Not all languages are equal. C is a good language to learn, and most likely one to survive in at least few decades -- it's beautiful by design, and allows a programmer to apply any knowledge that he has about the program or a computer.
On the other hand, C++ is a two-headed monster, people who learned it without knowing C first are a danger to society, or at least to everything in it that programs or uses software. Java is, among other things, a colossal monument to few people's egos, opinions and idiosyncrasies, with some OO programming ideas mixed in. Perl... it's useful but one should face it, its design is a mess, and it's getting only messier -- studying such a thing may be useful but better should be done after C. Pascal actually was good for studying because it stands behind the programmer with a huge stick, and beats the programmer mercilessly when it looks like programmer is doing something wrong. Unfortunately exactly the same thing makes Pascal so impractical in a real life, and causes people to be so confused when they switch to C, it isn't worth the trouble.
IBM did disappear. And so did AT&T -- what now carries those names can be only describe as corpses of the companies, with neither ambitions, influence or abilities of the original companies.
Wrong. Market has absolutely nothing to do with cost of operations -- if it was the case, no branch of industry or technology will be extinct by now, and you would be able to buy stone axes and bone fish hooks.
Monopoly can charge unreasonable price and still spend all those money on continuing "operations" that include unreasonable luxury, rampant investment around, buying the competition and other spendings that would be unnecessary to produce the product but are necessary to make monopoly maintenance worthwhile.
It's hardware -- either use overlay-incapable VGA and capture image produced by a regulae DVD player, or use TV output and a capture card -- with good enough equipment the only loss would be reduced bit depth. All the decoding would be done by perfectly licensed decoder in a player.
A few machines making millions of copies, or millions of machines making a few copies still leads you to a situation in which millions of copies exist.
The difference is, on what market those copies will be. "Industrial" bootleggers sell to people who are likely to spend money on DVDs. Homemade copies are usually given to friends that would either get that, or borrow the disk but definitely won't rush to the store to buy a DVD even if there were no homemade copies.
Could you provide a link to something that backs up your claim that consumer drives can make bit-for-bit copies?
I have never claimed that. Consumer equipment however can do re-capture/re-encoding of video (re-encoding is necessary because consumer media isn't _large_ enough to store DVD's equivalent -- again, with or without DeCSS).
Repetition while logged in doesn't make flase things true. Consumer-grade hardware is physically incapable of producing large enough number of copies to affect anything -- only either large professional bootleg setups (that definitely can copy the disks bit for bit, re-encode decrypted movie at the CRT, etc.) or copying over the network. First ones don't depend on any software or licenses, second gives a good idea what a "fair" price of the thing is -- no one is going to waste hours downloading things that can be bought cheap in the store. So really it all comes to the same thing -- movie industry jacking up the prices.
could easily just use exported data - but I really like having the data linked live. While working on the report the charts are updated in near realtime from the original SQL queries. That means when I print that report in 1200 dpi color glory the data is within five minutes of being accurate.
Is this the best way? Well, probably not. I had some nice little scripts that produced raw stats and dumped them to a text file, but you know how management is - they want charts and colors and exploding pie pieces.
If you can get a data from a database, you DEFINITELY can immediately generate those graphs. This would be so far superior to copying graphs from a spreadsheet, it's not even funny.
In the Windows world, I can switch between any number of apps that handle my data rather well (though I admit its far from perfect). Exporting to flat files, then re-importing will indeed work, but its far from real-time and its far from flexible in certain cases.
If you have data in the database you don't even have to export it -- there are a lot of tools that will do that directly from the datbase on the fly, and they are easier to use than to make your spreadsheet. In your "Windows world" you simply follow one rule -- to do everything by a small set of huge overbloated programs by tinkering with their endless "features", even if it doesn't make any sense, and requires constant repeating of simple but time-consuming procedures. Unix users don't follow this philosophy -- they would rather get a proper tool, and make things work once and for all, so every report can be generated by just clicking on a button in a web form -- then your boss wouldn't even have to ask you for those reports, if this is the goal, or you can email him a link to the automatically generated report about whatever you wish him to look at. But no, in "Windows world" it's all about documents being pushed between desks, be it through paper, email or telepathy if one will be invented -- the idea is that amount of busywork is more important than the result or even convenience -- the more amount of mouse-clicking is involved, the better, and the only alternative the user sees is "exporting to text" that he is trained to be afraid of.
