Except that doesn't always happen on Vista. My Sony mp3 player simply pops up an explorer window, as does my phone when I plug it in to charge it (I doubt I clicked 'always do this action' as that never seems to work when I do want it, and for the phone, last thing I want is to use it as a memory stick)
Businesses always want to charge as much as possible; a competitive marketplace always wants to make them charge as little as possible.
By releasing your code as OS, even under a BSD licence, means that the entry barriers to entering the marketplace are greatly reduced, thus increasing competition and reducing prices overall. This still applies if a company takes your freely-given code and makes a proprietary product with it, either they compete with other similar proprietary products (thus reducing prices), or they compete with a free version. Either way, the end users win.
In some cases, allowing a business to be created to sell a proprietary product using the free software can be more beneficial than a free one - mainly because the company will develop it much further than volunteers would. Obviously this applies more to smaller markets, or ones where there are few volunteers to develop for the free software, in these places a company would probably not start up at all unless they had a more guaranteed way of realising the revenue needed to develop it.
free under this GPL thingy, or pay us $$$ and do whatever you want with it
Yep, because that worked so well for increasing QT adoption in companies.
I understand what he's talking about - you do want the OSS to spread, then the GPL takes it only so far (to other OSS products), if you want it to spread everywhere and become a standard, then you need a BSD-style licence. I'm sure Microsoft wouldn't have used the tcp/ip stack if it was GPL, and they'd have created their own network protocol instead. Fortunately they got to 'steal' that code and we've all benefited.
Now open source is reaching a level of maturity in the marketplace, and in end-user's minds, its time to start thinking of reducing the amount of GPL licences out there, whilst you needed GPL to get us to this point, we need LGPL or BSD to take us all the way to global software domination.
// I paid my $4 and all I got was this lousy comment
How many LOC are there in the project? I feel an alternative to code golf coming up - how to write a very short, tiny, simple task in as many lines of code as possible. Some outsourced developers I know stand a very good chance of winning.
you forgot the biggest RAM consumer in managed code: the user.
I've found that everyone who has been taught "GC means you don't have to worry about RAM anymore" really take it to heart, so appending to a string means you end up with 3 temporary copies, but its ok, the GC will clean them up (and RAM is cheap, remember!), reading from a DB takes up so much RAM its crazy - probably XML formatting the recordset with many, many throw-away instances of all that string data.
As people write libraries that think like this, and others use them, and write code like it, all the little bits of memory that is left for the GC add up to a tremendous amount. Take a look at perfmon sometime and see how many collections the GC does per second for an idea of the work going on under that managed cover providing you with an 'easy to use' environment. That's why they won't write the OS using it.
Looks like SoftwareForge isn't FOSS, but Collabnet's TeamForge (which is fair enough, but disappointing as I wanted to run military open-source software, just to tell my boss that if its good enough for the DoD, its good enough for us!
it works well with YUM too - in fact, Webmin is one of the best admin things around. I think every project should be mandated to create a webmin administrative module before being allowed into the wild:)
Yeah, it's crazy that book authors should want to make a few bucks to eat and pay rent. They're just greedy bastards.
yep, but how much of a book's cover price goes to the author? That's the problem here - not that an author deserves everything he could get for his book (often that's nothing, too many books have tiny, insignificant sales), but that the publisher gets very rich on their back.
Self-publishing through something like the book-on-demand would be a very good thing, if it lowered the cost of publishing a book, and cut out the publisher.
yeah, like ethernet and tcp/ip never caught on in a unified manner.
The problem is not the standard (though overly complicated ones as designed byt he committes that submit them to the standards bodies certainly don't help), but the individual companies (or company) that decides to make a few changes to suit itself and doesn't care about interoperability with everyone else. The "de facto" standard caused by monopoly abuse.
"just works", like Windows you mean. (see, I posted about this the other day, the "just works" meme has been dragged up to kick Linux at every opportunity).
What you don't realise is that it does just work, the default installation on Ubuntu has had all that choice stuff taken away from you so you get a single, supported set of options. If you don't like any of the options you can swap them to something else (Kubuntu for example), yet the original install that those 90% of users get works, pure and simple.
As a corporate, you get the same power - the IT department can make up a distro of the bits they think works best and ship it out to all their users. They get the power and flexibility of being able to do that, yet still get a single install that those 90% of their users can work with, without any hassles.
For example, they were talking about "Cairo" (which was like V+3) long before Windows 95 showed up. It was really going to blow away NextStep!
And it did! Shame they had to wait until processing power had gone up ten times, graphics power up a hundred times, and ram capacity increased 1000 times!
Just think, they might produce something that blows Compiz away in another 5 years of hardware advances!
I'm not so sure you've been right that far back though.
