Fair enough. If someone wants to write software and then only license it to people with blue hair born on a wednesday, then they can.
My reply was in particular to the line "Why in the world would anyone want to write a replacement grep?" There are good reasons for doing so... commercial viability for one.
It's practically impossible to build a large-scale system nowadays without drawing on existing work (if not code, then on concept). So, we need to build on existing work when we're producing commercial products.
These non-GPL replacements are often written by the people who *need* this non-GPL code for their work, and so they write it.
And my ass isn't lazy. It gets regular exercise (see the Intel "Ottoman")
Ah, depends what you mean by "use". If "use" is just running the software yourself, then fine. However, if it means "use it in a service or product", then no.
I wasn't trying to slag off the GPL there. Just point out that sometimes GPL'ed software isn't appropriate for use in a particular circumstance. Thus, re-engineering code under the BSD license *does* make sense. The original point, I think.
What's all this jealousy shite? Both are free (as in 'beer') for the most part, and I have the choice to use either.
Jealousy in this case implies that for some reason I can't choose Linux, and am envious of it's features. Sure, the grass is always greener, but it doesn't make any difference when there's no fence between.
There's nothing stopping anyone from using GNU tools in *BSD. However, the aim is to make it a clear option so that those who don't agree with the GPL can choose to avoid GNU code.
The idea is "Give people the choice". It's better than the choice of "Either accept the GPL or don't use *any* of the software."
You must admit, the GPL is controversial. Whether you agree with them or not, I hope you can see that some people have a valid reason not to like it. The first page or two of the GPL (the preamble) is effectively a political manifesto. How 'free' is software licensed such that you have to accept a particular political belief?
Anyway, I digress. The fact is, I've worked at companies where the GPL has prohibited them for use in certain products/services. The BSD license didn't. Nothing I could do about it, but there you are. So there *is* a practicality to it.
This guy just didn't state it very well (or at all).
Yeah, FreeBSD took a while to get SMP working (although I was using it two years ago in production.)
That's because they chose to engineer it carefully rather than just throwing in the first implementation sent to them.
Agreed.. Linux tends to have better support for whizzy new gadgets. FreeBSD tends to concentrate on good stable support for good stable hardware. Since FreeBSD aims more at the server market than the home market, you tend to engineer a new machine specifically for FreeBSD, rather than throwing one together. (Not that you *can't* just throw one together)
It's not that OpenBSD is secure.. it's that OpenBSD's primary goal is security.
Completely different matter. FreeBSD's primary goal is a good BSD on the x86 platform.
BSDI's primary goal is to make money (pretty much the definition of commercial. Sometimes even the legal definition -- in the UK the Companies Act states that the primary legal responsibility of the directors of a company is to make the company more money.)
Since OpenBSD's goal is security, the other BSDs can draw on that experience to make them more secure... similar to the way that FreeBSD and Linux have both drawn on each other's experience in certain areas.
Well, don't all of these draw from the original UNIX code base? Perhaps there are no original lines of code left in, but I can't think of a completely clean-room UNIX clone out there. (Clean room in the sense that the authors don't even *look* at a preexisting line of code).
Thus, by your reasoning, Linux is just as fragmented.
The sense in which the BSDs aren't fragmented is that they don't *try* to synchronise code. Occasionally, code is copied from one to the other, but only by choice. They are maintained as completely separate projects, which may happen to collaborate at times.
Linux distros share code left, right and centre without explicit human intervention. *That's* why they're fragmented. Without core developers carefully controlling every bit of code that migrates, you get major consistency problems.
On the whole, I agree that that article was just as flamey as the average Linux advocacy article. However, how many people here would read something that wasn't 'controversial'?
(Channel 4 News is one of the major news shows on in the evening on UK tv. It's main plus-point is the anchorman Jon Snow, who is pretty damn good at asking nasty questions)
They had a rather ill-informed report, mentioning the Cult of the Dead Cow and Back Orifice, and then went on to a head-to-head between the MD of MI2G and some woman from Microsoft.
