That sounds more like the Apple approach, because after buying all the competitors, you effectively control everything, and can then make it work by design.
Personally, I've been expecting Sun to enter the field, what with Java's ubiquity, OpenGL binding, and their massively multi-threaded Niagara chips, I'd think they could come up with a competitive console.
I wasn't trying to claim that any of the three approaches succeeded in making everything "Just Work", I was just illustrating the only three ways it would be possible. Each of the three fails due to a lack in some portion of the implementation.
1.) Either the access point isn't an Apple product, and thereby outside of their control, or there really is a problem in their wifi implementation, which is a failure to adhere to their approach, not a failure of their approach in general.
2.) Microsoft's hegemony isn't universal, and it is certainly not omnipotent. They can get most people to follow their rules, but even those that choose to will not always follow all of the rules, or follow them properly. Again, this is because they don't control everyone, which is a failure to fully implement the approach, not a failure of the approach itself.
3.) You're quite right that not every user wants to hack their system to make it work. The F/OSS implementation is to provide enough users who can and do hack their systems to make it work, and having those users share the fruits of their labor with the rest of the community. Ideally, the manufacturers of hardware and developers of software would become a part of that community, and therefore they would be the "geeks", spreading the use of their product would be their "itch", and the hacking would be done by them to the benefit of their users. Again, the implementation of this approach is not universal, and so it doesn't make everything "Just Work".
There may be other approaches that I've not thought about, but if any of these three were to be successfully implemented, then they could make everything "Just Work". As it is, some organizations have produced various incarnations of these approaches with mixed levels of success, with Apple coming the closest to realizing the ideal of their chosen approach.
Please, don't be one of those guys who preach about open source in a RMS religious zealot style to end users who just want their goddamn iPod to work on their home machine There are 3 ways to make things "Just Work":
1.(The Apple way) Be able to control everything, from the metal to the display. If you can make everything the way you want it to be, you can make things work by design.
2. (The Microsoft way) Be able to contol everyone, from the hardware manufacturers to the software developers. If you can make everybody make things the way you want them to be, you can make things work by fiat.
3. (The F/OSS way) Be able to know everything, from the hardware registers to the software code. If you know everything about the components you use, you can make things work by hacking.
The reason RMS is so adamant about making things free is because we, as end users, have no other way to make them work for us.
While Ive tried Koffice and i quite like gnumeric, there's nothing much interesting in either. Office suites aren't exactly interesting or exciting pieces of software. They're the work-horse of software. That said, OO.o's extension framework may provide some of those exciting features from third-party developers.
I've never used KOffice much, but it was Krita that first prompted me to install it. Krita is a pretty good raster image editor, especially for those who don't like Gimp's layout and menus.
#1: DB2 is a quite capable database, you should have no trouble doing what you want within DB2 itself.
#2: By "on the database itself" I mean in the SQL query itself, or in temporary tables if necessary. You won't need to change the data stored in the database.
#3: SQL was made to do pretty much exactly this. You can sum up fields, grouped by another field, and programatically replace one value with another, depending on the content of yet a third.
#4: Not only are databases "capable" of running searches across multiple tables, this is exactly what any modern RDBMS is designed and optimized to do. There is no scenario I can think of where running such a search on Excel, which has no ability to index data or optimize queries, would be faster to run or easier to configure than a database.
We're querying data out of a database and trying to do simple processing on it (the type that Excel does very well) in the simplest ways we can, and present it to the bosses. What kind of processing are you having to do that can't be done on the database itself?
when someone says they have a reason to use more than X of something in your product, and all it would cost you to give it to them is (I think) changing the types of a bunch of variables, and maybe adding a couple of extra converter methods, you don't tell them, "No one should ever need that many! Only an idiot would even ask for that!" You either say, "Well, we don't currently have enough demand for that feature to be worth the trouble," or you just darn well do it! I'm sure there will be more to change than just that, and probably some unintended consequences of such a change as well.
And not to defend someone who is acting like a stuck up git (I haven't read the quote), chances are that he's right, it sounds like you're using a speadsheet to do the job of a database. When someone tells you you're using a hammer to cut wood, you can't just tell them that it costs them little to put serrated edges on the hammer's head and that they should just darn well do it.
The thing is OO isn't really open source, in that its entirely built and controlled by sun, no community project seams to be interested in making an innovative office tho. Aside from the fact that OO.o is not entirely built by Sun, there is the KOffice suite, and the slightly less cohesive GnomeOffice suite.
So far, are we even sure any laptops of reached the [intended] target audience? If so, have we heard from them? Yes and yes.