You have a vulnerabilities in a spreadsheet and have graphs and averages in it? Where does the data come from, does the spreadsheet by itself try to exploit your boxes? And if not so, can't you just re-generate everything automatically when making a document, using whatever source that stores the data -- and make graphs suitable for the document, what unlikely to be the same as what you have in a spreadsheet?
Or is it that you actually use spreadsheet to keep all the data (data collection) _and_ do all the graphs (presentation), _and_ keep all the descriptions (secure long-term "write-only" storage)? That's a poor handling of information, and the behavior of the mouse is the least of your problems. Personally I would just write a simple script in perl or php, and keep data where it belongs -- in databases, with properly assigned permissions, and presentation where it belongs -- in HTML, handled by few simple scripts.
You EXPORT it. Into a file. With all the options that you need. Then you switch to another program and insert that file into its document. Same functionality, one more click, much more flexibility.
1. "Windows clustering" AFAIK, only includes load-balancing and some transactions processing, designed for web servers primarily. No one uses it though because application-specific designs are usually better. Cluster management also exists, but I don't think, it's helpful in general.
2. There probably won't be much trouble to put MPI or PVM on a bunch of Windows boxes, and run something clustered. It's however unlikely that this thing will be easy to manage, and certainly the performance is likely to suck because Windows networking was "optimized" in various ways that make sense in a, say, web or mail server, but offer no help otherwise.
3. Windows has no equivalent of MOSIX or other advanced clustering architectures, and messiness of its design doesn't give much hope for it.
who has realized that today's way of computing is doomed and will be taken over by a new, more distributed way.
It was "more distributed" than anything you ever knew on win32 for decades already -- with technologies that have absolutely nothing to do with DCOM and.NET, and especially a lunacy that RPC and SOAP are. As opposed to Microsoft-world where you live, technologies don't appear because an ugly guy and a fat guy give everybody else an API to implement them. Technologies just appear when someone thinks, he can implement them, and distributed computing, done in ways much more flexible and efficient that nearsighted "frameworks" from Microsoft and Sun, is being used right now by many people while Windows monkeys were getting excited every two years after hearing each Microsoft announcement that the happy days of automagically working distributed object framework's days are near, and everybody should better learn COM sooner because this is where they are going to put it.
If that's what you want, feel free to work on it, or to pay someone to work on it. Don't feel free to tell volunteers where they can and cannot put their time and effort. It's rude and stupid.
"Volunteers" happen to care about what is being said about their ideas -- after all the whole project is supposed to be implemented by a large number of people, one that express their negative opinion included.
Don't like a product? Don't use it. Simple as that.
If Gnome just was a closed system and a product of some company that was developing it with its own resources, it would be reasonable. If it tries to be a framework that many people already use, they have a right to demand that carpet will not be pulled from under them in mid-development. Many people will have to make a decision if their programs will be made based on or compatible with Gnome, and it will be only fair to warn Miguel and other how their decisions would affect that. For example, if Gnome would require JVM or a even a single DLL that must be compiled with Microsoft tools, or will demand some hideously convoluted "OO framework", such as COM, XPCOM, DCOM and especially.NET support from applications just to perform their basic functionality and configuration, or any of that bullshit in session manager or mandatory pieces of its framework, I promise to never develop anything for it, never use it again, and delete it from all my computers despite the fact that now I am using it exclusively.
And I certainly feel that I have a right to tell that to Miguel now instead of doing it later.
Trying to tell someone that doesn't give a crap about "Desktops" that they must work on one only works when you are willing to pay them a salary.
And probably no one should tell anything to politicians unless with regularly scheduled bribes? Gnome would be an unremarkable piece of shit if it wasn't a project supported by a large community. Look at CDE -- that's your "Gnome" made by a bunch of shit-headed companies (including such "intellectual giants" as HP) over, I think, decades of hard work. That's what you get when you proudly work in a vacuum, and all your developers care so little about what others think about their code that they probably write that code while sitting in Exceed or possibly even in Devstudio. That's what you get when "screw all the users!" is an acceptable attitude. This is what happen when the only recouces are 1. to beat up the developers and 2. start your own project.
