Unfortunately there are many pieces of Mac OS functionality that aren't available through Cocoa. For example, "aliases". Aliases are the Mac's answer to symbolic links. I not going to go into whether they're good or bad - if you're working with the Mac you will probably have to deal with them at some point. You can't create or resolve aliases without invoking Carbon. OS X has been out for years now and we don't see a lot of functionality migrating out of Carbon and into Cocoa.
I don't care how long it takes to learn. The problem isn't the learning curve - it's that nothing else on the planet besides OS X has good support for it. Not only do you wind up with an app tied to the Cocoa UI framework but all of your code is in a language that doesn't have good support on any other platform.
I just did an app for OS X - we did it in Java/Cocoa. Works fairly well, looks like a good Mac application. And 90% of the code is directly portable to just about any other platform.
Fortunately there are not groups of idiots who run around trying to put cats under the hoods of cars. There are, unfortunately, groups of idiots who think it's a great idea to crash as many computers as possible.
Re:The first ever "bargain" Mac
on
Apple Revises eMac
·
· Score: 2, Funny
But, hey, it's CostCo! If it doesn't work you can return it and get a bunch of beef jerky instead.
Why do you assume that everyone else was "sloppy" and just allocated a big array? If you're going to write a big-int routine it's not that much more difficult to write an arbitrary precision routine than it is to write a "max FOO" precision routine.
I do think that they should have re-adjusted for the misprint, but it's pretty small of you to simply denigrate everyone else's performance. Perhaps they had been exposed to some other ways of working with large numbers than allocating arrays to the max precision. When I was in high school we had a pi calculator we used to run using DEC BASIC's arbitrary precision string arithmetic (this was before high-res graphics or we would probably have been wasting our time on fractals). I'm sure if I'd been given the problem I probably would have done something involving strings which probably would have worked.
I just finished development on a little utility that runs on Mac OS X. Has a nice GUI interface, we think it's easy to use. The GUI and GUI-RELATED code has been 90% of the work.
What's GUI-RELATED you ask? Well, consider a simple file copy utility that runs from the command line. When you run it, you give it the arguments on the command line and off it goes. If you want to use the same arguments twice you make a shell script or an alias - the file copy utility DOES NOT WORRY about persistent arguments.
OK, now we make a GUI file copy utility. Oh, you'd like to use the same arguments again. OK, that's reasonable, let's make a place to store those. Oops, now we need a way to manage them (create, edit, delete - what about concurrent access?). Hey, wait, this is Mac OS - you know, the path to a volume can change when it gets unmounted and remounted. Are you doing the right thing to specify the VOLUME the user wanted and not just a dead path? Oh, what about when the path no longer exists - should we fail, create it silently, pop up a dialog saying it's missing and let the user create it...
John Gruber is dead on the money - you can't just wrap a GUI around a CLI and expect things to be easy to use. There's a whole big layer of foundation code under the GUI that needs to be created to make things work the right way.
This is silly. LinuxToday had already made its fair use - it excerpted the article. The link blocking is just CMP being buttheads - which they have a right to be. It's as though you had the article posted inside your office and asked people who wanted to view it who told them about it and denied access based on that.
The Linux Today editor put it well: I am personally saddened by this move, as I feel it only brings harm to CMP, not Linux Today.
CMP is simply depriving themselves of free referrals. But, they have the right to be buttheads - I'm sure that's in the Constitution somewhere.
Strangely, if one types "Apple Computer" into the Address field of Internet Explorer this is exactly what happens. Mozilla does even better - it takes you directly to the Apple web site!
I don't think the issues are patents. (Don't even start with SCO - their claim to the SMP support in Linux is not based on patents but on their weird derivative code theory. According to this theory, the SMP code that Sequent and IBM developed for their own versions of Unix is a derivative of the original Unix base code and hence SCO owns the rights to it. It's not a patent thing, it's a fraud thing) Multiple companies have done SMP Unix systems. The issue is: it's HARD. The Unix kernel was built with single threadedness in mind. You have to go through and find all the right places and put locks in and then tune the puppy. When SGI went to their full SMP kernel back in the early 90's it took them forever to get it tuned well. I used to work on mini-supercomputers (a weird name, for a weird class) back in the late 80's and we had 8-way multiprocessor machines but we left the kernel single-threaded, again because it was HARD. Also, for scientific computing, you spend most of your time in the application, not in the kernel, so a single threaded kernel isn't a huge hit.
Today, a lot of people have pulled it off, tools are better and a lot of the issues are known. Multi-cpu computers are a readily available to small developers. I wish them luck, I'm sure they'll be able to pull it off, but it's not a matter of just waving the magic SMP wand at the code.
We're not talking about using, we're talking about OWNERSHIP rights. People who don't own things don't get to make policy for the people who do.
