Slashdotters don't always have a consistant opinion on reverse engineering (which is natural given the large number of us with differing opinions.) It wouldn't be the first time a vocal group opposed reverse engineering though - when Linus threw a hissy fit because Tridge had "reverse engineered" (ie telnetted in and typed "help") BitKeeper, with BitKeeper's control-freak creators terminating kernel developer's gratis usage as a result, there were enormous numbers of Slashdotters who sided with Torvalds. Indeed, he still seems to be surprisingly popular as an "open source" figure despite employing a policy for a while that was openly hostile to Free Software.
So no, if you can't even telnet to a proprietary repository for a free software project and type "help" without a large percentage of Slashdotters protesting, you can't really assume that Slashdotters are, as a group, pro-reverse-engineering. This is one of those things with vocal proponents on both sides.
I'm finding this highly improbable. I'm not saying you don't believe that's what they said, but there has to be more to it than that.
Back in 1998, I visited the US for the first time (I'm British.) I needed more cash, so I went to an ATM in the middle of Boston, put in my card, and withdrew some money. When I came back to Britain, and got my next bank statement, the charge showed up. Which is what you'd expect.
I'm finding it just a little bit difficult to believe that this would have been possible if the ATM had to search through a database containing EVERY BANK CARD IN THE WORLD, essentially made up of arbitrary card numbers, to find out which bank account my card refered to. I can't imagine why anyone would implement something that likely to be the victim of database synchronization and duplicate number errors.
It's notable that there is an official format for financial cards which works the way most of us would expect such a thing to work, identifying features such as account numbers and institution dependent features.
Xorg has its drivers. They're woefully out of date (X1xx series cards being completely unsupported) and relatively few contain 3D support. And fglrx is everything that'd bad about proprietary software. My laptop (a Thinkpad T60) can't even boot into X using fglrx if it's running from battery power (you just get a blank screen.) Use "totem" and "Unreal Tournament 2004" often enough, and one will stop working - if you're lucky. If you're unlucky, you'll get enough kernel panics to require a full-on reboot.
I'm honestly wondering if I'd been better off going the GMA950 route. Intel's drivers aren't completely Free Software, but there's at least a cleaner design in there.
I've said it before, but I really think the FSF needs to do more than just fund reverse engineering efforts. While ATI may hate these efforts, they make ATI products more valuable and ultimately help ATI. A concerted campaign to raise capital to start a rival graphics card manufacturer strikes me as a better solution if this is possible. I'm aware of open source graphics card efforts (involving programming FPGAs et al) but I think the next step needs to be taken. I would invest in it.
Yeah, damn it, if they hadn't put massive restrictions on owning guns in Britain seventy years ago, Britain wouldn't be seeing a slight increase in one type of crime today!
You do know that the probability of a criminal finding their victim was armed hasn't changed one iota in the last few decades?
You and Timothy need to step away from the NRA propaganda. Despite high profile "bans" on handguns in the last decade (and Kalashnikovs the decade before), realistically, few people in Britain had access to them even before these bans. There has been no decrease in deterence against the usual group of criminals supposedly fearful of guns because of any reduction in gun circulation.
Given the importance of OpenGL, and SGI's relunctance to offer a permanent royalty-free class-license for its patents in OpenGL, this could actually have some fairly dire consequences.
A company could legally buy SGI's patents, turn around, and start requiring license fees from OpenGL users.
SGI may be an irrelevent hardware company to most people, but it does have potential when seen in terms of the IP it controls.
Not to mention the fact this is supposedly being announced at the WWDC.
The WWDC. World Wide Developers Conference. DEVELOPERS CONFERENCE.
I expect Jobs to be running around showing the new features of Leopard, but I highly doubt that a movie rental service is going to be one of them. The only people that will be able to make use of the service are those with Leopard? That makes no sense whatsoever. Ok, well, he'd be announcing a new feature of iTunes at the WWDC? In God's name, why? What next? Maybe he'll do a presentation on leather iPod cases there too.
And then, perhaps, at MacWorld SF next year, he can spend half the presentation talking about the great new features in Xcode.
This is a ridiculous rumour from an increasingly irrelevent rumours site. The basic concept's nice and workable, but it's clear the rumour is, actually, completely made up. It carries about as much wieght as my prediction that Apple is planning to put Core 2 Duos in the iMac. I mean, yeah, that would make lots of sense, but that doesn't mean I actually have been told this by anyone in the know. Further, if I add that Apple's intent is to call the new 64 bit iMac the "CrapMac", it should become fairly obvious I know bugger all.
