It doesn't break too many command line tools, because few programs are set up expecting that there are two files named identically except for case. Programs just exhibit slightly different behavior. For example FILE *f = fopen("Readme.TXT", "r") might return a file handle on OS-X when it would return NULL on other Unix systems. For it to cause an actual problem, someone would have to do something like try to preflight the filename by calling readdir() and comparing filenames themselves.
The biggest problem that I can think of is tar, which will overwrite one file with another if the archive contains two files identical except for case.
It probably should, but there are other major Unix systems that have similar screwy naming systems.
System V,
System V release 2,
System V Release 3,
System V release 3.2,
System V release 4.
I would much rather have MacOS X now and wait for DVD functionality.
This is assuming that DVD player development and OS development have to happen in in series. Apple could have farmed off the development of a DVD player to some sort of outside shop. (probably one of the old NeXT developers.)
The interface is serial. The four important wires are the data, clk, pwr and gnd. The clock and the data bidirectional, because the keyboard and the computer both send and receive data. The clock is really more of an ACK signal, it gets toggled by the receiver when it has successfully read that current status of the data line. Its a little difficult (but not impossible) to reverse engineer, since looking at the keyboard disconnected from the computer won't give you any output (since no one is toggling the clk.) and when connected it gets confusing who is toggling what.
The AT and the PS/2 interface is the same, electrically, there are just different connectors.
You have the right to distribute your own original software any way you please, but you have absolutely no right to tell anyone else how they can distribute theirs.
The best way I've found of explaining the viral nature of theGPL to economically minded people is to expain it as an alternate form of renumeration.
If someone believe that software can be morally owned, they believe that the right to use that software can be sold or bartered. Its easy to find source code that people are selling to other developers. What do they want for it? It depends on what the author wants. It ranges from a case of beer to thousands of dollars. The author has a right to ask for whatever he wants. The potential recipiant can then accept the offer and use the source code or reject the offer and not use it.
One way of viewing the GPL is that in exchange for the license to to use the code, the author is asking for a similar license in return. The recipient of the code has a choice. Accept the license terms and use the code or reject the terms. Some people find the price of the GPL too high to pay, some find it a bargin. The purpose of the GPL is orthoginal to the whole deal.
The fact that copyright is granted on the creation of a work is a fairly recent change the copyright law. It was the Berne convention that changed that in the US.
The is a difference in protection allowed to the copyright owner of a work if they file an application with the copyright office. In cases of infringement, copyright owners that have applications on file can sue for punitave damages, where copyright owners who haven't filed can only sue to suspend further infringment.
There are GUI designs that don't prevent programatic interaction. One example is Apple's recommendation on how to implement Mac OS's Apple Events. A well factored application has its actions divided into fairly concrete set of verbs, and then some special nouns for them to interact with. Finally when it gets time to develop the UI portion of the application, the programmer hooks up the menus so that it sends Apple Events back to itself. A program designed this way can then be manipulated programatically with just as much functionality as with manual manipulation.
Before someone pipes up about the deficiancies of Applescript, I just want to point out that Applescript is just one implementation of a system that uses Apple Events to interact with other programs. Its designed to be overly verbose and "English-like" to allow people to read scripts that they might not have the ability to write on their own. There are other languages that can send Apple Events to correctly factored programs. Frontier and Perl come to mind.
TCL/Tk was originally designed for a similar type of interaction. The original idea for TCL was for it to be a language embedded into an application. The functionality of the program could be coded up as new TCL commands, and then small scripts could be written to bundle the functionality together. Tk was then an additional set of TCL commands that implemented a GUI toolkit which could interect with these extended TCL commands. A program built up as a set of TCL commands could be run either graphically as a Tk app or non-graphically through TCL. Then you can add to that the Tk "send" command which allows you to pass TCL commands to a running Tk app, allowing you to programatically interact with programs that are currently running graphically.
It doesn't need to be graphical, but doing it that way is a convenient way of getting all sort of contextual clues accross with a minimal amount of screen realestate and a minimal amount of recognition by the user.
Radio buttons or popup menus for exclusive selections, checkboxes for non-exclusive ones, text boxes for optional arguments and file dialogs for files. It gets the point accross retty easily and quickly.
Another point about commando that hasn't been mentioned so far is that it had a text area on the bottom that built the command for you as you edited the UI elements. This kept it as being a learning tool instead of a crutch.
I don't quite see the advantages of your smartshell alternative. The tab completion requires a smart terminal, so it already throws the unix stream paradigm out the window. It wouldn't be a mouse vs. keyboard thing, since MPW made extensive use of keyboard shortcuts.