Unfortunately, neither the MI2G guy or Jon Snow actually pinned her down to anything, and let her get away with the party line of "Isolated incident.. not a problem.. all the fault of the hackers.. E-mail's never secure anyway."
He almost got her on a few, like "Wasn't the service up for a while after you noticed before you pulled the plug", but didn't follow up when she fluffed him, and they didn't bring up the possibility of it being Microsoft's fault/responsibility. Jon Snow finally summed up with a "Let the viewers decide" line.
Bit of a shame. I feel they didn't really research it too well. Jon Snow did a Bill Gates interview once, and asked him something like "Your personal fortune could supply running water and good sanitation to every person on the planet. How do you feel about that?" Ended up making Billy-boy seem like the devil incarnate. =)
"Good morning. Thank you for calling Linus Torvalds. If you would like a quote concerning the following subjects, please type in your corporate expense credit card number now..."
At my last job, I set up a news server and a few Perl scripts to grab content off about three or four sites: here, Dilbert, BBC News and um.. that's about it. All my mailing lists fed in as well. Once that was working, I only browsed for work-related things (software, documentation, etc.) I rarely even clicked through to other sites.
The reason? I couldn't be bothered to browse. I think people are realising that the web is really about as exciting as the telephone network, and are using it as such... calling who they know occasionally, rather than phoning everyone on the planet.
In terms of number of hits versus site-count, the roportion of hits going to minor sites is probably dropping, while proportion of hits to the big portals is rising.
For my uses, in my opinion, etc, the best thing about FreeBSD is the centralisation of source.
There is one FreeBSD. It's not a bloody pick-n-mix, but a unified software product with a clear across-the-board versioning strategy, so you know what's being installed, and where, rather than a free-for-all like Linux.
The whole pedantic argument about Linux being a kernel and not a full operating system that RMS goes on about is exactly the problem.
At the moment, buying a computer is sometimes more difficult/dangerous than buying a house or a car. You can go the easy route and get ripped off, or spend every waking minute reading up on the latest news to build yourself one.
Linux can suffer from this. If you can find a good distro, you're okay, but chances are, another distro will have features you want, and you end up with a patchwork system. How can you have any level of confidence in that?
FreeBSD.. you have a single tree maintained by a well-defined group. Everything concentrates on that (okay, let's forget about the other *BSDs here...)
As far as Darwin's concerned? I like the fact that Apple recognises the other *BSD teams and at least says they're going to work *with* them on the new MacOS NeXTy things.
Shouldn't it stand for "Gnu's Not Useful-without-a-working-kernel"? =)
This isn't a rhetorical question, but a real "what-if": What if Linus had done the kernel, and then went looking for tools and not found any?
I personally think that someone (possibly Linus) would've written them, rather than mothball Linux. Alternatively, he might've taken a bit of BSD (ugh: BSD/Linux!)
If the Hurd had been finished earlier, maybe we'd all be using that instead, and it would be a true GNU system. However, right now, it sounds like sour grapes from Stallman.
I also get a little annoyed by his insistence on making everything a political issue. It devalues the good work he's done. I don't blame anyone for not putting on the GNU prefix. People might think that you agree with RMS's radical stance. A few decades ago, I'm sure even mentioning "GNU" would've put you up in front of McCarthy. The whole free will argument goes out the window when you don't even have free will to choose what you call something.
Well, fine. As long the logo isn't compulsory, I figure most will go with it. There's enough kudos for everyone.
Oddly enough, Netscape runs much worse on FreeBSD for me than it did when I was using Linux.
Y'know, that's the main reason I started using a desktop environment. I use KDE's browser thingy instead.. much less crashy than Netscape =)
Still doesn't come close to IE5 for stability for me, and I thought I'd *never* say something like that.
Re:Only 10% of the IP addresses are used?
on
IP Address Shortage
·
· Score: 1
You'll also find that most of the big commercial class 'A' users have slapped up big firewalls and proxies anyway, so NATting the lot of them would not reduce their functionality.
Revoke the whole lot of them, if you ask me, and then let them deal with it.