A [mesh] is more likely to change day to day. In Africa, there are still a lot of wars going on too. I think keeping a laptop charged would not be something to worry about. The internet is a huge mesh network, and your routes to any given website on the internet will change at least daily, probably more often. And not all of Africa is currently engaged in active warfare, most places have no military violence at all. And even if they were, life goes on during war, and given the nature of the OLPC, they will still work when violence damages all the other infrastructure in the area. Bombs may disable your 120v wall outlet and broadband internet access, but kids with an OLPC still have the airwaves and cows, so they're not affected.
Yes, and the production of work must be completed by a given [date]. You act as if there's no due date. And you act as if outsourced coding is usually ready by the due date. Could their delivery be late? Sure. Will that make them different than outsourcing to any other country? No.
Actually more and more reports of outsourced projects failing have been coming up over the years. Hell, I even have a friend and the company he bailed on ended up out of business because of outsourcing problems; poor code, huge time differences made [communications] almost impossible, and the language barrier ended up creating code that wasn't even close to spec. I never said outsourcing wasn't a bad idea, I said it was currently happening despite all the problems that you're concerned about for African developers.
He works for an [international] company now, and they have these problems as well. I rest my case.
Again, you're making a lot of assumptions here. One that someone won't let the laptop lose power, and two that there WILL be another laptop around to complete the mesh and three that one of those will have network connectivity. Given the laptop's ability to reach other laptops over a kilometer away, and the idea of giving them out to every child in a village, and the ability to charge the laptops any time, by an assortment of devices, the chances of my assumptions being correct are probably pretty good. Good enough for freelance outsourcing anyway.
I can't imagine though trying to do work when my laptop needs to communicate through five others to a small internet connection, that may or may not work. You already have to go through multiple computers to communicate with anything, the difference is that those are (usually) managed by companies rather than individuals.
Well if you're supposed to be a 9 to 5 coder, reliablity is important too. So it's great you can power the laptop, that was never really my point. My point is that it doesn't sound reliable enough to keep your job if you're in one of these small remote villages. They won't start out as 9 to 5 coders, no. Then again, most contracts I've seen for Indian coders were for the production of a given work, not for a range of hours to be worked. Even many American consultant contracts work this way.
We haven't even talked about any language barriers either. That never stopped outsourcing before.
Now I realize the role familiarity is playing here, but it wasn't just that. It's probably not so much a familiarity thing, but rather a difference in concepts. Linux isn't just a different Windows, things that you may think are inherent in computers themselves, like drive letters (C:\) are actually Window-isms, while things like mount points, which you would never have come across on Windows, are pretty ubiquitous on other operating systems.
Another problem you will run into, especially when looking for online support, is that there is no single "Linux" setup. Depending on your disro, and even your individual setup, menu items and programs may be different, or not there at all. Imagine someone trying to write a support article that covered everything from Windows 3.1 to Vista with a single explanation. This is one of the reasons why most help you find online are for the command line, because that is where most Linux systems are most similar. The other reason is that command line instructions are smaller and harder to misunderstand than a laundry list of "go here and click, now go there and click, now go here again and click".
Linux has a steep learning curve if all you're familiar with is Linux. Then again, you'd have the same learning curve going to BSD, Mac OSX, Solaris, or any other non-Windows OS. But once you start learning the basics, like mounting drives, the separation of the GUI from the actual OS, and the like, you will not only be as comfortable in Linux as you are on Windows, you will actually understand what your computer is doing much better, and all that daunting "jargon" will feel as natural to you as "C drive", "Start menu" and "Control Panel" do now.
In this example, why would Exchange hold anyone back? Aren't there alternatives to Exchange? Why not switch to one of the alternatives? The specific scenario I was referring to was switching desktops to Linux. If a company uses Exchange for email/calendaring, there are no Linux clients that integrate as well as Outlook. Sure they could change their server software when they change their desktop software, but that reinforces my point, now you don't just need an alternative to Windows, you need an alternative to Outlook+Exchange as well.
Or are those alternatives simply not as good as Exchange (which is, interestingly enough, a Microsoft product)? There are many alternatives to Exchange that work just as well or better. From packaged suites like Lotus, Zimbra or IceWarp, or you can put together your own choices of Pop3/Imap, WebDAV, CalDAV and LDAP servers. The point is that when deciding whether or not to switch away from Windows, the MS monoculture forces you to have to switch server software too.
