Yup -- I tossed an article onto Slashdot about a year ago pointing this out.
I'm not familiar enough with Word internals to know how useful the schema would be in translating documents losslessly between formats (and am very dubious that it would be particularly easy), but it isn't even necessary to go that far -- the point is that the non-Pro copies of Word don't support XML format export.
Microsoft isn't going to give up the golden strength of a file format lock-in any time soon, even if they let companies use custom indexing tools on their store of documents (which is really what this whole XML business is about).
I never said symlinks and shortcuts were the same. I said MacOS *Aliases* and Shortcuts were equivalent.
Hmm. I think you may have mistyped. Here's your earlier post:
If you're going to equate aliases to symlinks, then you have to equate shortcuts to symlinks as well.
In either case, I think you're going to have a tough time arguing the point -- I can open() through a path containing a symlink or alias, but not a shortcut. Shortcuts operate as merely an Explorer convenience.
I'd suggest this is an application issue. I've no doubt applications could be written to follow shortcuts if necessary.
Sure, but they aren't and never will be -- *all* the file code would have to be changed in every application. You'd always have code that failed or didn't work properly, even in the unthinkable case of the industry running out and trying to fix this as a whole. The entire point of symlinks and aliases is that they don't have to have application level support.
Like what ? Symlinks shouldn't require much (if any) modifications to the kernel - they're a filesystem feature.
Ummm...huh? This doesn't make any sense -- this is like saying "adding UDP support shouldn't require much (if any) modification to the kernel -- they're a network feature".
But I sorta doubt they'll use it to track down spammers
It's probably just the case that this is an interesting item in information theory research. Maybe they didn't have an overarching political agenda -- AT&T does lots of research in this area.
At at least some research labs, you get promotions/raises/etc based on how many patents you churn out.
If you look at, say, an encryption forum, you'll probably see people posting attacks on encryption algorithms. This isn't because they hate the algorithm, but because they know that the Right Way to work is to hammer on things, and not simply trust broken things.
My guess is that the same thing, hammering on something that's known to be broken (existing antispam systems) happened here, and then the guy just did the regular ol' apply-for-a-patent-related-to-my-research thing when done so that he could get credit for his work. Is AT&T out to further or destroy spam? Probably not, at least not using this patent. It's just business as usual.
This actually brings up an interesting point -- AFAI can tell, a good deal of the more "revolutionary" jumps in fields occur when someone who is familiar with two different fields brings in ideas from the first field to the second, which is populated by people who have never heard of the idea before.
Better analysis of existing data may be more rewarding than producing new data.
Of course, you don't get fat research grants for reading old papers.:-(
I've run into CS professors that couldn't program.
Frankly, most CS professors that I've run into are not particularly impressive programmers. Except for the very rare professor that teaches nothing but programming, most just don't have time to spend on programming, nor is programming particularly interesting compared to the theory.
I agree that not programming at all is a bit extreme -- one would probably need to get a mathematics PhD a few years back who was interested in some area of math theory that applied nicely to CS to be a "CS professor" that literally has never programmed, but CS is not about programming. You can pick up a good understanding of programming with texts and the Internet in short order. CS is a much, much larger field.
The professors, surely, for not being adept. But the institution that hired them, as well, for hiring retards. The schools that gave these people their degrees and doctorates in the first place, as well. How about their secondary schooling? That's at fault as well, for not teaching them (at least) how to think critically (deductive logic) and learn on their own. I'd partially blame the plethora of students that tend to go to 4-year schools for education nowadays: it's turned your average university into simply a festering wound for these magots to crawl around in, get drunk, etc. - as opposed to an institution of higher learning that has prestigous requirements, schools their students well, and turns out a very high percentage of leaders. The excess of required programs that students are required to take are utter garbage, things that should have been learned in high school (first and second year english spring to mind).
Frankly, it's not possible to truly do well, to excel, if you do nothing but sit in your seat and slowly absorb what's being said. The curriculum is provided to ensure that you know a bare minimum of what's involved. A *bare minimum*. It's assumed that if you want to really understand things, you will go out and read about them yourself. Every time I've run out and read up on a subject, I've found that professors are delighted to talk about their area in office hours.
Frankly, if you want a good argument for requiring one or two English classes for a CS degree, I think you need look no further than Slashdot. (Though, honestly, most problems here seem to be grammatical, not issues with being able to read and analyze literature, which is most likely what an introductory college-level English course will be on).
When I was attending university, I was delighted the first semester I took three CS courses simultanously. Until, that is, they started to weigh down heavily on me. I lost my interest in independent CS work.
On the other hand, I think that taking the one history course and one public policyish course that I tried in university was a fantastic idea. I really enjoyed them -- perhaps because they were different, but because I was in a new field, and learning new things rapidly, for the first time in a while.
