Nothing stops you from using a GPLed stack to see how a TCP/IP stack should work.
Until people complain how you copied portions of code and force your product to be GPLed. Without a BSD version of TCP/IP, it would have never become a standard (my opinion), as commercial software developers would not have adopted it.
If Apple keeps the code free, its not. IF Apple turns it into a proprietary OS like OSX, its a very bad thing- its yet another proprietary piece of code that will eventually be used to lock in and fuck over consumers. Apple is no better than MS, the two companies show the same lack of ethics- Gate is just better at it.
No, evern if apple keeps their code to themselves, we still end up with an OS that has had many many hours of testing on the core OS components - hence, we have a well-tested OS that people are willing to pay for. And if they're not willing to pay for it, then clearly apple's value-add isn't good enough, and they use the BSD free version.
If apple's extra bits are so essential and they can't be developed for free in the BSD version, then people *should* be paying apple for the R&D that they did, if apple sees fit to charge for it.
And this is true under the GPL as well. If I write an app, you are completely free to improve it and sell the improved version.
Of course - if i make my source (that perhaps required PAYMENT to coders, to develop) available for free.
Why would I outlay money paying coders to develop code that I will not be able to effectively sell?
Like it or not, certain aspects of software design are *not* well suited to the free software development model.
Just as a random example - mining software. If I am in charge of a multi-billion dollar mining project, am I going to sit around waiting for a bunch of hippies to write the software for me, or am I going to pay a developer for it, so it can be ready to ship *THIS* year, and not at some random time "when we get around to it"?
And if I've paid lots of money for this development, would I give it away to my competitors?
Like fuck... i'd be "reinventing the wheel" (not using any GPL software) so that I can keep it guarded.
Which means that company B down the road develops their own entirely incompatible version to do the same thing - and the people who have to support both version's bug sets in the field get fucked. Wasted human resources spent developing/maintaining for what? If large parts of the software were BSD, each version could have it's own localised tweaks to get an edge, but the core would be well-tested and have common maintenance issues.
This is a core difference between the GPL and the BSD licenses.
GPL people want to enforce their agenda of "well I gave my stuff away, so you have to as well - i don't care if you've paid people to do it".
BSD people put stuff out, and merely want credit - but if you want to do your own proprietry enhancements, go ahead - we don't care, the software is unencumbered.
And the point is, that good solid foundations + commercial development is more likely to result in good solid software - which benefits everybody who has to deal with it.
Free (as in unencumbered - eg, BSD) software does not go away in this scenario - it's merely able to co-operate a lot better with commercial software that is not suited to free development.
The GPL does not encourage the use of a good solid GPL core in commercial software development in this way - and that is a HUGE downside, in my opinion.
Taking the code and making a proprietary branch doesn't improve the world.
This is where some, including myself would argue the point that it does.
Where would we be without a good reference implementation of TCP/IP (BSD network stack) for other operating systems/devices to use as they please?
How is having a stable, useful OS that Apple can quickly implement large portions of, on their hardware platform, a bad thing?
The point of the BSD license is that if you restrict the ability of people to take your code, enhance it, and keep some of their changes secret, you'll have people reinventing the wheel so that they can keep their buggy-ass codebase proprietry.
If someone takes my code and improves it to the point that it's worth paying money for vs my free version - good for them... they deserve the right to sell their enhancements for what people think is fair. My free code remains available - if it's "good enough" for others then they can use that. If i think that feature needs to be in the free version, then it's up to me, or someone else (willing to code for free) to code it.
All GPL does is encourage reinvention of the wheel...
Slashdot isn't really representative of Stallman's community. They transitioned from an Open Source site to a "Geek Culture" site some time ago.
You mean, back in 1997 or earlier?
I've been reading/posting to this site since at least 1997 (well before userIDs were implemented - I just didn't bother creating one for a couple of weeks), and it's *always* been a geek culture site as far as I'm concerned.