I once read that computers would be like aircraft - first we had fragile things made of wood, then the state of the art got better and better over time, until you ended up with 2 airplane models: the 747 bulk carrier for cheap, flights carrying lots of people, and Concorde which was a luxury, high-end, performance model for those who had far too much money.
Computers are reaching the same plateau - you get to a point where your PC is good enough for everything you need it to do, and you don't bother spending the extra to upgrade, unlike times past when a new CPU model or OS was an instant best seller. You can see this happening right now with XP and core2duo. People don't see the need for Vista, or for quad cores.
When MS comes out with Windows 7, and Intel make a 10-core CPU, no-one will buy them. (well, ok, they may well buy W7 because MS will kill off XP and give you no choice in the matter).
So yeah, I rank "just works" up there with TCO. It says little about what kind of experience I can expect.
which is exactly why the marketeers use the term, its like panning the camera away from the attack scene, your imagination fills in far more than those 2 words could ever actually mean.
Then someone else will come along and make a gui that abstracts that command interface, in a way that is much better than anything the guys who wrote the cli-based system would have made.
I think something like subversion is a case in point, excellent cmdline tool, some excellent guis too, many available for environments the svn devs have no interest or skills in.
lol. Slashdot has always been like this. The mainstream media noticed us and ignored us years ago.
I've noticed a lot of posters relatively recently that are popping up and basically saying "linux is not ready yet, until you plug it in and it 'just works' it won't be ready", either implying Windows does "just work" or explicitly stating it. I know no computer does that, there's niggles in everything, but I seem to hear that mantra more often than I ever did.
Maybe you havn't been paying attention to them, but they're there.
There are a lot of pro-MS postings, I've done them myself, but they tend to be more objective against trolls saying Linux is perfect at everything and Windows couldn't possibly be any good. Windows is a perfectly usable OS, I just consider Linux to be architecturally better and has the potential to be significantly better.
Thing is, Ballmer (and I doubt it is him directly, probably more fortuitous circumstance for MS that they had a research project out of J++ that came good) is doing more anti-competitive moves with the new technology. Its just now, they can push.NET massively knowing that all.net programmers are tied securely into Windows for ever.
It isn't a nasty anti-competitive move, but it is designed to keep Windows marketshare at other platform's expense. (jury's out whether it will become nasty if Mono comes good, and gets lots of market share, and Microsoft loses developers to it)(but it seems that won't happen while there are plenty of bits that are not supported in Mono).
This is the kind of thing that makes a difference. 70 million euro is a lot of money.
Also, note the quote at the end, usually referred to as a cost of migrating to Linux.
Moving from Microsoft XP to Vista would not have brought us many advantages and Microsoft said it would require training of users.
One day we will have a replacement for Exchange, Thunderbird will have groupware features built-in (or plugged in) and MS will have serious competition on their hands. Or possibly will heat up the marketing war.
With Linux this in general does not work because the underlying dependencies change too often and too much.
You've obviously not built anything that requires the Visual Studio 2005 service pack 1 redistributable. And MS now considers it ok to break old apps (ie require a recompile) for 2008 too.
When I upgraded to Vista, I found a few apps that wouldn't work - and quite a few drivers that suddenly weren't supported and weren't ever going to be. If I wanted those devices to work, I had to buy a new one.
quite true, so now what does this mean for 'unbreakable' Linux? Will it be slowly dropped, or kept around as the x86 alternative to Solaris? In the datacentre, I'm certain Solaris will be the old new OS for Oracle... what about smaller customers, I don't think Solaris for x86 ever really got anywhere, definitely not compared to Linux.
They have a couple of options -- either (A) hold the schools to account for what they're doing and not doing
the biggest issue here is that school are allowing kids access to the freaking internet. I'm sure none of the kids there give a damn about any gay/lesbian website - they're too busy talking crap with their mates on facebook. Instead of learning stuff.
So yeah, sure we should be outraged at some faceless company deciding what's allowable or not on the internet, but we should be equally outraged that kids have access to all the rest of the internet whilst at school.
now, I'm going to go back to surfing those sites my employer deems acceptable for me to waste... sorry, "profitably leverage my skills in a proactive self-learning manner" on.
not just that, but Samsung has also got Yahoo widgets and internet-access capabilities built into its new range of TVs. I leave it to the reader to dsecide if that's a good thing or not, but its a start.
Slashdot Mirror
Except that doesn't always happen on Vista. My Sony mp3 player simply pops up an explorer window, as does my phone when I plug it in to charge it (I doubt I clicked 'always do this action' as that never seems to work when I do want it, and for the phone, last thing I want is to use it as a memory stick)
Businesses always want to charge as much as possible; a competitive marketplace always wants to make them charge as little as possible.