Yet, apparently, Apple is going to announce a movie downloads system at a developer's conference. Not a special event, like it did the iTunes Music Store. Or the leather cases and iPod speakers. No. This will be told to a bunch of developers who are actually interested in what Apple's new operating system will be like, because, like, that'll not piss them off at all, and everyone knows that all the cool journalists hang out at the WWDC.
Never underestimate what you can do in a high level language if its responsible for generating its own assembler code. I can actually see a few ways of implementing a Java like language (whose compiler generates assembler code) that can implement its own GC in high level code.
And that's ignoring the obvious possibility of giving your garbage collection a big ass "byte[] ProcessMemory" object to work with.
For example, the base Object type can have some base methods including "long AddressOf", "long SizeOf", "void RelocateTo(long)", and "Object[] GetObjectsIReferTo". Add to that an object that abstracts the stack, and ensure your generated code conforms to certain conventions so object addresses are only ever located in certain positions on the stack or certain registers, or in objects themselves.
Now, at this point this is pretty low level. You're still working with addresses. And there's still the realistic possibility of screwing up and causing object corruption. You can, of course, simply prevent other classes from using the methods available to the garbage collector at a low level. But if you're really smart, you can do the above without the AddressOf or RelocateTo things using addresses. You ensure you know the order of the objects on the heap, and replace RelocateTo with some form of "DropDown" object. You also abstract the heap (or heaps, in the case of more recent Javas whose GC strategy involves three heaps.)
That's just the beginning, but you can probably figure out from the above that an unmanaged language (I hate the term "low level language" when applied to C. Stop redefining terms people! C is a high level language) or some very convoluted, self contained, assembly is a necessary component of the implementation of a managed language.
1. Archive source locally and offer to sell (for $10 or less) CDs for archived versions.
2. Use SourceForge or Savannah rather than your own personal website for distributing source and binaries.
3. Include the source code with the binaries in the same package.
Pretty easy really. Unless your software package is extraordinarily big, (3) is probably the easiest. (And yes, SF is on the hook, not you, if they lose the source. They're the ones redistributing the binary.)
Excellent point. Further, it could be argued that if a company can, before making a hiring decision, determine that it's better to hire two people for $X who'll live a particular standard of living than it is to hire one person for 2x$X living the same standard of living, then morally and economically, the company is making the right decision by chosing the first (the "outsourcing" option.)
Firing someone merely to give someone else a job is, obviously, a different matter.
We've also created entirely new fields of work, and we've made it possible for new industries to be created. Our technologies make things possible, making products and services possible, that could never be dreamt of without our help.
I'm not going to say I'm ticked off at outsourcing, because it's a zero-sum game at the end of the day, one person loses a job, another gains it. But I'm not going to say I agree with you that those complaining are engaging in "doublespeak". They're not. This is a fantastic industry that has created far, far, more than its obsoleted, and will continue to do so for decades to come.
FWIW, this usually comes up, and it's worth pointing out a few things.
1. Most people know that WINE is a reimplementation of an API, not a CPU emulator. I know a handful of people get confused here, but the majority, in my experience, do not.
2. Historically, the word emulator has, actually, included programs that simulate operating systems, not just programs that simulate CPUs. For example, the Atari ST and Commodore Amiga had several systems marketed as "Macintosh emulators". The non-hardware versions of each merely patched the Macintosh ROMs to use local hardware rather than Mac hardware. So called "CP/M" emulators for V20/V30 series CPUs (an NEC version of the 8086 with native 8080 code execution support) were also widely available. These are by far not the only things refered to as "emulators" that were operating system, not CPU, emulators.
3. The debate over the term is so commonplace, that at one time WINE stood for, officially, both as "Wine Is Not an Emulator" and "WINdows Emulator". This is perfectly legitimate, it isn't a CPU emulator, but it does emulate Windows. A quick look-up of the dictionary definition of the word "emulate" will put you in good stead here as to why both terms are legimate.
The word "emulator" does not mean "CPU emulator", it merely means "Something that emulates". Whether you're emulating (immitating) a CPU, or an operating system, you are, actually, an emulator. Jumping in with a correction over the word emulator because it's not a CPU emulator is entirely unnecessary, and largely wrong. I don't doubt your motives, unfortunately usually when someone says "emulator", a large proportion of listeners think they hear an unspoken "CPU" in front of it, but even those who do usually know that nobody's talking about CPU emulation in WINE. Nobody except the Darwine people, anyway, but that's another issue.