People with your attitude and language should go pay their money to Microsoft and cuss out their techs (who at least get paid to listen to your whining. If you want things for free you can at least be CIVIL.
Manufacturers like RealTek are looking for a free ride. They take the work of other people without compensating them in the way that they ask - in this case to distribute the changes made to their code so that they can learn and modify. If the Linux developers asked for cash in return for using their code would you say they should not worry about it? A deal's a deal. RealTek can BUY an OS if they want to keep their changes proprietary or they could use a differently licensed free OS, like BSD. No one is forcing them to use GPL'd code.
Are you a contributor to the Linux code base? If not, then your wishes as a CONSUMER of other people's work don't really count for a lot.
is only about 10%. 1000 KGs = 2200 lbs. 240 kg (or 528 lbs) is actually MORE than 1/4 of an American (short) ton.
Apparently slashdot readers are so dumb that not only do they need to be reminded that a metric ton is 1000 kg but also that a kg = 2.2 lbs and a standard American ton is 2000 lbs.
In a free country, citizens do not need to justify their lawful actions to the government. The government needs to justify its actions to the citizenry. The basic principle is "Everything which is not forbidden is allowed." Explain to me why having large amounts of cash in my house is illegitimate.
I hate to disagree, but laser is one example of a technology that had applications before the technology itself was available!
The applications may have been there but the inventors weren't aware of them. They were doing basic research. Arthur Schawlow, who was one of the inventors of the laser at Bell Labs, said "We thought it might have some communications and scientific uses, but we had no application in mind. If we had, it might have hampered us and not worked out as well."
Well, if there were no market and no clamor it would be called basic research. Often people can't figure out the use for things until after they exist. For example, lasers - when lasers were invented nobody had a good idea of what they would be used for. Today, they're ubiquitous. Likewise, regular LEDs. At one point HP was trying to decide whether they should continue research on LEDs. Marketing said "no - you'll never be able to have them compete with little lightbulbs" Bill Hewlett said "Go do it" and made a huge market for HP
However, in this case, the uses are obvious - back lights for LCD screens come to mind immediately. Replacements for basic lightbulbs as well. LEDs are currently produced as little specks. In order to replace a high wattage bulb you have to team a number of them together. This is expensive. This process would turn out SHEETS of light emitting material. Also, efficiency. Current lightbulbs (and the prototype panel) produce about 15 lumens per watt - they expect to push the technology to 100 lumens per watt. This, coupled with longevity and a low cost to manufacture will drive existing lightbulbs and compact flourescents off the market. There are gaps that exist that the technology is filling
I have a server that I rent from EV1. I've been planning to upgrade and was planning to get another server from them for that. This action has me seriously considering switching our business to another provider. My company uses Open Source software and wants to promote it. Anything that encourages SCO in their anti-Linux/anti-Open Source attacks is directly offensive to us.
I'm willing to take the risk of being sued by SCO. We're small enough that we'd be pretty far down the list. Further, I doubt that any of EV1's customers are large enough to be on SCO's initial list of targets. I don't see this as EV1 "protecting" me. I see this as them encouraging and abetting a criminal company who is out to extort money from me. The more people there are who legitimize SCO and SCO's "licenses" the easier it will be for SCO to pressure small companies into paying cumshaw.
I'm getting the impression that there are a lot of really POOR programmers out who have no idea how to make a program run fast and use C or C++ as a crutch.
No, there's a lot of programmers out there who are insecure and like to say they use C because they think you're more macho if you do. I did C for 10+ years and wrote kernel code for supercomputers (yes, I did have stuff running on a Cray Y-MP, thank you). I got big swinging ones when it comes to software and I feel no need to impress people with my choice of language. I like working in Java because I get a lot more done and spend a lot less time chasing silly bugs. When something needs to be super quick or really close to the metal I'll write some C or some assembler as required. There are very few things that Java's not fast enough for, though, and the ability to get work done faster is where the perfomance lies for me. Also, I can pull in a library to do just about anything at this point...AND IT DOESN"T CRASH THE REST OF MY APP. In C I usually wind up rolling my own because using other people's libraries is a sure road to disaster.
If indeed that is your fear or that is your perception then engage in the democratic process. Back up your argument, back up your belief with facts, marshal evidence in order to convince those who are engaged in the process of governance.
Vinh's attitude is that he is "governing" and that we have to come to him with information to change his mind. He does not view himself as a public servant obviously. It is his job to convince the citizens of the United States (not the "governed of the United States") that he needs the tools he has asked for. It is his job to convince the citizens that hsi approach is correct. We do not need to "convince" those who are currently tasked with governing the country. We need to vote their political masters out and get some people in with better attitudes.
Slashdot Mirror
Unfortunately there are many pieces of Mac OS functionality that aren't available through Cocoa. For example, "aliases". Aliases are the Mac's answer to symbolic links. I not going to go into whether they're good or bad - if you're working with the Mac you will probably have to deal with them at some point. You can't create or resolve aliases without invoking Carbon. OS X has been out for years now and we don't see a lot of functionality migrating out of Carbon and into Cocoa.