I may be wrong, but I seem to remember that some of the authors of those RFCs and documents which discusses CIDR as an approach to solving the shortage were from Class-B users themselves.
Whether their sales model fits your preferences, or whatever, the BeOS does what it's meant to do: supply scalable low-latency media support. Nothing else does this.
Quoting from be.com: "...it was designed from the ground up to handle the real-time manipulation of high bandwidth digital media on off-the-shelf personal computers."
The thing is, Linux (and other free unices) might be very good. However, they're built on a design which is a little on the old side. It was an excellent design, and still is, but anyone who claims that it's the perfect design, and seriously believes that Linux (et al.) will _always_ be the best OS is in cloud-cuckoo land.
There _are_ better ways of doing things. I've used the BeOS a little (I bought that crippled demo version). From what I can see, they've achieved what they set out to achieve AND MORE. It's not focussed as a server platform, or a general-purpose workstation platform. However, it (will be) bloody good at both of those tasks... it's got the basis for it.
The one thing I can really criticise them for is not aiming for the general-purpose market, when it's quite clear they could. Maybe they'll do this when they think it can compete. It's still not complete enough for JQPs.
As far as open-sourcing it goes, I wish they would. However, although I don't agree with them on this point, I can understand their decision. Could you really justify 9 years of coding (and a hell of a lot more man-years) when you're going to open-source it? Yeah, Apple, Microsoft, etc. can, but they have lots of money, and other revenue streams. This is now Be inc.'s only major form of revenue since the BeBox went down. They're just about to IPO. Damn good move financially.
If/When they actually get round to shifting some serious volume, maybe they can then start doing the sensible thing and open-source it. For now, what would stop anyone (eg. Microsoft) nicking the internal design, and putting some code-monkeys on rewriting every line of code?
There's some serious IPRs in the BeOS. No other OS is as pervasively multi-threaded. As a result, no other OS can claim such a good performance ratio (relative performance/processor). Until they get some market share, they've got no protection.
The line "No-one ever got fired for buying the BeOS" hasn't been uttered yet (although it might be true.. virtually no-one HAS bought the BeOS!). Once that's the norm, then if they release the source, people will still buy the BeOS rather than the hypothetically-cloned "Microsoft Windows 2000 Media Edition"... They will have that protection. Some commercial unices currently have that protection. People will still buy them, even though there are cheaper (and free) alternatives. People like Redhat don't _need_ that protection as much, as they haven't made such a massive investment while making little return on that investment.
One alternative would be huge amounts of patent claims. Scyeah.. I'm sure we'd all *love* that. Anyway, most of the stuff they're doing is prior art. It's just they're doing it a) well, and b) in the right combination.
Don't slam the guys for trying to stay afloat. I think the BeOS architecturally is/could be the way future OSes are built.
Re: User Interfaces. Well, unlike Windows, etc, the GUI ain't the only usable interface. You telnet to a BeOS-based machine, and you get 'bash'. It's like IRIX, etc. It's difficult to get rid of the GUI, but the whole machine isn't dependent on the GUI. This way, according to demand, they can easily remove the GUI, or allow it to be switchable. No big deal.
(yikes.. I didn't *mean* to write that much.. honestly.)
Slashdot Mirror
I'm a dual-national (UK/USA), but I haven't been over there since about 1985. I wasn't even born there.
Does anyone know what restrictions/advantages I have as opposed to a normal everyday UK citizen, w.r.t this sovereignty issue?
Fair enough. If someone wants to write software and then only license it to people with blue hair born on a wednesday, then they can.
My reply was in particular to the line "Why in the world would anyone want to write a replacement grep?" There are good reasons for doing so... commercial viability for one.
It's practically impossible to build a large-scale system nowadays without drawing on existing work (if not code, then on concept). So, we need to build on existing work when we're producing commercial products.
These non-GPL replacements are often written by the people who *need* this non-GPL code for their work, and so they write it.