Not to mention the fact that there are people on there everyone wishes were influential, but aren't. Ars Technica's John Siracusa is one such example. Everytime he criticises Mac OS X for something, Apple turns around and makes whatever it is worse...
Why an army
of attorneys haven't taken up pro-bono arms against the RIAA to
knock some sense into them is beyond me. Oh, no, wait.. it's not..
there's gold in them thar hills no matter which side you're on.
Maybe because in the vast majority of cases, the RIAA's lawsuits are, actually, legitimate (that is, copyright infringement has occurred, the person sued did do it, there is enough evidence against the defendent for a civil suit to be successful), and therefore the "army of lawyers" would spend most of their time protecting freeloaders rather than the collateral damage that this mother supposedly represents.
You might just as well ask why there isn't an army of lawyers taking up pro-bono arms against the cops and AG, after all, we all know the state routinely convicts innocent people, often being actively misleading about the evidence. It's for the same reason: yes, that happens, but it's the minority case.
Quite. My home has TVs in all four bedrooms and the living room, and quite honestly, the hassle of unplugging the system and all of its cables to move it around is more than a little bit of a PITA (not to mention the issues when different members of the family want to play different games at the same time.)
That's why when the PS3 comes out, my family intends to buy six (one for outdoors by the pool too). Now, I know that my son and daughter have very different opinions about what makes a great game, as do my wife and I, but there are games that appeal to all of us, and for those games to only work on one of the consoles is going to royally suck. Are we supposed to play it exclusively on the living room console or the pool console? What about late at night? What if some of us want to play the game, and the rest of us want to relax in the living room in peace and quiet?
In the end, the only solution is going to be to buy multiple copies, and I don't know about you, but I think that's an outrage as it is. It's bad enough that we have to buy seperate copies for use on the PS2-based system we have wired up in the back of both my wife and I's Hummers, for when we're taking the kids out on longer trips. We were thinking of including the Hummers's PS2s as another pair of consoles we were going to upgrade to PS3s when the PS3 comes out, but on hearing this, it sounds like it would be a complete waste of money.
With gas prices being what they are, the notion we should pay for six (or eight, if we're to include the Hummer PS3s) copies of each game on top of everything else is quite simply asking too much. Where are we supposed to find the money?
This is supposing the studios decide to use a BitTorrent-like-protocol rather than BitTorrent. It's also rather easy to abuse this particular scheme if you're a regular BitTorrent user, just use a new version that happens to use the same ports and packet identifiers that result in high-bandwidth at Hollywood's expense.
And given how simple router technology has to be, it's going to be hard for the Network Partisans to create a version that uses anything not easily duplicatable by third parties. I don't see how this can work.
Then why can you see them in the shell, instead of exclusively having commands to deal with them, and maybe a flag in "list" view (as in, when I run the list command) to show that they have metadata? I call shenanigans.
You can't see them in list view in Workbench. Yes, they're output by the list command on the command line, but ordinary use isn't supposed to be done at that level. The CLI isn't the environment in which this type of metadata is used. It's the Workbench in which the metadata is utilized. And in the Workbench, you cannot deal with.info files and their associated files seperately, the.info file is hidden from the user and is treated purely as a source of metadata.
They were easy for users to delete but not manage. There wasn't even a commandline tool for manipulating their contents! It had to be done through the GUI, or for all I know, with some ARexx crap.
That's not what I was refering to. Manage in this context meant "manage as a file", not "manage the contents of".
All of which leads me to state that info files suck
I'm not disagreeing with you. But they're not "the same as".PIFs or shortcuts. The fundamental concept that files should be associated with a tool to open them in, rather than this being assumed by type, isn't a bad one. One thing I will say though is that.info files are better than nothing, and it's a crying shame that in 2006, OVER TWENTY FUCKING YEARS SINCE WORKBENCH FIRST CAME OUT, GNOME, Windows, et al are still with "nothing", using derived data types and file extensions to launch "default" tools rather than allowing the association of specific tools with files, and Macintosh is, increadibly, joining them. I'd like the metadata to be in the file system, but failing that, icon files is an adequate, if far from ideal, alternative.