I guess, I'd suggest taking a look at they way commando is set up before you start suggesting "better" alternatives.
The copyright laws have exceptions for parody, not for satire. I vaguely remember a court case in which the defendant was found infringing because it was satire and not parody, but the closest I could find was the quote from Justice Souter Copyright, Fair Use, and the Law, an article about fair use for musicians. In it he seems to imply that satire might have some fair use protection, but not as strongly. For satire, it has to be shown why the artist needed to infringe in order to make their work.
I don't think that it could be said that protocol interoperability sessions could be seen as a parody of food baking competitions. They aren't doing it to make fun of pilsbury style competitions.
Trademarks and copyrights do not have the thorough examination that patents (supposedly) have. They are just that certify the date and time of a submission and then the parties have to go to the courts for a determination on what is infringing.
The -off suffix appended to anything in order to signify competition has a history long before 1971. There have been BBQ bakeoff's about as long as this country has been cooking pig ribs coated with sauce. There were dance-offs in the '30s to '50s. A party might be able to show that -off is a normal part of the language and that trademarking bake-off is about as valid as trademarking a word with the prefix re to mean again. As another example, when I ran a previous version of this article through a spelling checker, it turned up trademarking as an invalid word. Could I trademark the word trademarking as the act of applying for a trademark?
But you claim that they've successfully defended the trademark before. That is a bit worrisome, but it all depends on what they've done to defend it. If it has just been sending cease and desist letters and having their victims fold, thats one thing. A court decision is something else.
As for a "public relations fiasco", they've been susceptable to it before. Do you remember Ben and Jerry's "What's the Doughboy afraid of" ad campaign? Who is more sympathetic, the doughboy? or a couple of hippies trying to sell extra fat ice cream during the height of the '80s health food craze.
Archiving everything, including public free texts, sounds like what Barry Shein was or is trying to do with the Online Book Initiative
I don't know how much work he is actively putting into it. It seems to have some recent additions, there is a folder called Election2000, but it only has one realaudio file in it.
The BASIC your describing isn't the BASIC I was using in the '80s. Mine had stronly typed variables, record datatypes, separate compilation, and block conditional and loops. Digital did a good job with VAX BASIC. The BASIC you're describing were the variants designed to run on the 4KB microcomputers of the day, variants that make Kemeny and Kurtz cringe.
I think you're right in that Perl will evolve. Its intentionally derivative nature makes it easy to modify. It already has changed greatly between perl1 and perl5, and since work is already being done on perl 6, it seems to continue on its same path.
Although Byte had a large number of readers at the time CMP stopped publication, it was loosing advertisers. One could argue that it was loosing advertisers because of its editorial focus. It had articles about everything from software development, to chip design, to computer system and peripherals, to software applications reviews.And all in a relitivly OS neutral viewpoint. Developers of applications didn't want to advertise in a magazine about CPU design. OS vendors didn't want to advertise in a magazine reviewing hardware their product doesn't run under. CPU manufactures didn't want to advertise
in a magazine discussing OOP.
There are a few web pages discussing the end of Byte magazine, including Tom Halfill's Tom's Unofficial BYTE FAQ: The Death of BYTE Magazine
and Jerry Pournelle's The Byte Fiasco
Could this happen the TPJ too? Jon implies that the magazine was financially strong when he sold it to EarthWeb, but advertisers may have left in the past year. (Take a look at your average large web site, and see how many "house ads" are running.) I know there are some advertisers that want to target an audience who use perl, but is it enough. Take a look at other computer magazines, and see how many ads are there for compilers and libraries. Since the perl interpreter and CPAN are both free, it takes a big hit in those advertising sectors.
I started working for a particular company in March 1999. I can think of about a dozen cases where the tm_year field of localtime was improperly used. About half of them used the year directly for output. A few others compared the tm_year field to some arbitrary year (80 or 90 were common) and added 2000 to it. There were a few that padded a stringified version the tm_year field with an extra zero if the numeric version was less than 10.
None of these were Y2K corrections, these were all the original code written by a bunch of people who had incrrect assumptions about localtime(3)'s specification.
One of the biggest offenders of the misuse of localtime was a 50-ish guy. At one time he claimed that he didn't know what the big deal about y2k was. that he first noticed the problem with mortgage applications he was doing 30 years ago. That was a little scary, a co-worker admitting to 30 years of Y2K bugs.
Thats the whole point of having time be an abstract type like time_t rather than a specific size. The real 2038 problem is not in time_t, but rather in programs that use the "%d" specifier of printf() to write out a time_t(Of course thats also why ANSI refused to give exact sizes to char,int,and long. But then of course, we wind up with long long because some fool is worried about backwards compatibility.)