By releasing your code as OS, even under a BSD licence, means that the entry barriers to entering the marketplace are greatly reduced, thus increasing competition and reducing prices overall. This still applies if a company takes your freely-given code and makes a proprietary product with it, either they compete with other similar proprietary products (thus reducing prices), or they compete with a free version. Either way, the end users win.
In some cases, allowing a business to be created to sell a proprietary product using the free software can be more beneficial than a free one - mainly because the company will develop it much further than volunteers would. Obviously this applies more to smaller markets, or ones where there are few volunteers to develop for the free software, in these places a company would probably not start up at all unless they had a more guaranteed way of realising the revenue needed to develop it.
free under this GPL thingy, or pay us $$$ and do whatever you want with it
Yep, because that worked so well for increasing QT adoption in companies.
I understand what he's talking about - you do want the OSS to spread, then the GPL takes it only so far (to other OSS products), if you want it to spread everywhere and become a standard, then you need a BSD-style licence. I'm sure Microsoft wouldn't have used the tcp/ip stack if it was GPL, and they'd have created their own network protocol instead. Fortunately they got to 'steal' that code and we've all benefited.
Now open source is reaching a level of maturity in the marketplace, and in end-user's minds, its time to start thinking of reducing the amount of GPL licences out there, whilst you needed GPL to get us to this point, we need LGPL or BSD to take us all the way to global software domination.
// I paid my $4 and all I got was this lousy comment
How many LOC are there in the project? I feel an alternative to code golf coming up - how to write a very short, tiny, simple task in as many lines of code as possible. Some outsourced developers I know stand a very good chance of winning.
you forgot the biggest RAM consumer in managed code: the user.
I've found that everyone who has been taught "GC means you don't have to worry about RAM anymore" really take it to heart, so appending to a string means you end up with 3 temporary copies, but its ok, the GC will clean them up (and RAM is cheap, remember!), reading from a DB takes up so much RAM its crazy - probably XML formatting the recordset with many, many throw-away instances of all that string data.
As people write libraries that think like this, and others use them, and write code like it, all the little bits of memory that is left for the GC add up to a tremendous amount. Take a look at perfmon sometime and see how many collections the GC does per second for an idea of the work going on under that managed cover providing you with an 'easy to use' environment. That's why they won't write the OS using it.
Which in turn has a link to the software - http://www.collab.net/products/sfee/demo/
Looks like SoftwareForge isn't FOSS, but Collabnet's TeamForge (which is fair enough, but disappointing as I wanted to run military open-source software, just to tell my boss that if its good enough for the DoD, its good enough for us!
it works well with YUM too - in fact, Webmin is one of the best admin things around. I think every project should be mandated to create a webmin administrative module before being allowed into the wild :)
Yeah, it's crazy that book authors should want to make a few bucks to eat and pay rent. They're just greedy bastards.
yep, but how much of a book's cover price goes to the author? That's the problem here - not that an author deserves everything he could get for his book (often that's nothing, too many books have tiny, insignificant sales), but that the publisher gets very rich on their back.
Self-publishing through something like the book-on-demand would be a very good thing, if it lowered the cost of publishing a book, and cut out the publisher.
yeah, like ethernet and tcp/ip never caught on in a unified manner.
The problem is not the standard (though overly complicated ones as designed byt he committes that submit them to the standards bodies certainly don't help), but the individual companies (or company) that decides to make a few changes to suit itself and doesn't care about interoperability with everyone else. The "de facto" standard caused by monopoly abuse.
"just works", like Windows you mean. (see, I posted about this the other day, the "just works" meme has been dragged up to kick Linux at every opportunity).
What you don't realise is that it does just work, the default installation on Ubuntu has had all that choice stuff taken away from you so you get a single, supported set of options. If you don't like any of the options you can swap them to something else (Kubuntu for example), yet the original install that those 90% of users get works, pure and simple.
As a corporate, you get the same power - the IT department can make up a distro of the bits they think works best and ship it out to all their users. They get the power and flexibility of being able to do that, yet still get a single install that those 90% of their users can work with, without any hassles.
yeah, its true - "Linux, it just works".
PS. :)
"Windows, just about works"
If they just invested in government bonds, why did their investments tank, taking up most of the reduction in quarterly profits?
Seems someone got a little greedy and thought they could ride the wave of CDOs et al to make more money than the boring safe investment would deliver.
For example, they were talking about "Cairo" (which was like V+3) long before Windows 95 showed up. It was really going to blow away NextStep!
And it did! Shame they had to wait until processing power had gone up ten times, graphics power up a hundred times, and ram capacity increased 1000 times!
Just think, they might produce something that blows Compiz away in another 5 years of hardware advances!
I'm not so sure you've been right that far back though.