And my ass isn't lazy. It gets regular exercise (see the Intel "Ottoman")
Ah, depends what you mean by "use". If "use" is just running the software yourself, then fine. However, if it means "use it in a service or product", then no.
I wasn't trying to slag off the GPL there. Just point out that sometimes GPL'ed software isn't appropriate for use in a particular circumstance. Thus, re-engineering code under the BSD license *does* make sense. The original point, I think.
(Not rising to this one, but other comments..)
What's all this jealousy shite? Both are free (as in 'beer') for the most part, and I have the choice to use either.
Jealousy in this case implies that for some reason I can't choose Linux, and am envious of it's features. Sure, the grass is always greener, but it doesn't make any difference when there's no fence between.
There's nothing stopping anyone from using GNU tools in *BSD. However, the aim is to make it a clear option so that those who don't agree with the GPL can choose to avoid GNU code.
The idea is "Give people the choice". It's better than the choice of "Either accept the GPL or don't use *any* of the software."
You must admit, the GPL is controversial. Whether you agree with them or not, I hope you can see that some people have a valid reason not to like it. The first page or two of the GPL (the preamble) is effectively a political manifesto. How 'free' is software licensed such that you have to accept a particular political belief?
Anyway, I digress. The fact is, I've worked at companies where the GPL has prohibited them for use in certain products/services. The BSD license didn't. Nothing I could do about it, but there you are. So there *is* a practicality to it.
This guy just didn't state it very well (or at all).
Yeah, FreeBSD took a while to get SMP working (although I was using it two years ago in production.)
That's because they chose to engineer it carefully rather than just throwing in the first implementation sent to them.
Agreed.. Linux tends to have better support for whizzy new gadgets. FreeBSD tends to concentrate on good stable support for good stable hardware. Since FreeBSD aims more at the server market than the home market, you tend to engineer a new machine specifically for FreeBSD, rather than throwing one together. (Not that you *can't* just throw one together)
It's not that OpenBSD is secure.. it's that OpenBSD's primary goal is security.
Completely different matter. FreeBSD's primary goal is a good BSD on the x86 platform.
BSDI's primary goal is to make money (pretty much the definition of commercial. Sometimes even the legal definition -- in the UK the Companies Act states that the primary legal responsibility of the directors of a company is to make the company more money.)
Since OpenBSD's goal is security, the other BSDs can draw on that experience to make them more secure... similar to the way that FreeBSD and Linux have both drawn on each other's experience in certain areas.
Well, don't all of these draw from the original UNIX code base? Perhaps there are no original lines of code left in, but I can't think of a completely clean-room UNIX clone out there. (Clean room in the sense that the authors don't even *look* at a preexisting line of code).
Thus, by your reasoning, Linux is just as fragmented.
The sense in which the BSDs aren't fragmented is that they don't *try* to synchronise code. Occasionally, code is copied from one to the other, but only by choice. They are maintained as completely separate projects, which may happen to collaborate at times.
Linux distros share code left, right and centre without explicit human intervention. *That's* why they're fragmented. Without core developers carefully controlling every bit of code that migrates, you get major consistency problems.
On the whole, I agree that that article was just as flamey as the average Linux advocacy article. However, how many people here would read something that wasn't 'controversial'?
(Channel 4 News is one of the major news shows on in the evening on UK tv. It's main plus-point is the anchorman Jon Snow, who is pretty damn good at asking nasty questions)
They had a rather ill-informed report, mentioning the Cult of the Dead Cow and Back Orifice, and then went on to a head-to-head between the MD of MI2G and some woman from Microsoft.
Unfortunately, neither the MI2G guy or Jon Snow actually pinned her down to anything, and let her get away with the party line of "Isolated incident.. not a problem.. all the fault of the hackers.. E-mail's never secure anyway."
He almost got her on a few, like "Wasn't the service up for a while after you noticed before you pulled the plug", but didn't follow up when she fluffed him, and they didn't bring up the possibility of it being Microsoft's fault/responsibility. Jon Snow finally summed up with a "Let the viewers decide" line.