There was a Macintosh developer, I think it was Symetry but it might have been Nisus, that used piracy as a marketing gimmick. They held an amnesty promotion for people who had pirated version of their product. The deal basically was that if you sent a screenshot of a copy of their product that you hadn't purchased, you could buy the product for some amount that as probably around their upgrade promotion price.
I'm not sure if they did that to try to guage the amount of piracy that was really going on with their product, just for the publicity, or some other reason.
LucasArts specifically were targetting a younger market. I was working in a software store around the mid '80s and one of the selling points the LucasArts marketing people wanted the retail sales people to stress was the lack of dying. It was supposed to be "less fustrating to the child"
(Now that I think of it, the marketing stuff was obviously aimed at the parent buying software for the child.)
For Infocom games, the single most important skill was learning to save often, and save many different scenarios in different files. I wish that I had RCS back then. I had to mimic the branches of saved games with file name convensions.
But the only consistancy Windows has had since day one has been its name. The API has radically changed over the years. You are right, the marketing has been great. How else does someone need to convince the public that continual ground up rewrites are the magnificent evolvement of a product line.
Unix on the other hand has been extended, but its core concepts have not been removed. It is a testament to good design.
Apple used to have a pretty good policy about upgrades. The only problem was that they were originally on the pricy side.
My Mac 512 got upgraded to a 512E and then a Mac Plus towards the end of the Plus' product life. (when the price went down.) My IIcx later became a IIci. I seriously thought of getting the IIci to Quadra 700 upgrade, but never brought it to that point. Just this year I changed the LaserWriter IINT I rescued from a dumpster into a IIf.
It was soon after that point where the upgradable models became fewer. By now, they've stopped completely. But this is a fairly recent development, not how it "has always been".
Yes, that was one of the original plans for Rhapsody, but Apple moved away from that strategy in May of 1998.
Apple's original Rhapsody plan didn't get warmly received by some of the major cross platform developers, (Microsoft, Macromedia, Adobe, etc.) Rhapsody allowed easy cross-platform development. (with its yellowbox DLL's for windows), but these companies already had a cross platform development strategy. If they had to throw away their old MacOS code to work on the new Mac platform, they probably wouldn't have replaced it with OS-X's frameworks, they would have dropped mac support entirely.
So Apple came up with the idea for Carbon, a MacOS subset that could run on OS-X. And since they had this strategy to keep the cross platform developers through MacOS support, there was little need for Apple to offer the yellowbox on intel DLLs to coax Windows developers.
But the way that the features were added to Unix, they are just hacks, and appear that way, rather than an integral part of the architecture.
Take a look at
the way multics handled dynamic linking. Calling a non-existant symbol caused the process to suspend. Someone could write
a replacement for the missing subroutine and resume the process.
Multics had a hierarchical administration system, unix has sudo(1).
Multics was designed for large systems. Unix was designed for small systems and grew large.
Apple may have some information about it on their OS X Developer Documentation page. I notice a mention on page 17 of their document Kernel Environment. In a traditional microkernel, the box marked "BSD" would be outside the bluish grey box marked "Kernel Environment".
Saying moved the BSD kernel into the same kernel space as the Mach microkernel. isn't quite the same was saying moved FreeBSD into
the Mach kernel, though.
There are two major differences between OS X and the Hurd.
The first difference is that OS X is a single server and the Hurd is a multi-server. That is, on OS X all the Mach stuff communicates with is one large Free-BSD kernel with its hardware dependent stuff ripped out. The Hurd on the other hand has each system call handled by a separate thread of execution.
The second difference is that Apple moved the BSD kernel into the same kernel space as the Mach microkernel. This means that they don't have the context switching overhead that traditional Mach based systems have.
Slashdot Mirror
The biggest problem that I can think of is tar, which will overwrite one file with another if the archive contains two files identical except for case.
Morale level? I think the morale of my companies engineering department would go up if there were fewer people in marketing.
It probably should, but there are other major Unix systems that have similar screwy naming systems. System V, System V release 2, System V Release 3, System V release 3.2, System V release 4.
This is assuming that DVD player development and OS development have to happen in in series. Apple could have farmed off the development of a DVD player to some sort of outside shop. (probably one of the old NeXT developers.)
The interface is serial. The four important wires are the data, clk, pwr and gnd. The clock and the data bidirectional, because the keyboard and the computer both send and receive data. The clock is really more of an ACK signal, it gets toggled by the receiver when it has successfully read that current status of the data line. Its a little difficult (but not impossible) to reverse engineer, since looking at the keyboard disconnected from the computer won't give you any output (since no one is toggling the clk.) and when connected it gets confusing who is toggling what.
