MySQL, as they both are database servers, while the database in OpenOffice is supposed to work with out a lot of server configurations in the same way as Microsoft Access works in MS Office.
First of all, it probably will help the FOSSies. Second it will hurt Microsoft.One or two people running Linux will not hurt them, but when a couple of millions do, then there will be a market for Linux oriented products, more hardware venders will ship Linux drivers, more software will be ported to Linux from other platforms and as more people get aware of Linux some Linux software will be ported to Windows.
The main reason for chosing windows is that all your friends have it, all your documents are created by programs that only run on windows, you can go blindfolded into a store and buy a WiFi card and it will work, and not the least almost all PC:s ship with windows included. Some people may not even know that there is choise, and the people who do, may think that this choice is only for geeks. Nobody runs Windows because it is such good software. They run it because it is good enough, and get the work done.
Most Linux distros will run just fine, on these low price boxes, in fact it will look much better than any windows+hardware combination Microsoft can offer. The lower the threashhold there is to switch to something else than windows, the bigger the chance that poeple will switch, especially if the alternative is free. Free as in free of licensing costs. Free as in free of speach. Free as in free of problems with copyprotection, activation keys and licence management.
In the end this will make it impossible for Microsoft to sell Windows Vista Ultimate, or its successor for $49 and they will have to switch to a more service based business where they give away their OS completely for free. In that situation they will have nothing to gain from not being interoperable with other systems. I'm not saying Microsoft won't be able to make money from this, but most Linux companies will have a head start in this new market. Above all, they will be in a position where
they don't have a near monopoly advantage. So yes, a new ultra low price PC market will hurt Microsoft in the long run.
You also have to consider that the number of poor people in the world that doesn't allready have a PC, but can afford one if it is priced low enough, most likely is much bigger than rich people that allready have a powerful PC that they replace every three years or so. This means that there will be much more room for expansion in hte low price market than in the high price market segment.
OS-X is very nice for the desktop, but I would stay away from it on the server side, not because there is all that much technically wrong with it, but because Apple don't seam to get the server market.
By hard earned experience with Apple server products I have learned that you can't trust them to support their products over long times. The all of a sudden discontinues products without any resonable migraton paths to the successr, if there even is a successor. E.g. they dicontiued A/UX and replaced with an Apple version of AIX that they then dropped totally in just a couple of years.
When they distribute updates they have more than once totally destroyed, customized settings, and the open source software that comes with the server version of MacOS-X is often incomplete or lacking in functionality compared to the same software on Linux or Solaris.
Chosing between Windows and Mac, I would choose Mac any day. MacOS-X is at least simple to use.
Nobody is surprised that Vista isn't a success (perhaps with the exception of one or two Microsoft employees).
XP did/do the job for most people. So, why upgrade? The only time Vista is worth to consider is if you buy a
new machine. But even then, Vista makes you machine more expensive, both in terms of hardware and software. Then
there is the question if it will work well with your old existing network of XP or even win2k boxes.
Microsoft had the same problem to get people to upgrade from win2k to XP, but XP didn't look like such a total
failure. The reason for that was that there were a lot of win9x users that left that platform for XP. Unlike the win2k users these customers actually got good value for their money, so it was not so hard to make them upgrade.
Another factor is that the competition is much harder now than when they released XP. Apple is starting to get
back in the game, and Linux looks better and better and evolving fast.
Google is in no way dependent of Microsoft, If Microsoft was gone tomorrow people would still find a way to use the web and Google services. If they are dependent of anything it would be good network services. Actually, if Microsoft was gone, it would leave google more room to expand in areas like e-mail handling and office application services.
1. ...No need for any useless login, active desktop, active directory etc.
No login, seams to be a bit of a security risk. A lot of sensitive information could get into the wrong hands
if a somebody misplaces his eee device.
Accessibility is important as it opens new markets for Linux. In many countries this kind of features are required for government use. I would also imagine that not caring for people with disabilities could leave corporate users in some countries open to lawsuits or additional cost when integrating employees that happen to get blind or otherwise disabled during their employment.
I agree with you real usability tests are far better than this, but they are also much more resource intensive to perform. This is why they are not performed as often as one would wish in free software projects.
/etc/, /dev, /proc /lib /bin /sys are rarely used from the GUI, as people who know what to do with them most likely edit them from a terminal using vi or some other cli text editor.
Just consider a simple thing like the use of a file manager.
For one thing, you could get information like, users in general use move file twice as often as they use copy file and 999 time more often than they make symbolik links in konqueror. This could be used in e.g. KDE filemanagers to determine the best order on the "copy here", "move here", "link here", "cancel" popup menu that appears when you drop files over a drop target. (figures just fictional). Or it may turn out that the best thing would be to remove the menu altogheter if one item turn out to be much more used than the others.
You could get information on what directories users opens and edit files in using gui editors.
This could result in new desktop shortcuts or, that some seldom used directories should act like hidden files in the GUI. My suspicion is that
Having information like this may help the Gnome team to not simplify/remove things that people actually usem and the KDE team to simplify/remove things that people almost never use.
To get the most out of this kind of data, I would try to make statistics of the most probable user action given that a certain action just have been performed. And the frequency of each action. Combine this information with the file types if any involved and user serveys and you would probably get a quite good picture on how the GUI is used.
This will give Microsoft valuable information that will help them develop better software in the future. The quesiton is how many users are prepared to trust them enough to allow this. But as it is comleatly volentary I see no problem in doing so, other than that they may get some biased results as people that value privacy and security most likely will not participate. However, personally, I wourld like to be paid a little more than just with some free software to do this kind of testing for Microsoft.
I would actually like to see open source projects such as Gnome and KDE to do a similar thing (As long as it is volentary). Knowing how the users use their is essential to create good usable systems, and very few usability tests are performed on software in the FOSS world, so something like this would probably be even more beneficial to these kind of projects. I would also think that more people would be prepared to volontear this kind information to e.g. the KDE or Gnome team than they would to big evil Microsoft. After all very few Gnome or KDE users would worry about that a big evil company might discover unlicenced software on their systems.
This statement have high probablility to become just as famous as "There will be a total market for five computers in the world" and "nobody will ever need more than 640k memory".
Over this time new algorithms are likely to turn up, and many of the encryption technologies we today consider safe will turn into the digital eqivalent of a simple, easily broken padlock.