Nope. PIFs and shortcuts co-exist (in terms of what the user sees) with the files they refer to. By comparison,.info files are hidden from the user, with the user merely seeing the information they represent, not icons for the.info files themselves. Nor are shortcuts set up automatically by applications that save files with the correct metadata in them. The nearest equivalent to icons today is probably the.DS_Store mechanism of Mac OS X's current finder.
(To understand what I mean, beg, borrow, or steal an Amiga, and find a directory with files in it, some with, some without,.info files. The root of a Workbench disk is fine. With Workbench 1.3 or lower, you'll just see those files that have metadata. For Workbench 2.0, you can also select a menu option to view the directory with files that don't have metadata ("View all files" vs "Icons only" or something.) Viewing all files doesn't result in the.info files suddenly appearing seperately. In terms of the metaphor the Amiga's designers were after, icon files were internal, something the user should thing do not even exist. I think part of the reason for seperating their handling into icon.library was because the whole idea of storing the information in seperate files was ugly, and I suspect the long term aim was to incorporate the metadata into the filesystem itself. Alas, that never happened.)
It wasn't a terribly good way to store metadata, it wasn't terribly efficient Workbench acted oddly with files that didn't have associated.info files (by default hiding them altogether) while.info files were all over the place when you started to use the shell or poorly designed file selectors, but that was their purpose and they weren't supposed to be treated as files themselves. The one positive aspect to them is that they were easy to manage for users who did want to go to a low level.
BitTorrent isn't going to help the movie companies if it's blocked, whether they can "afford the fast lane" or not. Remember, BT is a peer-to-peer protocol. Most of the traffic will not be originating from the publisher, so there's no way a publisher is going to be able to go to an ISP and say "I want these torrents accellerated and I'm prepared to pay", because there's no way to identify that traffic.
XP will cease being sold in the next six months, to be replaced by Vista. At that point, I suspect the "familiarity" argument will go out the Window. Vista is quite radically different, even if it shares similarities in some areas.
That said, the familiarity argument's never really sit right with me. I use a variety of systems in my home, and have plenty of "computer illiterate" guests, none of whom ever have any serious problems with the systems available. Mac OS X is far removed from Windows, much further than a default Ubuntu GNOME GNU/Linux desktop, and nobody seems to have a problem getting the web browser up and figuring out what to do, or at least they require no more pointing than they do switching between Windows 9x and Windows XP.
Speaking as someone who stopped running it on the desktop from 2002 to 2006 (after using it since 1996), but recently tried it again, I have to say that GNU/Linux most certainly is ready for the desktop (finally.) At least, if it isn't, then Windows certainly isn't either.
Recent GNOMEs seem pretty intuitive and well designed. There's no want for software any more in most areas (and Windows is arguably deficient in as many areas), and Ubuntu appears to have done a remarkable job getting everything to "just work". I'm sure you can come up with complaints, but I doubt they affect "Grandma" more than similar complaints about Windows.
You're describing a completely unrelated problem. The problem you had was with the seperation of type data from the filename. I'm refering to opener information. The opener only refers to what application can open the file by default, a file doesn't become impossible to open in an entirely seperate application because the opener information is wrong.
However, if the type information is wrong, and the application insists on opening files only of a particular type, then, yes, you're buggered, but that's the exact opposite of what I was proposing. The fact the Mac stored type information as well (and did it badly) doesn't change the fact that the default opener of an application should not be tied to a specific type.
Not unless you were using a custom shell like SID or something.
Amiga files had icons (no, seriously, that was the technical name for the metadata files, these were stored as files with ".info" as the extension in the same directory as the source file) which contained the full pathname of the application (tool) that was supposed to open them by default. Renaming them most certainly didn't cause them to be opened in any other file.
Interestingly, you give another example of how silly it would have been to do so. IFF is not an image format, it's a container format. DPaint would open IFF ILBMs, but not other types of IFF.
Did you really have problems loading valid IFF ILBM files with random (ie not ".iff") extensions into DPaint? I don't recall having that problem myself.
I think basing it purely on file type though is a mistake.
I don't particuarly want to edit Resources/SuperappXSupport/defaults.xml using the same "editor" as Projects/MegaDB/ExportedData.xml. I don't want to edit ~/Desktop/sieve.c in the same "editor" as Projects/CorporateManagementSystem/1.0/src/somemod ule/minorplugin/extensions.c. I don't want to edit "Letter to Grandma.doc" in the same word processor as "My Novel.doc".