That perhaps a lot more people other than "Open Source" people know about it now is besides the point - I don't think Rob ever intended it to be wholly devoted to open-source 100% - it started out as just a "this is cool" blog...
So, because RMS had a good idea many years ago, he can never ever be wrong, is that what you're trying to say?
I call bullshit... no one is correct, 100% of the time. RMS is an idealist. In a utopian world perhaps all of his ideas would work. But that isn't the world we live in, I'm afraid.
DRM isn't dangerous... DRM is simply encryption, and encryption isn't bad. I don't think anyone here wants encryption restricted in any way. Everyone has the right to encrypt any data in any way they want, period!
You're talking about RMS here, who has previously been against the idea of even having passwords for computer access... (http://mannu.livejournal.com/10626.html)... you think he's going to be all for people encrypting their data?:D
I applaud Microsoft for at least trying, because unlike Google or Apple, they realize that the future is in a database driving relational file system and not stop gap pseudo-solutions like indexing. Its obviously a difficult concept to implement, but once anyone is able to implement the idea, it will be a VERY welcomed concept and improve the functionality and usability of an operating system. I for one would switch to and swear by ANY OS that implements this idea properly, whether its Linux, OS X, or yes, even Windows.
Being ready now (google, spotlight) and "good enough" > being ready "real soon now", but being perfect.
Don't get me wrong, I too, applaud microsoft for trying, and if it ever gets released and works, it will be the superior solution. However, people need the type of searching ability *now* - and if google/spotlight are able to provide a "better" solution today, then that is what will be used...
Hey kiddo, I wrote an IPSEC howto and MPD howto for FreeBSD (currently lost, however when my previous employer took the webserver in question down) - what have YOU done?
Hmmm... moderated down as over-rated... typical Fedora fanboys with moderator access, I see...:D
My "point" I guess is that KDE/Gnome are not really "Fedora" as such. I mean, shit... I can run Gnome 2.16/KDE 3.5 on FreeBSD 4.4 if I really wanted to. Ditto for CUPS.
As to upgrading your Kernel on Fedora - not sure here, I don't run it. Tried core4 out for a couple of days, thought it was hugely over-rated, and went back to Ubuntu/FreeBSD:D
This is one of the MAJOR reasons I prefer the BSDs - the apps are seperated from the OS. When FreeBSD X.Y comes out, you know there's been major changes to the way core O/S components work, compared to X-1.Y... and if you're happy with your old version, you can simply upgrade the APPS and leave the perfectly good (for your purposes) base O/S alone:)
Oh, and just on the whole locking up 6 aircraft and shooting them all down from huge ranges - that would be the phoenix missile.
Unfortunately, it's good against un-manouverable bomber/strike aircraft, but against a modern fighter-bomber, it's not able to turn well enough.
Don't mistake me for a tomcat hater - I think it's a way better plane overall than the superpig, but there were several fairly good reasons it's been retired.... yet another one being that the airframes are simply getting too old, and grumman no longer exists to build new ones...
Unfortunately, the mission that the tomcat was designed for (long range fleet defense against strike aircraft) no longer exists.
The F/A-18E ("Super" hornet), whilst being a comparatively shitty airframe in terms of flight dynamics, is far, far stealthier and far, far cheaper to maintain.
So, it's a swing wing with 1 pivot - 1 wing forward, the other back?
I recall reading about this design in an aircraft book that was published in the late 70s (the F16 was mentioned in it was the brand new mass-production USAF fighter).
Like em or not, microsoft made a good, reliable operating system, and they maintained it,
For *very* limited definitions of "good", "reliable" and "maintained".
The funny/sad thing is that microsoft operating system "game supremacy" (in terms of overall game availability) has been evident for a very long time, yet none of the other operating system developers have done anything (with their knowledge of that fact) to boost their own OS sales. (unless i'm mistaken?) They could have opened game houses of their own, for instance.