Bit of a shame. I feel they didn't really research it too well. Jon Snow did a Bill Gates interview once, and asked him something like "Your personal fortune could supply running water and good sanitation to every person on the planet. How do you feel about that?" Ended up making Billy-boy seem like the devil incarnate. =)
"Good morning. Thank you for calling Linus Torvalds. If you would like a quote concerning the following subjects, please type in your corporate expense credit card number now..."
That's the way I'd play it. =)
Hey, the form submit mails them to an AOL account.
Figures.
At my last job, I set up a news server and a few Perl scripts to grab content off about three or four sites: here, Dilbert, BBC News and um.. that's about it. All my mailing lists fed in as well. Once that was working, I only browsed for work-related things (software, documentation, etc.) I rarely even clicked through to other sites.
The reason? I couldn't be bothered to browse. I think people are realising that the web is really about as exciting as the telephone network, and are using it as such... calling who they know occasionally, rather than phoning everyone on the planet.
In terms of number of hits versus site-count, the roportion of hits going to minor sites is probably dropping, while proportion of hits to the big portals is rising.
Exactly.
For my uses, in my opinion, etc, the best thing about FreeBSD is the centralisation of source.
There is one FreeBSD. It's not a bloody pick-n-mix, but a unified software product with a clear across-the-board versioning strategy, so you know what's being installed, and where, rather than a free-for-all like Linux.
The whole pedantic argument about Linux being a kernel and not a full operating system that RMS goes on about is exactly the problem.
At the moment, buying a computer is sometimes more difficult/dangerous than buying a house or a car. You can go the easy route and get ripped off, or spend every waking minute reading up on the latest news to build yourself one.
Linux can suffer from this. If you can find a good distro, you're okay, but chances are, another distro will have features you want, and you end up with a patchwork system. How can you have any level of confidence in that?
FreeBSD.. you have a single tree maintained by a well-defined group. Everything concentrates on that (okay, let's forget about the other *BSDs here...)
As far as Darwin's concerned? I like the fact that Apple recognises the other *BSD teams and at least says they're going to work *with* them on the new MacOS NeXTy things.
As an Emacs user, I want *more* obscure keys! The Windows keys are useful for binding to Meta, and the other modifiers.
Anyway, I'm sure this is just a ploy to stop people playing things like Xpilot.
Also, the 'harddrive' and 'terrabyte' misspellings were in the quoted text from the contributor.. not Rob.
Aren't those things just cut-n-pasted from the submission?
Anyone know if anywhere sells toilet seats that look like iBooks? Just what I need when I'm remodelling the bathroom. =)
I'll probably get one.. a little VNC terminal with wireless networking for using around the house.
Sod the Linux idea.. just use it to VNC into my Unix and Windows boxes when I'm at home. I really don't want to be doing unixy stuff on *any* laptop!
Shouldn't it stand for "Gnu's Not Useful-without-a-working-kernel"? =)
This isn't a rhetorical question, but a real "what-if": What if Linus had done the kernel, and then went looking for tools and not found any?
I personally think that someone (possibly Linus) would've written them, rather than mothball Linux. Alternatively, he might've taken a bit of BSD (ugh: BSD/Linux!)
If the Hurd had been finished earlier, maybe we'd all be using that instead, and it would be a true GNU system. However, right now, it sounds like sour grapes from Stallman.
I also get a little annoyed by his insistence on making everything a political issue. It devalues the good work he's done. I don't blame anyone for not putting on the GNU prefix. People might think that you agree with RMS's radical stance. A few decades ago, I'm sure even mentioning "GNU" would've put you up in front of McCarthy. The whole free will argument goes out the window when you don't even have free will to choose what you call something.
Well, fine. As long the logo isn't compulsory, I figure most will go with it. There's enough kudos for everyone.
Oddly enough, Netscape runs much worse on FreeBSD for me than it did when I was using Linux.
Y'know, that's the main reason I started using a desktop environment. I use KDE's browser thingy instead.. much less crashy than Netscape =)
Still doesn't come close to IE5 for stability for me, and I thought I'd *never* say something like that.