I'm not really sure why we've had the obsession since the mid-eighties with associating applications to file types (except to implement "defaults"), we should be recording the most appropriate application in the same metadata. This is how the Mac (used to) do it. It's how the Amiga did it. It's how it ought to be.
The mental image I have is of a Hollywood director or script writer fighting for a vision of a film and being overruled on the grounds of marketability. This kind of thing happens all the time. But then Hollywood turns around and tells everyone else, "don't you dare touch our creative vision!"
Well, in fairness, there are two points. First, pretty much every movie ends up being a compromise that both parties try to be happy with. Directors can, and often do, have their names removed from movies they're not happy with.
The second is that it wasn't the studios that went against Cleanflix, it was the Director's Guild. So it was the right body to be defending the moral rights of artists. We're not talking about the MPAA suing.
Slashdot Mirror
Slashdotters don't always have a consistant opinion on reverse engineering (which is natural given the large number of us with differing opinions.) It wouldn't be the first time a vocal group opposed reverse engineering though - when Linus threw a hissy fit because Tridge had "reverse engineered" (ie telnetted in and typed "help") BitKeeper, with BitKeeper's control-freak creators terminating kernel developer's gratis usage as a result, there were enormous numbers of Slashdotters who sided with Torvalds. Indeed, he still seems to be surprisingly popular as an "open source" figure despite employing a policy for a while that was openly hostile to Free Software.
So no, if you can't even telnet to a proprietary repository for a free software project and type "help" without a large percentage of Slashdotters protesting, you can't really assume that Slashdotters are, as a group, pro-reverse-engineering. This is one of those things with vocal proponents on both sides.
I'm finding this highly improbable. I'm not saying you don't believe that's what they said, but there has to be more to it than that.
Back in 1998, I visited the US for the first time (I'm British.) I needed more cash, so I went to an ATM in the middle of Boston, put in my card, and withdrew some money. When I came back to Britain, and got my next bank statement, the charge showed up. Which is what you'd expect.
I'm finding it just a little bit difficult to believe that this would have been possible if the ATM had to search through a database containing EVERY BANK CARD IN THE WORLD, essentially made up of arbitrary card numbers, to find out which bank account my card refered to. I can't imagine why anyone would implement something that likely to be the victim of database synchronization and duplicate number errors.
It's notable that there is an official format for financial cards which works the way most of us would expect such a thing to work, identifying features such as account numbers and institution dependent features.
I'm aware of that, and it also shares memory. But the fglrx drives are SO BAD it's probably a superior solution under GNU/Linux.
Xorg has its drivers. They're woefully out of date (X1xx series cards being completely unsupported) and relatively few contain 3D support. And fglrx is everything that'd bad about proprietary software. My laptop (a Thinkpad T60) can't even boot into X using fglrx if it's running from battery power (you just get a blank screen.) Use "totem" and "Unreal Tournament 2004" often enough, and one will stop working - if you're lucky. If you're unlucky, you'll get enough kernel panics to require a full-on reboot.
I'm honestly wondering if I'd been better off going the GMA950 route. Intel's drivers aren't completely Free Software, but there's at least a cleaner design in there.
I've said it before, but I really think the FSF needs to do more than just fund reverse engineering efforts. While ATI may hate these efforts, they make ATI products more valuable and ultimately help ATI. A concerted campaign to raise capital to start a rival graphics card manufacturer strikes me as a better solution if this is possible. I'm aware of open source graphics card efforts (involving programming FPGAs et al) but I think the next step needs to be taken. I would invest in it.
Yeah, damn it, if they hadn't put massive restrictions on owning guns in Britain seventy years ago, Britain wouldn't be seeing a slight increase in one type of crime today!
You do know that the probability of a criminal finding their victim was armed hasn't changed one iota in the last few decades?
You and Timothy need to step away from the NRA propaganda. Despite high profile "bans" on handguns in the last decade (and Kalashnikovs the decade before), realistically, few people in Britain had access to them even before these bans. There has been no decrease in deterence against the usual group of criminals supposedly fearful of guns because of any reduction in gun circulation.
Given the importance of OpenGL, and SGI's relunctance to offer a permanent royalty-free class-license for its patents in OpenGL, this could actually have some fairly dire consequences. A company could legally buy SGI's patents, turn around, and start requiring license fees from OpenGL users.