No, the problem is that the garage door manufacturers, much like certain large OS vendors, did not consider eventual security concerns in the least when building their product.
If they *did*, they'd be doing RSA, or at least have a shared secret.
The code to my garage is encoded using a 10-bit switch. Since anybody could open my garage in a reasonable period of time by simply trying all possible codes, I don't don't keep valuable items in my garage unless they are secured through some other means (i.e: vehicle key).
Frankly, I suspect that any kind of sane thief would just break a window...
The idea is that a couple of ambitious companies tried getting themselves an artificial monopoly by claiming that people producing compatible devices were bypassing copy protection schemes and were therefore criminally liable.
No. NTFS supports hardlinks. I haven't heard of it supporting symlinks.
Shortcuts are not symlinks. They cannot be used by file-level access calls -- open() cannot contain a shortcut in a path. The only thing in Windows that groks shortcuts is Explorer (and Explorer-provided widgets, like the open and save dialogs). I can't just replace my id1/foobar directory with a symlink.
If you run around replacing directories with shortcuts, almost all Windows programs will just break. On the other hand, doing so is feasible with both aliases and symlinks on classic Mac OS and UNIX.
The only time Windows programs can use symlinks that I'm aware of is when they're running in WINE.
Microsoft is extremely twitchy about touching their kernel. All sorts of functionality that should have been put in the kernel has had halfassed or no equivalents in Explorer. Really awful file locking workarounds and a lack of symlinks are my two should-be-done-properly pet peeves.
I'm sorry, but you are wrong. While I can't comment about whether Bin Laden is Evil according to your definition (probably not - most religious fanats won't be), Saddam Hussein clearly is.
We are talking about a man who needs no further justification for his actions than "That's what I want to do" and "That's what will promote me".
You do realize that if you ran a PR campaign and isolated the least moral and put up your more negative assumptions about motivations on just about anyone (yes, including Bush), you can make them look quite "evil".
"Evil" is a pretty poorly defined concept, but in general, I've found that when you actually know and talk to someone, even someone widely disliked, their viewpoint isn't all that unreasonable. They're just another human with a different background and somewhat different set of goals. Be it Bill Gates, Fidel Castro, Richard Nixon, or George Bush, I suspect that friends and family can appreciate their own position.
Furthermore, Saddam is *not* particularly religious -- as a matter of fact, Saddam is also supposed to be on the al Queda shit list because of the fact that he has seriously secularized an Islamic country.
The point being, this wasn't a random computer store - he brought it there because he knew I worked there and knew we did good work. I'm sure he just forgot, or didn't know how to turn it off.
[shrug] Okay, fair enough. That wasn't clear to me in the initial post.
I'd care if I was taking my machine in for repair because it's incredibly rude to impose that sort of thing on others. There's a difference between leaving your porn in a folder somewhere when it goes into the shop, and having it be your desktop and screensaver.
[sigh] This is where we get into gender differences. Imposition wasn't at all the issue I thought you were bringing up -- rather, the possibility that he might be shamed.
I suppose the closest equivalent I could think of would be if I ran into a woman's computer with pornography of men, or perhaps a guy with gay porn on his computer. I just...I can't say I'd feel imposed upon. I'd probably get a chuckle out of it, but not feel uncomfortable.
To be honest, the tint of dishonesty is what bothers me more. Obviously, we try to lie to each other, to present a different image from what we really are. People clean up the house when they have relatives coming over, job interviewees wear formal business attire for interviews (even if they never would on the job), and generally sexuality is Not Spoken Of. It kind of bothers me -- I wish everyone could be entirely open and honest -- if they happen to find tentacle porn appealing, it'd be nice if folks could admit it. I've had a couple friends that are gay, and the degree of net emotional agony they go through with having to deal with hiding the fact much of the time seems like a mind-bogglingly awful social issue.
It may be that two hundred years from now, folks will be as far beyond us today regarding honesty in sexuality as we are beyond Victorian times and the kind of hidden social issues that Freud turned up.
A windows fan claims Longhorn is/will be better than Panther. Will we run a follow-up story when Apple fans claim in this slashdot story that Panther is in fact better than Longhorn?
To Slashdot's credit and my surprise, people actually aren't biting the troll. There are a lot of related comments, but very few comments along the lines of "MS sucks and the stuff from other people is better."
By the way, Longhorn does exist, inasmuch as leaked builds can be found on the Internet.
I agree that there will probably be a Windows OS (or suite of Windows OSes released within two years of Longhorn's projected ship date).
I'm somewhat dubious that the content of those OSes is going to be necessarily all that similar to what's being distributed today as Longhorn.