Gnome is one of the projects that started exactly because of this (and Trolls showing just a little bit of similar attitude at the time, but this is unimportant). So unless someone wants to repeat the history again, it's better to express the displeasure at Miguel's ideological decisions now than to have to fork the project, driving Ximian out of business in the process.
Why would anyone want to do such a thing? It's like trying to "copy" AutoCAD drawing into Photoshop -- sure, one may implement it, and maybe AutoCAD already does that, however applications that use drastically different ways to represent their data must somehow force naive user to tell them how the data is supposed to be converted instead of frustrating everyone else with idiotic defaults.
"Elektronika"/"SM" line -- clone of PDP-11 line, often with some creative changes (high-density floppies, graphics controlers on a second PDP-11 CPU), then some VAXen
"DWK"/"UKNC" line -- same as "SM", but made as a desktop. "DWK" models 3 and 4 were built as a single unit with terminal (keyboard was separate), "UKNC" was a very nice flat box with builtin keyboard and extension connectors at the top, connected to a separate monitor.
"BK-0010" -- can be described as a PDP-11 squeezed into Sinclair's case, everything was in the keyboard, with TV output, tape recorder connector, and on some models a serial port.
"Elektronika-85" -- Dec Pro/350 clone. Was hated just as much as its prototype.
"ES-1840","Iskra-1030" lines -- IBM PC clones, usually with some changes. Appeared in early 90's and soon were replaced by conventional PC clones.
"Radio-86RK","Specialist" -- hobbyist 8080-based boxes, never were mass-produced but popular among various computer enthusiasts.
"Sinclair" clones
There were some others, however I have mentioned the most popular ones.
Copyrights, trademarks and patents are supposed to be ortogonal to each other and supposed to be unrelated parts of "intellectual property" law that may apply to the same issue but deal with different aspects of it (one can infringe on all of them at the same time, yet issues raised by each infringement should be resolved separately). Mixing them together or justifying one by another being protected in some likely situation is a kind of legal illiteracy.
I don't know, my projector is rather limited, and I would prefer film over aanything that iy can produce... Maybe film projectors' power is limited by air-cooling of the film, and colors are limited by the need to have all pigments on all frames, and under threat of fading, while LCDs are easier to cool and can use any reasonable way to get red, blue and green as close to the optimal as one would want...
I have a "digital projector" with a modest 1024x768 resolution projected on a 112x84" (2.8x2.1M) wall, and since I used old LCD panel from ebay, it has rather limited contrast and color. Still the whole thing (Proxima Ovation 944+ panel and 3M 9700 overhead projector) was less than $1000.
Modern projectors have little excuse for being expensive -- separate red, green and blue panels are easier in production, and less prone to "fallen out pixels".
The problem I see with this article is...
on
What is .NET?
·
· Score: 2
...not that it's inaccurate technically when it describes the particular technology, but that it tries to claim that what it describes is really ".NET". In fact ".NET" is whatever Microsoft calls ".NET", and at this point in includes CLR, framework, and existing implementation that is infested with Microsoft "technologies" tie-ins (COM, VB, C#) at the extent that, among the other properties, the whole thing absolutely certainly is unportable.
Another problem is, ".NET", and in fact,.NET framework, is being compared to Java, as if Java is the only thing that opposes it. My opinion is that Java sucks in its own right, and the only alternative that I accept is the lack of "framework" until the time when people will learn how to use existing tools, and develop models that actually produce a framework that benefits developer, and not just locks him into something that framework's authors think, will benefit them, or pamper their idiosyncrasies. My opinion on "frameworks" that exist today is here, and it fully applies to this article. There are millions of possible ways to build a very complex sand castle in software -- Java, COM, SOAP,.NET are only few of them, and without any doubt people will invent more. The problem is, no one needs sand castles, people need something they can live in, and kids on the beach need to grow up before they will be able to build houses. And some mentally deficient and whiny kids, such as Microsoft, are better kept away not only from building houses but sand castles as well.
The question I ask then is, does anyone know of application 'quotas' or anything similar going on? Is it their policy to let a certain %% through, give or take, or is it just the fact that like it or not, and stupid and BAD ideas or not, the same percentage which meet the necessary patent requirements hasn't changed in 5 years.
Stupid and bad ideas still can be patented, the only requirement is describing a distinct device or process that did not exist before. The problem is, no one knows how "idea" in general is different from "device" or "process", and where algorithms and mathematical formulas fit between those things.