SGI may be an irrelevent hardware company to most people, but it does have potential when seen in terms of the IP it controls.
Not to mention the fact this is supposedly being announced at the WWDC.
The WWDC. World Wide Developers Conference. DEVELOPERS CONFERENCE.
I expect Jobs to be running around showing the new features of Leopard, but I highly doubt that a movie rental service is going to be one of them. The only people that will be able to make use of the service are those with Leopard? That makes no sense whatsoever. Ok, well, he'd be announcing a new feature of iTunes at the WWDC? In God's name, why? What next? Maybe he'll do a presentation on leather iPod cases there too.
And then, perhaps, at MacWorld SF next year, he can spend half the presentation talking about the great new features in Xcode.
This is a ridiculous rumour from an increasingly irrelevent rumours site. The basic concept's nice and workable, but it's clear the rumour is, actually, completely made up. It carries about as much wieght as my prediction that Apple is planning to put Core 2 Duos in the iMac. I mean, yeah, that would make lots of sense, but that doesn't mean I actually have been told this by anyone in the know. Further, if I add that Apple's intent is to call the new 64 bit iMac the "CrapMac", it should become fairly obvious I know bugger all.
Yet, apparently, Apple is going to announce a movie downloads system at a developer's conference. Not a special event, like it did the iTunes Music Store. Or the leather cases and iPod speakers. No. This will be told to a bunch of developers who are actually interested in what Apple's new operating system will be like, because, like, that'll not piss them off at all, and everyone knows that all the cool journalists hang out at the WWDC.
Okay.
Never underestimate what you can do in a high level language if its responsible for generating its own assembler code. I can actually see a few ways of implementing a Java like language (whose compiler generates assembler code) that can implement its own GC in high level code. And that's ignoring the obvious possibility of giving your garbage collection a big ass "byte[] ProcessMemory" object to work with.
For example, the base Object type can have some base methods including "long AddressOf", "long SizeOf", "void RelocateTo(long)", and "Object[] GetObjectsIReferTo". Add to that an object that abstracts the stack, and ensure your generated code conforms to certain conventions so object addresses are only ever located in certain positions on the stack or certain registers, or in objects themselves.
Now, at this point this is pretty low level. You're still working with addresses. And there's still the realistic possibility of screwing up and causing object corruption. You can, of course, simply prevent other classes from using the methods available to the garbage collector at a low level. But if you're really smart, you can do the above without the AddressOf or RelocateTo things using addresses. You ensure you know the order of the objects on the heap, and replace RelocateTo with some form of "DropDown" object. You also abstract the heap (or heaps, in the case of more recent Javas whose GC strategy involves three heaps.)
That's just the beginning, but you can probably figure out from the above that an unmanaged language (I hate the term "low level language" when applied to C. Stop redefining terms people! C is a high level language) or some very convoluted, self contained, assembly is a necessary component of the implementation of a managed language.
Options:
1. Archive source locally and offer to sell (for $10 or less) CDs for archived versions.
2. Use SourceForge or Savannah rather than your own personal website for distributing source and binaries.
3. Include the source code with the binaries in the same package.
Pretty easy really. Unless your software package is extraordinarily big, (3) is probably the easiest. (And yes, SF is on the hook, not you, if they lose the source. They're the ones redistributing the binary.)
Excellent point. Further, it could be argued that if a company can, before making a hiring decision, determine that it's better to hire two people for $X who'll live a particular standard of living than it is to hire one person for 2x$X living the same standard of living, then morally and economically, the company is making the right decision by chosing the first (the "outsourcing" option.)
Firing someone merely to give someone else a job is, obviously, a different matter.
Yeah, we've made a lot of jobs obsolete.
We've also created entirely new fields of work, and we've made it possible for new industries to be created. Our technologies make things possible, making products and services possible, that could never be dreamt of without our help.
I'm not going to say I'm ticked off at outsourcing, because it's a zero-sum game at the end of the day, one person loses a job, another gains it. But I'm not going to say I agree with you that those complaining are engaging in "doublespeak". They're not. This is a fantastic industry that has created far, far, more than its obsoleted, and will continue to do so for decades to come.
FWIW, this usually comes up, and it's worth pointing out a few things.
1. Most people know that WINE is a reimplementation of an API, not a CPU emulator. I know a handful of people get confused here, but the majority, in my experience, do not.