Therefore, would it not be correct to say that, in this instance, the business stays with Microsoft because it's actually the best option for their needs? No, it's just re-enforcing my original point, that Microsoft makes it hard to leave, not easy to stay. Exchange is a terrible email/calendaring system, everyone I've seen use it on a large scale has problems with it. But the fact of the matter is that once you've implemented the entire Windows+Outlook+Exchange stack, you can't switch just one piece of it, you have to switch it all.
So it largely depends on if other's in the area are using the laptop.. Actually no, the mesh networking continues to operate as a relay even when the laptop isn't on, using minimal battery power while doing so. All you need is a laptop within range, with some amount of batter power, and you can use it to reach other laptops and/or internet access points.
India is a lot more centralized that many nations in Africa, which from what I understand is where the XO laptops are going. The laptops are being targeted to many countries in Africa, Asia and South America. I believe they are also planning on sending them to some remote areas of Appalachian America. Some of these areas have concentrated populations, some don't. Some have existing power and communication infrastructure, some don't. The OLPC was designed to operate in both situations.
Are you seriously comparing the reliablity of the eletric / communications grid in the US with peddling a bike or relying on cow droppings? I'm comparing the availability of human power to the availability of electrical power. Everywhere you go, human power is available to you. Not so with the grid.
Really? OLPC is setting up wireless / wired network connectivity? Not so much the wired connectivity, but the laptops themselves form a "mesh" network among each other, and the OLPC program is setting up internet access points for them to use. This means that if any one laptop in the "mesh" can access the internet access point, then all other laptops in the "mesh" can access it as well. You can be miles away from the access point, and still get internet connectivity. But even if the internet access isn't available, the laptops themselves can still network amongst each other.
Do you really think a company is going to hire someone, site unseen in a 3rd world nation? It happens all the time, what industry have you been working in? Most likely there will be "consultant" firms, like India's Tata Corp, that will sign up these young coders, then advertise their services to western companies. But heck, even if they only get work off rentacoder.com, it'll still pay them more than they can get in their own country.
How exactly do you get reliable internet connectivity if you need a bicycle to power the laptop? Come on, get real here. How do you reliably get internet connectivity if you need a 120v wall outlet to power your laptop? Seems to me a bicycle is more portable, and definitely more available in those countries. Then again, they also can use a pulley, a solar cell, a cow, and yes, even a 120v wall outlet, plus many more
The problem with die-hard Linux advocates is that they continually insist that the only reason Microsoft is on top is because of marketing. Actually the claim is usually that Microsoft is on top only because they are on top. More specifically, due to the ubiquity of Microsoft's mono-culture, you can't just provide a better alternative to one piece of MS software, you have to be able to replace _all_ of the pieces simultaneously.
Consider the fact that one of the main reasons holding business back from using Linux on the desktop is Exchange, which has absolutely nothing to do with the OS. Or the fact that people don't switch to OpenOffice mainly because of file formats, which have nothing to do with the quality of the software itself.
People don't switch to Linux, not because Windows is better, but because there is some critical piece of their Windows environment that they can't get on Linux (like Photoshop), or because they don't want to change their entire environment just to get the benefits of a better OS. Either way, it is resistance to change, rather than deciding on quality, that keeps people using Windows. If the tables were turned, and Linux, OpenOffice and Firefox currently had 80% market share, and Microsoft was trying to compete with Vista, MS Office and IE, nobody would be switching to them.
I would agree. I had no problem with them using closed-source firmware for their mesh networking, because there wasn't an open-source solution that could provide the same functionality, and the mesh networking was more important than having open-sourced firmware.
That argument doesn't hold up for the OS, however, because there are multiple open-source solutions that can provide the same functionality as Windows or OS X.
Not if you don't have anything of value to start with. Computer equipment costs money, building communications costs money, etc. Both of which are provided by the OLPC laptop.
Tell me... how does one build a data center when you have no capital to buy the equipment and put the infrastructure in place? Why would a C/Java/Python developer in a 3rd world country need a data center, when their employers in the first world will provide that? All they need it the knowledge to do the work, the equipment to do it on, and a way to communicate and transmit the work to their employer. Seems to be that OLPC provides all that.
That's fine, as long as access to the source code, and authorization to freely modify and distribute is one of the requirements.
why? Because without access to the source code, the children can't learn how the OS works. And without the ability to modify and distribute the source code, the children can't use what they learn.
Do you think it really matters to the underprivileged kids in Africa whether their software is free as in speech or free as in beer or just a license to use? Again, you're assuming the point is to give them a computer, it's not, the computer is the means to the end, not the end itself. The purpose of OLPC is to teach these kids about computers, and that means letting them see how it works, and letting them change how it works. Windows isn't going to teach these kids anything.