The AT and the PS/2 interface is the same, electrically, there are just different connectors.
The best way I've found of explaining the viral nature of theGPL to economically minded people is to expain it as an alternate form of renumeration.
If someone believe that software can be morally owned, they believe that the right to use that software can be sold or bartered. Its easy to find source code that people are selling to other developers. What do they want for it? It depends on what the author wants. It ranges from a case of beer to thousands of dollars. The author has a right to ask for whatever he wants. The potential recipiant can then accept the offer and use the source code or reject the offer and not use it.
One way of viewing the GPL is that in exchange for the license to to use the code, the author is asking for a similar license in return. The recipient of the code has a choice. Accept the license terms and use the code or reject the terms. Some people find the price of the GPL too high to pay, some find it a bargin. The purpose of the GPL is orthoginal to the whole deal.
The fact that copyright is granted on the creation of a work is a fairly recent change the copyright law. It was the Berne convention that changed that in the US.
The is a difference in protection allowed to the copyright owner of a work if they file an application with the copyright office. In cases of infringement, copyright owners that have applications on file can sue for punitave damages, where copyright owners who haven't filed can only sue to suspend further infringment.
There are GUI designs that don't prevent programatic interaction. One example is Apple's recommendation on how to implement Mac OS's Apple Events. A well factored application has its actions divided into fairly concrete set of verbs, and then some special nouns for them to interact with. Finally when it gets time to develop the UI portion of the application, the programmer hooks up the menus so that it sends Apple Events back to itself. A program designed this way can then be manipulated programatically with just as much functionality as with manual manipulation.
Before someone pipes up about the deficiancies of Applescript, I just want to point out that Applescript is just one implementation of a system that uses Apple Events to interact with other programs. Its designed to be overly verbose and "English-like" to allow people to read scripts that they might not have the ability to write on their own. There are other languages that can send Apple Events to correctly factored programs. Frontier and Perl come to mind.
TCL/Tk was originally designed for a similar type of interaction. The original idea for TCL was for it to be a language embedded into an application. The functionality of the program could be coded up as new TCL commands, and then small scripts could be written to bundle the functionality together. Tk was then an additional set of TCL commands that implemented a GUI toolkit which could interect with these extended TCL commands. A program built up as a set of TCL commands could be run either graphically as a Tk app or non-graphically through TCL. Then you can add to that the Tk "send" command which allows you to pass TCL commands to a running Tk app, allowing you to programatically interact with programs that are currently running graphically.
It doesn't need to be graphical, but doing it that way is a convenient way of getting all sort of contextual clues accross with a minimal amount of screen realestate and a minimal amount of recognition by the user.
Radio buttons or popup menus for exclusive selections, checkboxes for non-exclusive ones, text boxes for optional arguments and file dialogs for files. It gets the point accross retty easily and quickly.
Another point about commando that hasn't been mentioned so far is that it had a text area on the bottom that built the command for you as you edited the UI elements. This kept it as being a learning tool instead of a crutch.
I don't quite see the advantages of your smartshell alternative. The tab completion requires a smart terminal, so it already throws the unix stream paradigm out the window. It wouldn't be a mouse vs. keyboard thing, since MPW made extensive use of keyboard shortcuts.
I guess, I'd suggest taking a look at they way commando is set up before you start suggesting "better" alternatives.
I don't think that it could be said that protocol interoperability sessions could be seen as a parody of food baking competitions. They aren't doing it to make fun of pilsbury style competitions.
The -off suffix appended to anything in order to signify competition has a history long before 1971. There have been BBQ bakeoff's about as long as this country has been cooking pig ribs coated with sauce. There were dance-offs in the '30s to '50s. A party might be able to show that -off is a normal part of the language and that trademarking bake-off is about as valid as trademarking a word with the prefix re to mean again. As another example, when I ran a previous version of this article through a spelling checker, it turned up trademarking as an invalid word. Could I trademark the word trademarking as the act of applying for a trademark?
But you claim that they've successfully defended the trademark before. That is a bit worrisome, but it all depends on what they've done to defend it. If it has just been sending cease and desist letters and having their victims fold, thats one thing. A court decision is something else.
As for a "public relations fiasco", they've been susceptable to it before. Do you remember Ben and Jerry's "What's the Doughboy afraid of" ad campaign? Who is more sympathetic, the doughboy? or a couple of hippies trying to sell extra fat ice cream during the height of the '80s health food craze.
I don't know how much work he is actively putting into it. It seems to have some recent additions, there is a folder called Election2000, but it only has one realaudio file in it.