Remember what Windows XP looked like (okay, I admit that this is a UI issue) in the Whistler betas. Nothing at all like what shipping Windows XP looks like.
The whole task-based business is coming back to life *again*?
Sigh.
It was supposed to be big with Apple's OpenDoc. Neat research idea that didn't map very well to the actual metaphors in use in their UI. Microsoft tried doing it with OLE. The idea is that you have one big monolithic application tied into the OS that can do everything you'd possibly want using components.
And now, some guy is still harping on the "task-based approach". Urgh.
However, it seems like one of the big things in Longhorn will be the WinFS--which I understand to the the database-as-a-filesystem.
(a) Not a new idea. MS will be the first to try to put it into a popular OS, though.
(b) Apple has done similar before, though you'd be unlikely to guess it from the interface -- they had the desktop file, a constantly OS-maintained index of the filesystem.
(c) There are a number of technical and user interface issues with this approach -- there's a reason folks standardized on a hierarchical system. We'll see what happens, though.
Incidently, Windows is the only popular OS that still only supports a tree-style filesystem. Classic Mac OS, modern Mac OS, BSD and Linux all support any fully connected graph you might desire, thanks to symlinks (and on classic Mac OS, aliases).
d) This isn't actually all that new even from a UI perspective -- think of using Apple's Find File. Perhaps you toss in a few more search parameters to get at more metadata and data. The real difference is that traditionally, you must *also* assign a file a position in a hierarchical filesystem (though your hierarchical filesystem could potentially consist of just a single directory node with lots of files in it, a la the My Documents MS approach).
e) I'm remembering the last time the database research people at Microsoft convinced everyone that going all database would be a great idea. MS SQL, for a period of time, used tables internally for *everything*. Performance sucked, but they did it anyway. It's sexy from a theory point of view, because it simplifies things. Unfortunately, it throws out a lot of area-specific design knowledge that's been built up over the years. We know a lot about how to do a good filesystem, and there are features that apply nicely to filesystems that are less convenient with a traditional database. It's going to be tough to make a better system by throwing everything out.
f) I've heard ominous rumblings about WinFS being removed from Longhorn. It may or may not be living up to internal expectations.
g) Anytime something like this is announced *this* far in advance and isn't getting shown off in final form, it means that the promises frequently come from the research people. Research people have all sorts of rosy views about their own work, plenty of pet ideas, and may not have spent a long time doing usability tests. (This comes from one of those research people.) I wouldn't get excited about this any more than I would the frequent announcements on Slashdot about nonexistent new storage technologies ("in four years, we're all going to be using five terabyte 1 cubic centimeter Jell-O blobs to store our data!"). Yeah...come back when you have something shipping instead of a bunch of theoretical maximum numbers coming from a research team.
When you don't like a company, their methods, or their products, a free-market way of dealing with this would be to simply buy the products of their competitors. Is something like this not possible with employment as well?
Unfortunately, I don't believe so.
The simple free market economic model that is taught in Microeconomics I makes a number of assumptions that don't hold here. The first is that of perfect information. We have enough trouble getting accurate and trustable information to the public in the first place when it comes to products. Because of NDAs, the traditional secrecy surrounding salaries, and the fact that workers are generally only very familiar with their own company, people tend to not have a information basis for making comparative decisions about a company. Plus, as I pointed out, senior management exists in a position that is very easy for them to abuse, and because it is a relatively small number of people relative to the rest of the people in a company, usually isn't enough to sink the company.
People indirectly punish this behavior -- they will buy cheaper from another company that does not do the same thing -- but because upper managmeent is a small group, unit price generally isn't impacted all that much.
Finally, it's very difficult to differentiate workers on skill period -- workers want to positively represent themselves, worries about lawsuits prevent many past employers from describing an employee's past performance, and a bos only sees a portion of what an employee does. Finally, while it might be easy to rate a bricklayer (only a few metrics to gauge job performance on and all metrics are pretty visible -- you lay N bricks per hour and your laid bricks deviate from an ideally laid brick in M ways), CEO behavior tends to be very situation-specific, and difficult to judge.
The way I see it, the first step is to let the game companies sell the ROMS in whatever protected and limited fashon they are comfortable with, and then after they see a consumer demand for it, market forces would start letting competitive companies offer better EULAs
The main problem with this idea is that owners of copyrighted material have a natural monopoly over that material. Nobody else can *offer* that material but them, so there is no competition. You can get Phantasy Star from but one company.
There *are* multiple RPGs that you might play instead, but if you're, say, a Mega Man fan, Mario is not a simple substitute, you know?
I agree with your points, though I think that your claim that it's necessarily because someone else "worked harder" isn't necessarily true. Wanting to be the top dog is completely disconnected from how hard other people worked.