You'll also find that most of the big commercial class 'A' users have slapped up big firewalls and proxies anyway, so NATting the lot of them would not reduce their functionality.
Revoke the whole lot of them, if you ask me, and then let them deal with it.
I may be wrong, but I seem to remember that some of the authors of those RFCs and documents which discusses CIDR as an approach to solving the shortage were from Class-B users themselves.
Does anyone know if this stuff (darwin and/or MacOS X Server) will work (badly/well) on an iMac?
I'm a FreeBSD fan, and I'm sick of beige boxes. =)
Oh, it wouldn't be a public server.. just another box to play with, and perhaps do some development on.
???
How about a new category: "Linux Kernel Releases", so we can disable it in our User Prefs, if we're not interested?
Whether their sales model fits your preferences, or whatever, the BeOS does what it's meant to do: supply scalable low-latency media support. Nothing else does this.
Quoting from be.com: "...it was designed from the ground up to handle the real-time manipulation of high bandwidth digital media on off-the-shelf personal computers."
The thing is, Linux (and other free unices) might be very good. However, they're built on a design which is a little on the old side. It was an excellent design, and still is, but anyone who claims that it's the perfect design, and seriously believes that Linux (et al.) will _always_ be the best OS is in cloud-cuckoo land.
There _are_ better ways of doing things. I've used the BeOS a little (I bought that crippled demo version). From what I can see, they've achieved what they set out to achieve AND MORE. It's not focussed as a server platform, or a general-purpose workstation platform. However, it (will be) bloody good at both of those tasks... it's got the basis for it.
The one thing I can really criticise them for is not aiming for the general-purpose market, when it's quite clear they could. Maybe they'll do this when they think it can compete. It's still not complete enough for JQPs.
As far as open-sourcing it goes, I wish they would. However, although I don't agree with them on this point, I can understand their decision. Could you really justify 9 years of coding (and a hell of a lot more man-years) when you're going to open-source it? Yeah, Apple, Microsoft, etc. can, but they have lots of money, and other revenue streams. This is now Be inc.'s only major form of revenue since the BeBox went down. They're just about to IPO. Damn good move financially.
If/When they actually get round to shifting some serious volume, maybe they can then start doing the sensible thing and open-source it. For now, what would stop anyone (eg. Microsoft) nicking the internal design, and putting some code-monkeys on rewriting every line of code?
There's some serious IPRs in the BeOS. No other OS is as pervasively multi-threaded. As a result, no other OS can claim such a good performance ratio (relative performance/processor). Until they get some market share, they've got no protection.
The line "No-one ever got fired for buying the BeOS" hasn't been uttered yet (although it might be true.. virtually no-one HAS bought the BeOS!). Once that's the norm, then if they release the source, people will still buy the BeOS rather than the hypothetically-cloned "Microsoft Windows 2000 Media Edition"... They will have that protection. Some commercial unices currently have that protection. People will still buy them, even though there are cheaper (and free) alternatives. People like Redhat don't _need_ that protection as much, as they haven't made such a massive investment while making little return on that investment.
One alternative would be huge amounts of patent claims. Scyeah.. I'm sure we'd all *love* that. Anyway, most of the stuff they're doing is prior art. It's just they're doing it a) well, and b) in the right combination.
Don't slam the guys for trying to stay afloat. I think the BeOS architecturally is/could be the way future OSes are built.
Re: User Interfaces. Well, unlike Windows, etc, the GUI ain't the only usable interface. You telnet to a BeOS-based machine, and you get 'bash'. It's like IRIX, etc. It's difficult to get rid of the GUI, but the whole machine isn't dependent on the GUI. This way, according to demand, they can easily remove the GUI, or allow it to be switchable. No big deal.
(yikes.. I didn't *mean* to write that much.. honestly.)
Yeah, me too. Can be quite a problem, in my case.
Tom F^HD^H@^H(dammit)Gidden
Unfortunately, it comes with a built in heater
Make that: "Fortunately, it comes with a built in waffle iron"