...that IP rights are supposed to be balanced with "free market" (what I consider to be a political doctrine on the border of being a religious belief) and not consumer protection, freedom of expression, advance of technology, science and human thought, or other real things that are threatened by overbroad patents and other kinds IP abuse.
Slashdot Mirror
1. None of that has anything to do with "social skills".
2. I have never said that a lot of managers in "corporate world" have any social skills to begin with.
Microsoft employees are certainly active today.
If you make your hiring decisions based on social skills, you don't have them.
Not all languages are equal. C is a good language to learn, and most likely one to survive in at least few decades -- it's beautiful by design, and allows a programmer to apply any knowledge that he has about the program or a computer.
On the other hand, C++ is a two-headed monster, people who learned it without knowing C first are a danger to society, or at least to everything in it that programs or uses software. Java is, among other things, a colossal monument to few people's egos, opinions and idiosyncrasies, with some OO programming ideas mixed in. Perl... it's useful but one should face it, its design is a mess, and it's getting only messier -- studying such a thing may be useful but better should be done after C. Pascal actually was good for studying because it stands behind the programmer with a huge stick, and beats the programmer mercilessly when it looks like programmer is doing something wrong. Unfortunately exactly the same thing makes Pascal so impractical in a real life, and causes people to be so confused when they switch to C, it isn't worth the trouble.
IBM did disappear. And so did AT&T -- what now carries those names can be only describe as corpses of the companies, with neither ambitions, influence or abilities of the original companies.
Wrong. Market has absolutely nothing to do with cost of operations -- if it was the case, no branch of industry or technology will be extinct by now, and you would be able to buy stone axes and bone fish hooks.
Monopoly can charge unreasonable price and still spend all those money on continuing "operations" that include unreasonable luxury, rampant investment around, buying the competition and other spendings that would be unnecessary to produce the product but are necessary to make monopoly maintenance worthwhile.
It's hardware -- either use overlay-incapable VGA and capture image produced by a regulae DVD player, or use TV output and a capture card -- with good enough equipment the only loss would be reduced bit depth. All the decoding would be done by perfectly licensed decoder in a player.
A few machines making millions of copies, or millions of machines making a few copies still leads you to a situation in which millions of copies exist.
The difference is, on what market those copies will be. "Industrial" bootleggers sell to people who are likely to spend money on DVDs. Homemade copies are usually given to friends that would either get that, or borrow the disk but definitely won't rush to the store to buy a DVD even if there were no homemade copies.
Could you provide a link to something that backs up your claim that consumer drives can make bit-for-bit copies?
I have never claimed that. Consumer equipment however can do re-capture/re-encoding of video (re-encoding is necessary because consumer media isn't _large_ enough to store DVD's equivalent -- again, with or without DeCSS).
Repetition while logged in doesn't make flase things true. Consumer-grade hardware is physically incapable of producing large enough number of copies to affect anything -- only either large professional bootleg setups (that definitely can copy the disks bit for bit, re-encode decrypted movie at the CRT, etc.) or copying over the network. First ones don't depend on any software or licenses, second gives a good idea what a "fair" price of the thing is -- no one is going to waste hours downloading things that can be bought cheap in the store. So really it all comes to the same thing -- movie industry jacking up the prices.
could easily just use exported data - but I really like having the data linked live. While working on the report the charts are updated in near realtime from the original SQL queries. That means when I print that report in 1200 dpi color glory the data is within five minutes of being accurate. Is this the best way? Well, probably not. I had some nice little scripts that produced raw stats and dumped them to a text file, but you know how management is - they want charts and colors and exploding pie pieces.
If you can get a data from a database, you DEFINITELY can immediately generate those graphs. This would be so far superior to copying graphs from a spreadsheet, it's not even funny.
In the Windows world, I can switch between any number of apps that handle my data rather well (though I admit its far from perfect). Exporting to flat files, then re-importing will indeed work, but its far from real-time and its far from flexible in certain cases.