2. Historically, the word emulator has, actually, included programs that simulate operating systems, not just programs that simulate CPUs. For example, the Atari ST and Commodore Amiga had several systems marketed as "Macintosh emulators". The non-hardware versions of each merely patched the Macintosh ROMs to use local hardware rather than Mac hardware. So called "CP/M" emulators for V20/V30 series CPUs (an NEC version of the 8086 with native 8080 code execution support) were also widely available. These are by far not the only things refered to as "emulators" that were operating system, not CPU, emulators.
3. The debate over the term is so commonplace, that at one time WINE stood for, officially, both as "Wine Is Not an Emulator" and "WINdows Emulator". This is perfectly legitimate, it isn't a CPU emulator, but it does emulate Windows. A quick look-up of the dictionary definition of the word "emulate" will put you in good stead here as to why both terms are legimate.
The word "emulator" does not mean "CPU emulator", it merely means "Something that emulates". Whether you're emulating (immitating) a CPU, or an operating system, you are, actually, an emulator. Jumping in with a correction over the word emulator because it's not a CPU emulator is entirely unnecessary, and largely wrong. I don't doubt your motives, unfortunately usually when someone says "emulator", a large proportion of listeners think they hear an unspoken "CPU" in front of it, but even those who do usually know that nobody's talking about CPU emulation in WINE. Nobody except the Darwine people, anyway, but that's another issue.
Not to mention the fact that there are people on there everyone wishes were influential, but aren't. Ars Technica's John Siracusa is one such example. Everytime he criticises Mac OS X for something, Apple turns around and makes whatever it is worse...
Maybe because in the vast majority of cases, the RIAA's lawsuits are, actually, legitimate (that is, copyright infringement has occurred, the person sued did do it, there is enough evidence against the defendent for a civil suit to be successful), and therefore the "army of lawyers" would spend most of their time protecting freeloaders rather than the collateral damage that this mother supposedly represents.
You might just as well ask why there isn't an army of lawyers taking up pro-bono arms against the cops and AG, after all, we all know the state routinely convicts innocent people, often being actively misleading about the evidence. It's for the same reason: yes, that happens, but it's the minority case.
Quite. My home has TVs in all four bedrooms and the living room, and quite honestly, the hassle of unplugging the system and all of its cables to move it around is more than a little bit of a PITA (not to mention the issues when different members of the family want to play different games at the same time.)
That's why when the PS3 comes out, my family intends to buy six (one for outdoors by the pool too). Now, I know that my son and daughter have very different opinions about what makes a great game, as do my wife and I, but there are games that appeal to all of us, and for those games to only work on one of the consoles is going to royally suck. Are we supposed to play it exclusively on the living room console or the pool console? What about late at night? What if some of us want to play the game, and the rest of us want to relax in the living room in peace and quiet?
In the end, the only solution is going to be to buy multiple copies, and I don't know about you, but I think that's an outrage as it is. It's bad enough that we have to buy seperate copies for use on the PS2-based system we have wired up in the back of both my wife and I's Hummers, for when we're taking the kids out on longer trips. We were thinking of including the Hummers's PS2s as another pair of consoles we were going to upgrade to PS3s when the PS3 comes out, but on hearing this, it sounds like it would be a complete waste of money.
With gas prices being what they are, the notion we should pay for six (or eight, if we're to include the Hummer PS3s) copies of each game on top of everything else is quite simply asking too much. Where are we supposed to find the money?
This is supposing the studios decide to use a BitTorrent-like-protocol rather than BitTorrent. It's also rather easy to abuse this particular scheme if you're a regular BitTorrent user, just use a new version that happens to use the same ports and packet identifiers that result in high-bandwidth at Hollywood's expense.
And given how simple router technology has to be, it's going to be hard for the Network Partisans to create a version that uses anything not easily duplicatable by third parties. I don't see how this can work.
Nope. PIFs and shortcuts co-exist (in terms of what the user sees) with the files they refer to. By comparison, .info files are hidden from the user, with the user merely seeing the information they represent, not icons for the .info files themselves. Nor are shortcuts set up automatically by applications that save files with the correct metadata in them. The nearest equivalent to icons today is probably the .DS_Store mechanism of Mac OS X's current finder.