I don't care how long it takes to learn. The problem isn't the learning curve - it's that nothing else on the planet besides OS X has good support for it. Not only do you wind up with an app tied to the Cocoa UI framework but all of your code is in a language that doesn't have good support on any other platform.
I just did an app for OS X - we did it in Java/Cocoa. Works fairly well, looks like a good Mac application. And 90% of the code is directly portable to just about any other platform.
Fortunately there are not groups of idiots who run around trying to put cats under the hoods of cars. There are, unfortunately, groups of idiots who think it's a great idea to crash as many computers as possible.
But, hey, it's CostCo! If it doesn't work you can return it and get a bunch of beef jerky instead.
Because it's hard to levitate a significant amount of material?
Nah, we had more fun than that - we shipped it :-)
Why do you assume that everyone else was "sloppy" and just allocated a big array? If you're going to write a big-int routine it's not that much more difficult to write an arbitrary precision routine than it is to write a "max FOO" precision routine.
I do think that they should have re-adjusted for the misprint, but it's pretty small of you to simply denigrate everyone else's performance. Perhaps they had been exposed to some other ways of working with large numbers than allocating arrays to the max precision. When I was in high school we had a pi calculator we used to run using DEC BASIC's arbitrary precision string arithmetic (this was before high-res graphics or we would probably have been wasting our time on fractals). I'm sure if I'd been given the problem I probably would have done something involving strings which probably would have worked.
I just finished development on a little utility that runs on Mac OS X. Has a nice GUI interface, we think it's easy to use. The GUI and GUI-RELATED code has been 90% of the work.
What's GUI-RELATED you ask? Well, consider a simple file copy utility that runs from the command line. When you run it, you give it the arguments on the command line and off it goes. If you want to use the same arguments twice you make a shell script or an alias - the file copy utility DOES NOT WORRY about persistent arguments.
OK, now we make a GUI file copy utility. Oh, you'd like to use the same arguments again. OK, that's reasonable, let's make a place to store those. Oops, now we need a way to manage them (create, edit, delete - what about concurrent access?). Hey, wait, this is Mac OS - you know, the path to a volume can change when it gets unmounted and remounted. Are you doing the right thing to specify the VOLUME the user wanted and not just a dead path? Oh, what about when the path no longer exists - should we fail, create it silently, pop up a dialog saying it's missing and let the user create it...
John Gruber is dead on the money - you can't just wrap a GUI around a CLI and expect things to be easy to use. There's a whole big layer of foundation code under the GUI that needs to be created to make things work the right way.
Don't you get those from the Franklin Mint?
This is silly. LinuxToday had already made its fair use - it excerpted the article. The link blocking is just CMP being buttheads - which they have a right to be. It's as though you had the article posted inside your office and asked people who wanted to view it who told them about it and denied access based on that.
The Linux Today editor put it well: I am personally saddened by this move, as I feel it only brings harm to CMP, not Linux Today.
CMP is simply depriving themselves of free referrals. But, they have the right to be buttheads - I'm sure that's in the Constitution somewhere.
Strangely, if one types "Apple Computer" into the Address field of Internet Explorer this is exactly what happens. Mozilla does even better - it takes you directly to the Apple web site!
I don't think the issues are patents. (Don't even start with SCO - their claim to the SMP support in Linux is not based on patents but on their weird derivative code theory. According to this theory, the SMP code that Sequent and IBM developed for their own versions of Unix is a derivative of the original Unix base code and hence SCO owns the rights to it. It's not a patent thing, it's a fraud thing) Multiple companies have done SMP Unix systems. The issue is: it's HARD. The Unix kernel was built with single threadedness in mind. You have to go through and find all the right places and put locks in and then tune the puppy. When SGI went to their full SMP kernel back in the early 90's it took them forever to get it tuned well. I used to work on mini-supercomputers (a weird name, for a weird class) back in the late 80's and we had 8-way multiprocessor machines but we left the kernel single-threaded, again because it was HARD. Also, for scientific computing, you spend most of your time in the application, not in the kernel, so a single threaded kernel isn't a huge hit.
Today, a lot of people have pulled it off, tools are better and a lot of the issues are known. Multi-cpu computers are a readily available to small developers. I wish them luck, I'm sure they'll be able to pull it off, but it's not a matter of just waving the magic SMP wand at the code.
We're not talking about using, we're talking about OWNERSHIP rights. People who don't own things don't get to make policy for the people who do.
People with your attitude and language should go pay their money to Microsoft and cuss out their techs (who at least get paid to listen to your whining. If you want things for free you can at least be CIVIL.