Game development these days is not something that can be easily done on a low budget - very few people/small teams contain decent programmers, artists AND musicians willing to work for free - and no one is going to pay decent wages to develop a game that has a very limited potential market (chicken and egg situation).
Kill Microsoft, and games with be available on multiple platforms again (as they were in the Amiga/C64/ST/Console days.
Besides, a lot of the good Windows games can be run under linux anyway (Transgaming).
"Everyone in the US remembers just fine"... oh ok so that makes it just dandy.
The imperial system is not intuitive (stones? miles? farenheit?), or even standardized (US gallons? british gallons? dry gallons?) at all.
All clinging to imperial does is cause problems dealing with the rest of the world - and yes, contrary to the commonly held US belief structure, there's more people OUTSIDE the US than IN it.
Dividing by 10 does work well (decimal point shifting) - you have far fewer base units to learn/remember. Most people can get by with kilos, metres, litres - most of the other units are just base 10 multiples/fractions of that. Also, many of the units are easily calculated (eg, volume vs distance, volume vs weight, etc).
But before you even consider that, how about the fact that the SI measurements are largely metric anyway? It's not just the rest of the world you have to deal with - SI standard measurements need to be converted as well.
I'm sure NASA would disagree with the idea that things are just fine the way they are:D
I have enough experience with imperial measure (feet, inches, psi, stones, pounds, pints... especially pints:D) to know that it's nowhere near as simple and "clean" as metric - and it seems the majority of the civilized world agrees with me:D
But hey, I don't particularly care - if you lot want to make things hard on yourselves (in the long run - continuing for the rest of eternity to convert metric back to imperial), be my guest....:)
If documentation is a pet hate of linux, then check out FreeBSD. It has more of a "commercial unix" feel/flavour to it, and the documentation in my experience has always been pretty top notch.
It's also many, many times faster/more responsive than SCO...
Hmmm, my bad... according to TFA, some companies have been charged $1399 per cpu license of Linux (and thus, SCO is more expensive I presume)? Hahaha....
Training in SCO's EdgeBuilder developer kit will be offered at the conference; attendees completing it will be given US$1000.
Minus $699 SCO license, minus cost of conference attendance, minus cost of user licenses, minus pain and suffering dealing with the most (IMHO) painful unix variant I have ever dealt with (I've only used/been admin on a couple of SCO variants, Solaris, FreeBSD and most of the major Linux distros though)...
I just don't get it personally. The rest of the scientific world (and hell, most of the civilized world full stop) uses metric units - why are they so entrenched in the USA? "It's too hard to change" is just a cop out...
So uh... is it a general purpose machine, or has it got an instruction set like brainfuck? No point in running at 500ghz if it's going to take 10,000 clock cycles to perform addition...
Slashdot Mirror
You're correct. IMHO, not using well-tested code *EVERYWHERE* (i.e., all software, commercial or not) is a loss to humanity as a whole.
Until people complain how you copied portions of code and force your product to be GPLed. Without a BSD version of TCP/IP, it would have never become a standard (my opinion), as commercial software developers would not have adopted it.
No, evern if apple keeps their code to themselves, we still end up with an OS that has had many many hours of testing on the core OS components - hence, we have a well-tested OS that people are willing to pay for. And if they're not willing to pay for it, then clearly apple's value-add isn't good enough, and they use the BSD free version.
If apple's extra bits are so essential and they can't be developed for free in the BSD version, then people *should* be paying apple for the R&D that they did, if apple sees fit to charge for it.
Of course - if i make my source (that perhaps required PAYMENT to coders, to develop) available for free.
Why would I outlay money paying coders to develop code that I will not be able to effectively sell?
Like it or not, certain aspects of software design are *not* well suited to the free software development model.
Just as a random example - mining software. If I am in charge of a multi-billion dollar mining project, am I going to sit around waiting for a bunch of hippies to write the software for me, or am I going to pay a developer for it, so it can be ready to ship *THIS* year, and not at some random time "when we get around to it"?