(To understand what I mean, beg, borrow, or steal an Amiga, and find a directory with files in it, some with, some without, .info files. The root of a Workbench disk is fine. With Workbench 1.3 or lower, you'll just see those files that have metadata. For Workbench 2.0, you can also select a menu option to view the directory with files that don't have metadata ("View all files" vs "Icons only" or something.) Viewing all files doesn't result in the .info files suddenly appearing seperately. In terms of the metaphor the Amiga's designers were after, icon files were internal, something the user should thing do not even exist. I think part of the reason for seperating their handling into icon.library was because the whole idea of storing the information in seperate files was ugly, and I suspect the long term aim was to incorporate the metadata into the filesystem itself. Alas, that never happened.)
It wasn't a terribly good way to store metadata, it wasn't terribly efficient Workbench acted oddly with files that didn't have associated .info files (by default hiding them altogether) while .info files were all over the place when you started to use the shell or poorly designed file selectors, but that was their purpose and they weren't supposed to be treated as files themselves. The one positive aspect to them is that they were easy to manage for users who did want to go to a low level.
BitTorrent isn't going to help the movie companies if it's blocked, whether they can "afford the fast lane" or not. Remember, BT is a peer-to-peer protocol. Most of the traffic will not be originating from the publisher, so there's no way a publisher is going to be able to go to an ISP and say "I want these torrents accellerated and I'm prepared to pay", because there's no way to identify that traffic.
XP will cease being sold in the next six months, to be replaced by Vista. At that point, I suspect the "familiarity" argument will go out the Window. Vista is quite radically different, even if it shares similarities in some areas.
That said, the familiarity argument's never really sit right with me. I use a variety of systems in my home, and have plenty of "computer illiterate" guests, none of whom ever have any serious problems with the systems available. Mac OS X is far removed from Windows, much further than a default Ubuntu GNOME GNU/Linux desktop, and nobody seems to have a problem getting the web browser up and figuring out what to do, or at least they require no more pointing than they do switching between Windows 9x and Windows XP.
Speaking as someone who stopped running it on the desktop from 2002 to 2006 (after using it since 1996), but recently tried it again, I have to say that GNU/Linux most certainly is ready for the desktop (finally.) At least, if it isn't, then Windows certainly isn't either.
Recent GNOMEs seem pretty intuitive and well designed. There's no want for software any more in most areas (and Windows is arguably deficient in as many areas), and Ubuntu appears to have done a remarkable job getting everything to "just work". I'm sure you can come up with complaints, but I doubt they affect "Grandma" more than similar complaints about Windows.
You're describing a completely unrelated problem. The problem you had was with the seperation of type data from the filename. I'm refering to opener information. The opener only refers to what application can open the file by default, a file doesn't become impossible to open in an entirely seperate application because the opener information is wrong.
However, if the type information is wrong, and the application insists on opening files only of a particular type, then, yes, you're buggered, but that's the exact opposite of what I was proposing. The fact the Mac stored type information as well (and did it badly) doesn't change the fact that the default opener of an application should not be tied to a specific type.
Not unless you were using a custom shell like SID or something.
Amiga files had icons (no, seriously, that was the technical name for the metadata files, these were stored as files with ".info" as the extension in the same directory as the source file) which contained the full pathname of the application (tool) that was supposed to open them by default. Renaming them most certainly didn't cause them to be opened in any other file.
Interestingly, you give another example of how silly it would have been to do so. IFF is not an image format, it's a container format. DPaint would open IFF ILBMs, but not other types of IFF.
Did you really have problems loading valid IFF ILBM files with random (ie not ".iff") extensions into DPaint? I don't recall having that problem myself.
I think basing it purely on file type though is a mistake.
I don't particuarly want to edit Resources/SuperappXSupport/defaults.xml using the same "editor" as Projects/MegaDB/ExportedData.xml. I don't want to edit ~/Desktop/sieve.c in the same "editor" as Projects/CorporateManagementSystem/1.0/src/somemod ule/minorplugin/extensions.c. I don't want to edit "Letter to Grandma.doc" in the same word processor as "My Novel.doc".
I'm not really sure why we've had the obsession since the mid-eighties with associating applications to file types (except to implement "defaults"), we should be recording the most appropriate application in the same metadata. This is how the Mac (used to) do it. It's how the Amiga did it. It's how it ought to be.
Well, in fairness, there are two points. First, pretty much every movie ends up being a compromise that both parties try to be happy with. Directors can, and often do, have their names removed from movies they're not happy with.
The second is that it wasn't the studios that went against Cleanflix, it was the Director's Guild. So it was the right body to be defending the moral rights of artists. We're not talking about the MPAA suing.