Publish a list of specs and requirements, and let anyone who can meet them submit their OS. Anything else and you're subsidizing a product that has no competition; that almost always leads to an inferior product. That's fine, as long as access to the source code, and authorization to freely modify and distribute is one of the requirements. If Microsoft makes a product that conforms to that, and works better and/or cheaper than Linux, then use it. If it turns out that BSD or Solaris is a better choice, go with them. It doesn't have to be Linux, it just has to be open.
Huh? How exactly are laptops going to help build up poorer nations? So they can know everything and still not be able to use it because there are virtually no natural resources to use to build an economy? The point is that you can build an economy on information resources, not just physical resources. Take a look at India's explosive economic growth over the last decade.
Anyone posting on this thread should be aware that "gnutoo" is a sockpuppet account of twitter. He's just shilling his own posts to pretend someone agrees with him. Which is ironic, because other people _will_ agree with him because, as trolls go, he at least produces original works that are mostly inline with Slashdot's demographic.
In TFA, Negroponte seems to be saying that countries like Egypt are holding back from potentially massive deployment because the XO doesn't run Windows. For whatever reasons, the customers are demanding Windows. I believe that orders for the XO have been less than hoped, and they're doing this to stay alive. If OLPC has to abandoned their reason for being in order to survive, and it shouldn't matter if they die.
If I founded an organization with a mission to "give free educational books to impoverished kids", and Egyptian authorities said they were only interest in Nancy Drew novels, I wouldn't suddenly change my organization's mission to "getting books in the hands of impoverished kids", I'd tell the Egyptian authorities that I don't sell Nancy Drew novels, and then explain how they will corrupt the morality of their children.
I assume the advantages of no license fees - an already field tested product - and an OS which works remarkably well with the hardware would only become more visible if M$ was allowed to give it a shot. That would require that somebody _sees_ the MS licensing fees or an alternative OS that works remarkably well with the hardware.
If Microsoft gets XP shipped with even _some_ of the laptops, then they will be the ones selling OLPC laptops to 3rd world countries, and you best believe that the license will be bundled with the hardware cost (like it is here in the USA), any performance problems will be blamed on the "cheap" hardware, and nobody will ever be told that there is a better alternative.
Look, M$ wants to try and compete here? I say let them OLPC wasn't founded to give Microsoft a new market to compete in. It was created to give impoverished children access to self-maintaible technology. They made sure that you didn't need an ISP to communicate between laptops. They made sure that you didn't need an AC grid to operate the laptops. They made sure that you didn't need GeekSquad to fix your laptop.
By picking open-source software, then even made sure you didn't need a corporation to fix or improve your software. If they shipped with Windows XP, without it being open-sourced, then they are failing in their objective, because the operation system of the computer could not be maintained by the owner/operator of the computer, but only by Microsoft.
Slashdot Mirror
That sounds more like the Apple approach, because after buying all the competitors, you effectively control everything, and can then make it work by design.
Personally, I've been expecting Sun to enter the field, what with Java's ubiquity, OpenGL binding, and their massively multi-threaded Niagara chips, I'd think they could come up with a competitive console.
I wasn't trying to claim that any of the three approaches succeeded in making everything "Just Work", I was just illustrating the only three ways it would be possible. Each of the three fails due to a lack in some portion of the implementation.
1.) Either the access point isn't an Apple product, and thereby outside of their control, or there really is a problem in their wifi implementation, which is a failure to adhere to their approach, not a failure of their approach in general.
2.) Microsoft's hegemony isn't universal, and it is certainly not omnipotent. They can get most people to follow their rules, but even those that choose to will not always follow all of the rules, or follow them properly. Again, this is because they don't control everyone, which is a failure to fully implement the approach, not a failure of the approach itself.
3.) You're quite right that not every user wants to hack their system to make it work. The F/OSS implementation is to provide enough users who can and do hack their systems to make it work, and having those users share the fruits of their labor with the rest of the community. Ideally, the manufacturers of hardware and developers of software would become a part of that community, and therefore they would be the "geeks", spreading the use of their product would be their "itch", and the hacking would be done by them to the benefit of their users. Again, the implementation of this approach is not universal, and so it doesn't make everything "Just Work".
There may be other approaches that I've not thought about, but if any of these three were to be successfully implemented, then they could make everything "Just Work". As it is, some organizations have produced various incarnations of these approaches with mixed levels of success, with Apple coming the closest to realizing the ideal of their chosen approach.