I once read that computers would be like aircraft - first we had fragile things made of wood, then the state of the art got better and better over time, until you ended up with 2 airplane models: the 747 bulk carrier for cheap, flights carrying lots of people, and Concorde which was a luxury, high-end, performance model for those who had far too much money.
Computers are reaching the same plateau - you get to a point where your PC is good enough for everything you need it to do, and you don't bother spending the extra to upgrade, unlike times past when a new CPU model or OS was an instant best seller. You can see this happening right now with XP and core2duo. People don't see the need for Vista, or for quad cores.
When MS comes out with Windows 7, and Intel make a 10-core CPU, no-one will buy them. (well, ok, they may well buy W7 because MS will kill off XP and give you no choice in the matter).
So yeah, I rank "just works" up there with TCO. It says little about what kind of experience I can expect.
which is exactly why the marketeers use the term, its like panning the camera away from the attack scene, your imagination fills in far more than those 2 words could ever actually mean.
in fact a good app will have a CLI.
Then someone else will come along and make a gui that abstracts that command interface, in a way that is much better than anything the guys who wrote the cli-based system would have made.
I think something like subversion is a case in point, excellent cmdline tool, some excellent guis too, many available for environments the svn devs have no interest or skills in.
lol. Slashdot has always been like this. The mainstream media noticed us and ignored us years ago.
I've noticed a lot of posters relatively recently that are popping up and basically saying "linux is not ready yet, until you plug it in and it 'just works' it won't be ready", either implying Windows does "just work" or explicitly stating it. I know no computer does that, there's niggles in everything, but I seem to hear that mantra more often than I ever did.
Maybe you havn't been paying attention to them, but they're there.
There are a lot of pro-MS postings, I've done them myself, but they tend to be more objective against trolls saying Linux is perfect at everything and Windows couldn't possibly be any good. Windows is a perfectly usable OS, I just consider Linux to be architecturally better and has the potential to be significantly better.
Thing is, Ballmer (and I doubt it is him directly, probably more fortuitous circumstance for MS that they had a research project out of J++ that came good) is doing more anti-competitive moves with the new technology. Its just now, they can push .NET massively knowing that all .net programmers are tied securely into Windows for ever.
It isn't a nasty anti-competitive move, but it is designed to keep Windows marketshare at other platform's expense. (jury's out whether it will become nasty if Mono comes good, and gets lots of market share, and Microsoft loses developers to it)(but it seems that won't happen while there are plenty of bits that are not supported in Mono).
interesting idea, not everyone's a coder after all.
Perhaps an Autumn of Documentation, followed by a Winter of Marketing, and a Spring of Sales.
Think how much goodness could be spread by some of the above!
This is the kind of thing that makes a difference. 70 million euro is a lot of money.
Also, note the quote at the end, usually referred to as a cost of migrating to Linux.
Moving from Microsoft XP to Vista would not have brought us many advantages and Microsoft said it would require training of users.
One day we will have a replacement for Exchange, Thunderbird will have groupware features built-in (or plugged in) and MS will have serious competition on their hands. Or possibly will heat up the marketing war.
With Linux this in general does not work because the underlying dependencies change too often and too much.
You've obviously not built anything that requires the Visual Studio 2005 service pack 1 redistributable. And MS now considers it ok to break old apps (ie require a recompile) for 2008 too.
When I upgraded to Vista, I found a few apps that wouldn't work - and quite a few drivers that suddenly weren't supported and weren't ever going to be. If I wanted those devices to work, I had to buy a new one.
To which product(s) are you referring?
all the other ones. All the application servers, Peoplesoft, Siebel (ugh)
really, the 'sell you everything in one stack' sounds more Microsoft than anything else.
At least Oracle's new hardware business will do better than Microsoft Bob and the Xbox.
quite true, so now what does this mean for 'unbreakable' Linux? Will it be slowly dropped, or kept around as the x86 alternative to Solaris? In the datacentre, I'm certain Solaris will be the old new OS for Oracle... what about smaller customers, I don't think Solaris for x86 ever really got anywhere, definitely not compared to Linux.
They have a couple of options -- either (A) hold the schools to account for what they're doing and not doing
the biggest issue here is that school are allowing kids access to the freaking internet. I'm sure none of the kids there give a damn about any gay/lesbian website - they're too busy talking crap with their mates on facebook. Instead of learning stuff.
So yeah, sure we should be outraged at some faceless company deciding what's allowable or not on the internet, but we should be equally outraged that kids have access to all the rest of the internet whilst at school.
now, I'm going to go back to surfing those sites my employer deems acceptable for me to waste... sorry, "profitably leverage my skills in a proactive self-learning manner" on.
not just that, but Samsung has also got Yahoo widgets and internet-access capabilities built into its new range of TVs. I leave it to the reader to dsecide if that's a good thing or not, but its a start.