Manufacturers like RealTek are looking for a free ride. They take the work of other people without compensating them in the way that they ask - in this case to distribute the changes made to their code so that they can learn and modify. If the Linux developers asked for cash in return for using their code would you say they should not worry about it? A deal's a deal. RealTek can BUY an OS if they want to keep their changes proprietary or they could use a differently licensed free OS, like BSD. No one is forcing them to use GPL'd code.
Are you a contributor to the Linux code base? If not, then your wishes as a CONSUMER of other people's work don't really count for a lot.
According to the Bloomberg article, Olympus and Sanyo are both licensees of the patents in question.
is only about 10%. 1000 KGs = 2200 lbs. 240 kg (or 528 lbs) is actually MORE than 1/4 of an American (short) ton.
Apparently slashdot readers are so dumb that not only do they need to be reminded that a metric ton is 1000 kg but also that a kg = 2.2 lbs and a standard American ton is 2000 lbs.
In a free country, citizens do not need to justify their lawful actions to the government. The government needs to justify its actions to the citizenry. The basic principle is "Everything which is not forbidden is allowed." Explain to me why having large amounts of cash in my house is illegitimate.
I hate to disagree, but laser is one example of a technology that had applications before the technology itself was available!
The applications may have been there but the inventors weren't aware of them. They were doing basic research. Arthur Schawlow, who was one of the inventors of the laser at Bell Labs, said "We thought it might have some communications and scientific uses, but we had no application in mind. If we had, it might have hampered us and not worked out as well."
That's a lot better than your average tungsten filament lightbulb.
Well, if there were no market and no clamor it would be called basic research. Often people can't figure out the use for things until after they exist. For example, lasers - when lasers were invented nobody had a good idea of what they would be used for. Today, they're ubiquitous. Likewise, regular LEDs. At one point HP was trying to decide whether they should continue research on LEDs. Marketing said "no - you'll never be able to have them compete with little lightbulbs" Bill Hewlett said "Go do it" and made a huge market for HP
However, in this case, the uses are obvious - back lights for LCD screens come to mind immediately. Replacements for basic lightbulbs as well. LEDs are currently produced as little specks. In order to replace a high wattage bulb you have to team a number of them together. This is expensive. This process would turn out SHEETS of light emitting material. Also, efficiency. Current lightbulbs (and the prototype panel) produce about 15 lumens per watt - they expect to push the technology to 100 lumens per watt. This, coupled with longevity and a low cost to manufacture will drive existing lightbulbs and compact flourescents off the market. There are gaps that exist that the technology is filling
No, MkLinux was an official project within the OS team and managed by the manager of the NuKernel group. Everyone on the OS team thought it was cool.
I have a server that I rent from EV1. I've been planning to upgrade and was planning to get another server from them for that. This action has me seriously considering switching our business to another provider. My company uses Open Source software and wants to promote it. Anything that encourages SCO in their anti-Linux/anti-Open Source attacks is directly offensive to us.
I'm willing to take the risk of being sued by SCO. We're small enough that we'd be pretty far down the list. Further, I doubt that any of EV1's customers are large enough to be on SCO's initial list of targets. I don't see this as EV1 "protecting" me. I see this as them encouraging and abetting a criminal company who is out to extort money from me. The more people there are who legitimize SCO and SCO's "licenses" the easier it will be for SCO to pressure small companies into paying cumshaw.
If you want point n' click Unix buy a Mac and stop whining. If you want to be cheap, RTFM.
I'm getting the impression that there are a lot of really POOR programmers out who have no idea how to make a program run fast and use C or C++ as a crutch.
No, there's a lot of programmers out there who are insecure and like to say they use C because they think you're more macho if you do. I did C for 10+ years and wrote kernel code for supercomputers (yes, I did have stuff running on a Cray Y-MP, thank you). I got big swinging ones when it comes to software and I feel no need to impress people with my choice of language. I like working in Java because I get a lot more done and spend a lot less time chasing silly bugs. When something needs to be super quick or really close to the metal I'll write some C or some assembler as required. There are very few things that Java's not fast enough for, though, and the ability to get work done faster is where the perfomance lies for me. Also, I can pull in a library to do just about anything at this point...AND IT DOESN"T CRASH THE REST OF MY APP. In C I usually wind up rolling my own because using other people's libraries is a sure road to disaster.
From the article:
If indeed that is your fear or that is your perception then engage in the democratic process. Back up your argument, back up your belief with facts, marshal evidence in order to convince those who are engaged in the process of governance.
Vinh's attitude is that he is "governing" and that we have to come to him with information to change his mind. He does not view himself as a public servant obviously. It is his job to convince the citizens of the United States (not the "governed of the United States") that he needs the tools he has asked for. It is his job to convince the citizens that hsi approach is correct. We do not need to "convince" those who are currently tasked with governing the country. We need to vote their political masters out and get some people in with better attitudes.