And if I've paid lots of money for this development, would I give it away to my competitors?
Like fuck... i'd be "reinventing the wheel" (not using any GPL software) so that I can keep it guarded.
Which means that company B down the road develops their own entirely incompatible version to do the same thing - and the people who have to support both version's bug sets in the field get fucked. Wasted human resources spent developing/maintaining for what? If large parts of the software were BSD, each version could have it's own localised tweaks to get an edge, but the core would be well-tested and have common maintenance issues.
This is a core difference between the GPL and the BSD licenses.
GPL people want to enforce their agenda of "well I gave my stuff away, so you have to as well - i don't care if you've paid people to do it".
BSD people put stuff out, and merely want credit - but if you want to do your own proprietry enhancements, go ahead - we don't care, the software is unencumbered.
And the point is, that good solid foundations + commercial development is more likely to result in good solid software - which benefits everybody who has to deal with it.
Free (as in unencumbered - eg, BSD) software does not go away in this scenario - it's merely able to co-operate a lot better with commercial software that is not suited to free development.
The GPL does not encourage the use of a good solid GPL core in commercial software development in this way - and that is a HUGE downside, in my opinion.
This is where some, including myself would argue the point that it does.
Where would we be without a good reference implementation of TCP/IP (BSD network stack) for other operating systems/devices to use as they please?
How is having a stable, useful OS that Apple can quickly implement large portions of, on their hardware platform, a bad thing?
The point of the BSD license is that if you restrict the ability of people to take your code, enhance it, and keep some of their changes secret, you'll have people reinventing the wheel so that they can keep their buggy-ass codebase proprietry.
If someone takes my code and improves it to the point that it's worth paying money for vs my free version - good for them... they deserve the right to sell their enhancements for what people think is fair. My free code remains available - if it's "good enough" for others then they can use that. If i think that feature needs to be in the free version, then it's up to me, or someone else (willing to code for free) to code it.
All GPL does is encourage reinvention of the wheel...
I've been reading/posting to this site since at least 1997 (well before userIDs were implemented - I just didn't bother creating one for a couple of weeks), and it's *always* been a geek culture site as far as I'm concerned.
That perhaps a lot more people other than "Open Source" people know about it now is besides the point - I don't think Rob ever intended it to be wholly devoted to open-source 100% - it started out as just a "this is cool" blog...
I don't even install the flash player on Windows...
I call bullshit... no one is correct, 100% of the time. RMS is an idealist. In a utopian world perhaps all of his ideas would work. But that isn't the world we live in, I'm afraid.
You're talking about RMS here, who has previously been against the idea of even having passwords for computer access... (http://mannu.livejournal.com/10626.html) ... you think he's going to be all for people encrypting their data? :D
Being ready now (google, spotlight) and "good enough" > being ready "real soon now", but being perfect.
Don't get me wrong, I too, applaud microsoft for trying, and if it ever gets released and works, it will be the superior solution. However, people need the type of searching ability *now* - and if google/spotlight are able to provide a "better" solution today, then that is what will be used...
Hey kiddo, I wrote an IPSEC howto and MPD howto for FreeBSD (currently lost, however when my previous employer took the webserver in question down) - what have YOU done?
My "point" I guess is that KDE/Gnome are not really "Fedora" as such. I mean, shit... I can run Gnome 2.16/KDE 3.5 on FreeBSD 4.4 if I really wanted to. Ditto for CUPS.
As to upgrading your Kernel on Fedora - not sure here, I don't run it. Tried core4 out for a couple of days, thought it was hugely over-rated, and went back to Ubuntu/FreeBSD :D
This is one of the MAJOR reasons I prefer the BSDs - the apps are seperated from the OS. When FreeBSD X.Y comes out, you know there's been major changes to the way core O/S components work, compared to X-1.Y... and if you're happy with your old version, you can simply upgrade the APPS and leave the perfectly good (for your purposes) base O/S alone :)
That must be a *very* lenient test for considering US students literate then?