I think you're right in that Perl will evolve. Its intentionally derivative nature makes it easy to modify. It already has changed greatly between perl1 and perl5, and since work is already being done on perl 6, it seems to continue on its same path.
There are a few web pages discussing the end of Byte magazine, including Tom Halfill's Tom's Unofficial BYTE FAQ: The Death of BYTE Magazine and Jerry Pournelle's The Byte Fiasco
Could this happen the TPJ too? Jon implies that the magazine was financially strong when he sold it to EarthWeb, but advertisers may have left in the past year. (Take a look at your average large web site, and see how many "house ads" are running.) I know there are some advertisers that want to target an audience who use perl, but is it enough. Take a look at other computer magazines, and see how many ads are there for compilers and libraries. Since the perl interpreter and CPAN are both free, it takes a big hit in those advertising sectors.
None of these were Y2K corrections, these were all the original code written by a bunch of people who had incrrect assumptions about localtime(3)'s specification.
One of the biggest offenders of the misuse of localtime was a 50-ish guy. At one time he claimed that he didn't know what the big deal about y2k was. that he first noticed the problem with mortgage applications he was doing 30 years ago. That was a little scary, a co-worker admitting to 30 years of Y2K bugs.
Thats the whole point of having time be an abstract type like time_t rather than a specific size. The real 2038 problem is not in time_t, but rather in programs that use the "%d" specifier of printf() to write out a time_t(Of course thats also why ANSI refused to give exact sizes to char,int,and long. But then of course, we wind up with long long because some fool is worried about backwards compatibility.)
I'm not sure if they did that to try to guage the amount of piracy that was really going on with their product, just for the publicity, or some other reason.
LucasArts specifically were targetting a younger market. I was working in a software store around the mid '80s and one of the selling points the LucasArts marketing people wanted the retail sales people to stress was the lack of dying. It was supposed to be "less fustrating to the child"
(Now that I think of it, the marketing stuff was obviously aimed at the parent buying software for the child.)
For Infocom games, the single most important skill was learning to save often, and save many different scenarios in different files. I wish that I had RCS back then. I had to mimic the branches of saved games with file name convensions.
Unix on the other hand has been extended, but its core concepts have not been removed. It is a testament to good design.
My Mac 512 got upgraded to a 512E and then a Mac Plus towards the end of the Plus' product life. (when the price went down.) My IIcx later became a IIci. I seriously thought of getting the IIci to Quadra 700 upgrade, but never brought it to that point. Just this year I changed the LaserWriter IINT I rescued from a dumpster into a IIf.
It was soon after that point where the upgradable models became fewer. By now, they've stopped completely. But this is a fairly recent development, not how it "has always been".
Apple's original Rhapsody plan didn't get warmly received by some of the major cross platform developers, (Microsoft, Macromedia, Adobe, etc.) Rhapsody allowed easy cross-platform development. (with its yellowbox DLL's for windows), but these companies already had a cross platform development strategy. If they had to throw away their old MacOS code to work on the new Mac platform, they probably wouldn't have replaced it with OS-X's frameworks, they would have dropped mac support entirely.
So Apple came up with the idea for Carbon, a MacOS subset that could run on OS-X. And since they had this strategy to keep the cross platform developers through MacOS support, there was little need for Apple to offer the yellowbox on intel DLLs to coax Windows developers.
So no intel based frameworks. Unfortunatly.
Take a look at the way multics handled dynamic linking. Calling a non-existant symbol caused the process to suspend. Someone could write a replacement for the missing subroutine and resume the process.
Multics had a hierarchical administration system, unix has sudo(1).
Multics was designed for large systems. Unix was designed for small systems and grew large.
Apple may have some information about it on their OS X Developer Documentation page. I notice a mention on page 17 of their document Kernel Environment. In a traditional microkernel, the box marked "BSD" would be outside the bluish grey box marked "Kernel Environment".
Saying moved the BSD kernel into the same kernel space as the Mach microkernel. isn't quite the same was saying moved FreeBSD into the Mach kernel, though.
There are two major differences between OS X and the Hurd.
The first difference is that OS X is a single server and the Hurd is a multi-server. That is, on OS X all the Mach stuff communicates with is one large Free-BSD kernel with its hardware dependent stuff ripped out. The Hurd on the other hand has each system call handled by a separate thread of execution.
The second difference is that Apple moved the BSD kernel into the same kernel space as the Mach microkernel. This means that they don't have the context switching overhead that traditional Mach based systems have.
The good point is that the user space hurd daemons can be overridden per process. Kernel modules insert functionality for the whole system.