1.(The Apple way) Be able to control everything, from the metal to the display. If you can make everything the way you want it to be, you can make things work by design.
2. (The Microsoft way) Be able to contol everyone, from the hardware manufacturers to the software developers. If you can make everybody make things the way you want them to be, you can make things work by fiat.
3. (The F/OSS way) Be able to know everything, from the hardware registers to the software code. If you know everything about the components you use, you can make things work by hacking.
The reason RMS is so adamant about making things free is because we, as end users, have no other way to make them work for us.
I've never used KOffice much, but it was Krita that first prompted me to install it. Krita is a pretty good raster image editor, especially for those who don't like Gimp's layout and menus.
#1: DB2 is a quite capable database, you should have no trouble doing what you want within DB2 itself.
#2: By "on the database itself" I mean in the SQL query itself, or in temporary tables if necessary. You won't need to change the data stored in the database.
#3: SQL was made to do pretty much exactly this. You can sum up fields, grouped by another field, and programatically replace one value with another, depending on the content of yet a third.
#4: Not only are databases "capable" of running searches across multiple tables, this is exactly what any modern RDBMS is designed and optimized to do. There is no scenario I can think of where running such a search on Excel, which has no ability to index data or optimize queries, would be faster to run or easier to configure than a database.
And not to defend someone who is acting like a stuck up git (I haven't read the quote), chances are that he's right, it sounds like you're using a speadsheet to do the job of a database. When someone tells you you're using a hammer to cut wood, you can't just tell them that it costs them little to put serrated edges on the hammer's head and that they should just darn well do it.
Another problem you will run into, especially when looking for online support, is that there is no single "Linux" setup. Depending on your disro, and even your individual setup, menu items and programs may be different, or not there at all. Imagine someone trying to write a support article that covered everything from Windows 3.1 to Vista with a single explanation. This is one of the reasons why most help you find online are for the command line, because that is where most Linux systems are most similar. The other reason is that command line instructions are smaller and harder to misunderstand than a laundry list of "go here and click, now go there and click, now go here again and click".
Linux has a steep learning curve if all you're familiar with is Linux. Then again, you'd have the same learning curve going to BSD, Mac OSX, Solaris, or any other non-Windows OS. But once you start learning the basics, like mounting drives, the separation of the GUI from the actual OS, and the like, you will not only be as comfortable in Linux as you are on Windows, you will actually understand what your computer is doing much better, and all that daunting "jargon" will feel as natural to you as "C drive", "Start menu" and "Control Panel" do now.
Consider the fact that one of the main reasons holding business back from using Linux on the desktop is Exchange, which has absolutely nothing to do with the OS. Or the fact that people don't switch to OpenOffice mainly because of file formats, which have nothing to do with the quality of the software itself.
People don't switch to Linux, not because Windows is better, but because there is some critical piece of their Windows environment that they can't get on Linux (like Photoshop), or because they don't want to change their entire environment just to get the benefits of a better OS. Either way, it is resistance to change, rather than deciding on quality, that keeps people using Windows. If the tables were turned, and Linux, OpenOffice and Firefox currently had 80% market share, and Microsoft was trying to compete with Vista, MS Office and IE, nobody would be switching to them.
I would agree. I had no problem with them using closed-source firmware for their mesh networking, because there wasn't an open-source solution that could provide the same functionality, and the mesh networking was more important than having open-sourced firmware.
That argument doesn't hold up for the OS, however, because there are multiple open-source solutions that can provide the same functionality as Windows or OS X.
why? Because without access to the source code, the children can't learn how the OS works. And without the ability to modify and distribute the source code, the children can't use what they learn.
If I founded an organization with a mission to "give free educational books to impoverished kids", and Egyptian authorities said they were only interest in Nancy Drew novels, I wouldn't suddenly change my organization's mission to "getting books in the hands of impoverished kids", I'd tell the Egyptian authorities that I don't sell Nancy Drew novels, and then explain how they will corrupt the morality of their children.
If Microsoft gets XP shipped with even _some_ of the laptops, then they will be the ones selling OLPC laptops to 3rd world countries, and you best believe that the license will be bundled with the hardware cost (like it is here in the USA), any performance problems will be blamed on the "cheap" hardware, and nobody will ever be told that there is a better alternative.
By picking open-source software, then even made sure you didn't need a corporation to fix or improve your software. If they shipped with Windows XP, without it being open-sourced, then they are failing in their objective, because the operation system of the computer could not be maintained by the owner/operator of the computer, but only by Microsoft.