Most of the emails, posts, chat and webpages I see hosted/posted from the US are not what I would call "literate".
Unless it's only the 15% most illiterate people you have who are posting stuff on-line, I'd say those statistics are a crock of shit.
Err... by "it" regarding turn performance... I'm talking about the phoenix... not the tomcat ;)
Unfortunately, it's good against un-manouverable bomber/strike aircraft, but against a modern fighter-bomber, it's not able to turn well enough.
Don't mistake me for a tomcat hater - I think it's a way better plane overall than the superpig, but there were several fairly good reasons it's been retired.... yet another one being that the airframes are simply getting too old, and grumman no longer exists to build new ones...
Unfortunately, the mission that the tomcat was designed for (long range fleet defense against strike aircraft) no longer exists.
The F/A-18E ("Super" hornet), whilst being a comparatively shitty airframe in terms of flight dynamics, is far, far stealthier and far, far cheaper to maintain.
I recall reading about this design in an aircraft book that was published in the late 70s (the F16 was mentioned in it was the brand new mass-production USAF fighter).
Forget the name of the book unfortunately...
Yawn...
For *very* limited definitions of "good", "reliable" and "maintained".
Game development these days is not something that can be easily done on a low budget - very few people/small teams contain decent programmers, artists AND musicians willing to work for free - and no one is going to pay decent wages to develop a game that has a very limited potential market (chicken and egg situation).
Kill Microsoft, and games with be available on multiple platforms again (as they were in the Amiga/C64/ST/Console days.
Besides, a lot of the good Windows games can be run under linux anyway (Transgaming).
If enough people clap their hands, won't the energy dissipated as heat from the clapping impact and air compression raise the temperature even more? :)
The imperial system is not intuitive (stones? miles? farenheit?), or even standardized (US gallons? british gallons? dry gallons?) at all.
All clinging to imperial does is cause problems dealing with the rest of the world - and yes, contrary to the commonly held US belief structure, there's more people OUTSIDE the US than IN it.
Dividing by 10 does work well (decimal point shifting) - you have far fewer base units to learn/remember. Most people can get by with kilos, metres, litres - most of the other units are just base 10 multiples/fractions of that. Also, many of the units are easily calculated (eg, volume vs distance, volume vs weight, etc).
But before you even consider that, how about the fact that the SI measurements are largely metric anyway? It's not just the rest of the world you have to deal with - SI standard measurements need to be converted as well.
I'm sure NASA would disagree with the idea that things are just fine the way they are :D
I have enough experience with imperial measure (feet, inches, psi, stones, pounds, pints... especially pints :D) to know that it's nowhere near as simple and "clean" as metric - and it seems the majority of the civilized world agrees with me :D
But hey, I don't particularly care - if you lot want to make things hard on yourselves (in the long run - continuing for the rest of eternity to convert metric back to imperial), be my guest.... :)
It's also many, many times faster/more responsive than SCO...
Hmmm, my bad... according to TFA, some companies have been charged $1399 per cpu license of Linux (and thus, SCO is more expensive I presume)? Hahaha....
Minus $699 SCO license, minus cost of conference attendance, minus cost of user licenses, minus pain and suffering dealing with the most (IMHO) painful unix variant I have ever dealt with (I've only used/been admin on a couple of SCO variants, Solaris, FreeBSD and most of the major Linux distros though)...
How does "fuck you SCO" sound? :)
"4 cycle" means 4-stroke, not 4 cylinder.
If it's the engine i'm thinking off (briggs+stratton typical thing) then it's a single cylinder 4 stroke.
I just don't get it personally. The rest of the scientific world (and hell, most of the civilized world full stop) uses metric units - why are they so entrenched in the USA? "It's too hard to change" is just a cop out...
So uh... is it a general purpose machine, or has it got an instruction set like brainfuck? No point in running at 500ghz if it's going to take 10,000 clock cycles to perform addition...