If you have data in the database you don't even have to export it -- there are a lot of tools that will do that directly from the datbase on the fly, and they are easier to use than to make your spreadsheet. In your "Windows world" you simply follow one rule -- to do everything by a small set of huge overbloated programs by tinkering with their endless "features", even if it doesn't make any sense, and requires constant repeating of simple but time-consuming procedures. Unix users don't follow this philosophy -- they would rather get a proper tool, and make things work once and for all, so every report can be generated by just clicking on a button in a web form -- then your boss wouldn't even have to ask you for those reports, if this is the goal, or you can email him a link to the automatically generated report about whatever you wish him to look at. But no, in "Windows world" it's all about documents being pushed between desks, be it through paper, email or telepathy if one will be invented -- the idea is that amount of busywork is more important than the result or even convenience -- the more amount of mouse-clicking is involved, the better, and the only alternative the user sees is "exporting to text" that he is trained to be afraid of.
You have a vulnerabilities in a spreadsheet and have graphs and averages in it? Where does the data come from, does the spreadsheet by itself try to exploit your boxes? And if not so, can't you just re-generate everything automatically when making a document, using whatever source that stores the data -- and make graphs suitable for the document, what unlikely to be the same as what you have in a spreadsheet?
Or is it that you actually use spreadsheet to keep all the data (data collection) _and_ do all the graphs (presentation), _and_ keep all the descriptions (secure long-term "write-only" storage)? That's a poor handling of information, and the behavior of the mouse is the least of your problems. Personally I would just write a simple script in perl or php, and keep data where it belongs -- in databases, with properly assigned permissions, and presentation where it belongs -- in HTML, handled by few simple scripts.
You EXPORT it. Into a file. With all the options that you need. Then you switch to another program and insert that file into its document. Same functionality, one more click, much more flexibility.
1. "Windows clustering" AFAIK, only includes load-balancing and some transactions processing, designed for web servers primarily. No one uses it though because application-specific designs are usually better. Cluster management also exists, but I don't think, it's helpful in general.
2. There probably won't be much trouble to put MPI or PVM on a bunch of Windows boxes, and run something clustered. It's however unlikely that this thing will be easy to manage, and certainly the performance is likely to suck because Windows networking was "optimized" in various ways that make sense in a, say, web or mail server, but offer no help otherwise.
3. Windows has no equivalent of MOSIX or other advanced clustering architectures, and messiness of its design doesn't give much hope for it.
who has realized that today's way of computing is doomed and will be taken over by a new, more distributed way.
It was "more distributed" than anything you ever knew on win32 for decades already -- with technologies that have absolutely nothing to do with DCOM and .NET, and especially a lunacy that RPC and SOAP are. As opposed to Microsoft-world where you live, technologies don't appear because an ugly guy and a fat guy give everybody else an API to implement them. Technologies just appear when someone thinks, he can implement them, and distributed computing, done in ways much more flexible and efficient that nearsighted "frameworks" from Microsoft and Sun, is being used right now by many people while Windows monkeys were getting excited every two years after hearing each Microsoft announcement that the happy days of automagically working distributed object framework's days are near, and everybody should better learn COM sooner because this is where they are going to put it.
Java was supposed to be platform-independent, too -- and, as opposed to MS, Sun actually wanted that to happen. http://www.dialpad.com anyone?
Personally I know only one true cross-platform thing -- properly written C or C++ source.
If that's what you want, feel free to work on it, or to pay someone to work on it. Don't feel free to tell volunteers where they can and cannot put their time and effort. It's rude and stupid.
"Volunteers" happen to care about what is being said about their ideas -- after all the whole project is supposed to be implemented by a large number of people, one that express their negative opinion included.
Don't like a product? Don't use it. Simple as that.
If Gnome just was a closed system and a product of some company that was developing it with its own resources, it would be reasonable. If it tries to be a framework that many people already use, they have a right to demand that carpet will not be pulled from under them in mid-development. Many people will have to make a decision if their programs will be made based on or compatible with Gnome, and it will be only fair to warn Miguel and other how their decisions would affect that. For example, if Gnome would require JVM or a even a single DLL that must be compiled with Microsoft tools, or will demand some hideously convoluted "OO framework", such as COM, XPCOM, DCOM and especially .NET support from applications just to perform their basic functionality and configuration, or any of that bullshit in session manager or mandatory pieces of its framework, I promise to never develop anything for it, never use it again, and delete it from all my computers despite the fact that now I am using it exclusively.
And I certainly feel that I have a right to tell that to Miguel now instead of doing it later.
Trying to tell someone that doesn't give a crap about "Desktops" that they must work on one only works when you are willing to pay them a salary.
And probably no one should tell anything to politicians unless with regularly scheduled bribes? Gnome would be an unremarkable piece of shit if it wasn't a project supported by a large community. Look at CDE -- that's your "Gnome" made by a bunch of shit-headed companies (including such "intellectual giants" as HP) over, I think, decades of hard work. That's what you get when you proudly work in a vacuum, and all your developers care so little about what others think about their code that they probably write that code while sitting in Exceed or possibly even in Devstudio. That's what you get when "screw all the users!" is an acceptable attitude. This is what happen when the only recouces are 1. to beat up the developers and 2. start your own project.
Gnome is one of the projects that started exactly because of this (and Trolls showing just a little bit of similar attitude at the time, but this is unimportant). So unless someone wants to repeat the history again, it's better to express the displeasure at Miguel's ideological decisions now than to have to fork the project, driving Ximian out of business in the process.
Why would anyone want to do such a thing? It's like trying to "copy" AutoCAD drawing into Photoshop -- sure, one may implement it, and maybe AutoCAD already does that, however applications that use drastically different ways to represent their data must somehow force naive user to tell them how the data is supposed to be converted instead of frustrating everyone else with idiotic defaults.
There were some others, however I have mentioned the most popular ones.
Copyrights, trademarks and patents are supposed to be ortogonal to each other and supposed to be unrelated parts of "intellectual property" law that may apply to the same issue but deal with different aspects of it (one can infringe on all of them at the same time, yet issues raised by each infringement should be resolved separately). Mixing them together or justifying one by another being protected in some likely situation is a kind of legal illiteracy.
I don't know, my projector is rather limited, and I would prefer film over aanything that iy can produce... Maybe film projectors' power is limited by air-cooling of the film, and colors are limited by the need to have all pigments on all frames, and under threat of fading, while LCDs are easier to cool and can use any reasonable way to get red, blue and green as close to the optimal as one would want...
I have a "digital projector" with a modest 1024x768 resolution projected on a 112x84" (2.8x2.1M) wall, and since I used old LCD panel from ebay, it has rather limited contrast and color. Still the whole thing (Proxima Ovation 944+ panel and 3M 9700 overhead projector) was less than $1000.
Modern projectors have little excuse for being expensive -- separate red, green and blue panels are easier in production, and less prone to "fallen out pixels".
...not that it's inaccurate technically when it describes the particular technology, but that it tries to claim that what it describes is really ".NET". In fact ".NET" is whatever Microsoft calls ".NET", and at this point in includes CLR, framework, and existing implementation that is infested with Microsoft "technologies" tie-ins (COM, VB, C#) at the extent that, among the other properties, the whole thing absolutely certainly is unportable.
Another problem is, ".NET", and in fact, .NET framework, is being compared to Java, as if Java is the only thing that opposes it. My opinion is that Java sucks in its own right, and the only alternative that I accept is the lack of "framework" until the time when people will learn how to use existing tools, and develop models that actually produce a framework that benefits developer, and not just locks him into something that framework's authors think, will benefit them, or pamper their idiosyncrasies. My opinion on "frameworks" that exist today is here, and it fully applies to this article. There are millions of possible ways to build a very complex sand castle in software -- Java, COM, SOAP, .NET are only few of them, and without any doubt people will invent more. The problem is, no one needs sand castles, people need something they can live in, and kids on the beach need to grow up before they will be able to build houses. And some mentally deficient and whiny kids, such as Microsoft, are better kept away not only from building houses but sand castles as well.
The question I ask then is, does anyone know of application 'quotas' or anything similar going on? Is it their policy to let a certain %% through, give or take, or is it just the fact that like it or not, and stupid and BAD ideas or not, the same percentage which meet the necessary patent requirements hasn't changed in 5 years.
Stupid and bad ideas still can be patented, the only requirement is describing a distinct device or process that did not exist before. The problem is, no one knows how "idea" in general is different from "device" or "process", and where algorithms and mathematical formulas fit between those things.
...that IP rights are supposed to be balanced with "free market" (what I consider to be a political doctrine on the border of being a religious belief) and not consumer protection, freedom of expression, advance of technology, science and human thought, or other real things that are threatened by overbroad patents and other kinds IP abuse.
IBM is a dead company -- it works as a amorphous conglomerate with no functioning central management.
Microsoft is not a company, it's more like a mob, or cult.