The entire point of mathematics is to produce perfect but abstract models. In some cases, these models are produced with no consideration for actual application. This is the job of the mathematician. He sets down a set of axioms, and granted those axioms, makes known-true claims about models. His claims are not "true" because they have anything to do with reality, but because they follow a commonly accepted set of axioms. The logician is a mathematician. Technically a statician is as well, but since statistics are usually used in an applied manner (a la science), it tends not to be found in math departments.
The philosopher (well, there are a lot of types of philosopher, but I like just going for the analytic ones and dumping all those ethicists:-) ) works with the next closest field. He deals with knowledge about *reality* given some set of (usually) precisely defined axioms. His work is inherently relevant to reality, but not necessary particularly useful.
The scientist (distinct from *all* these) produces knowledge about what models currently are *believed to best apply* to reality given a (not precisely defined, but generally accepted among most folks) set of axioms. The scientist is not interested in absolute truth -- he deals with our current best guesses about reality.
A "pure" scientist is still part of this process...he just proposes models for use in representing reality that may not be easily testable, and may be "better" than other models only more subtle rules, like Ockham's Razor.
If you are a smaller ISV, (Independent Service Provider, in case you didn't know)
The acronym and the words don't match up.
Independent Software (occasionally Service or Solution) Vendor might be better.
An ISV is just a company that sells software (or services or solutions:-) ) but doesn't sell the hardware platform that the software runs on. Adobe's a good example of an ISV.
And the poor schmucks that put up movies need to learn to use BitTorrent.
Slashdot Mirror
Yeah. It doesn't lock you in. The competitors just can't be GPLed software.
how many TCP/IP stacks do you know of that AREN'T derived from BSD?
Linux's, for one major example.
The only stack fully compliant with the TCP/IP spec.
Yup -- I tossed an article onto Slashdot about a year ago pointing this out.
I'm not familiar enough with Word internals to know how useful the schema would be in translating documents losslessly between formats (and am very dubious that it would be particularly easy), but it isn't even necessary to go that far -- the point is that the non-Pro copies of Word don't support XML format export.
Microsoft isn't going to give up the golden strength of a file format lock-in any time soon, even if they let companies use custom indexing tools on their store of documents (which is really what this whole XML business is about).
I never said symlinks and shortcuts were the same. I said MacOS *Aliases* and Shortcuts were equivalent.
Hmm. I think you may have mistyped. Here's your earlier post:
If you're going to equate aliases to symlinks, then you have to equate shortcuts to symlinks as well.
In either case, I think you're going to have a tough time arguing the point -- I can open() through a path containing a symlink or alias, but not a shortcut. Shortcuts operate as merely an Explorer convenience.
I'd suggest this is an application issue. I've no doubt applications could be written to follow shortcuts if necessary.
Sure, but they aren't and never will be -- *all* the file code would have to be changed in every application. You'd always have code that failed or didn't work properly, even in the unthinkable case of the industry running out and trying to fix this as a whole. The entire point of symlinks and aliases is that they don't have to have application level support.
Like what ? Symlinks shouldn't require much (if any) modifications to the kernel - they're a filesystem feature.
Ummm...huh? This doesn't make any sense -- this is like saying "adding UDP support shouldn't require much (if any) modification to the kernel -- they're a network feature".
Well...you aren't *supposed* to be able to patent mathematical systems.
Of course, RSA managed to get their encryption patent...
But I sorta doubt they'll use it to track down spammers
It's probably just the case that this is an interesting item in information theory research. Maybe they didn't have an overarching political agenda -- AT&T does lots of research in this area.
At at least some research labs, you get promotions/raises/etc based on how many patents you churn out.
If you look at, say, an encryption forum, you'll probably see people posting attacks on encryption algorithms. This isn't because they hate the algorithm, but because they know that the Right Way to work is to hammer on things, and not simply trust broken things.
My guess is that the same thing, hammering on something that's known to be broken (existing antispam systems) happened here, and then the guy just did the regular ol' apply-for-a-patent-related-to-my-research thing when done so that he could get credit for his work. Is AT&T out to further or destroy spam? Probably not, at least not using this patent. It's just business as usual.
This actually brings up an interesting point -- AFAI can tell, a good deal of the more "revolutionary" jumps in fields occur when someone who is familiar with two different fields brings in ideas from the first field to the second, which is populated by people who have never heard of the idea before.
:-(
Better analysis of existing data may be more rewarding than producing new data.
Of course, you don't get fat research grants for reading old papers.
I've run into CS professors that couldn't program.
Frankly, most CS professors that I've run into are not particularly impressive programmers. Except for the very rare professor that teaches nothing but programming, most just don't have time to spend on programming, nor is programming particularly interesting compared to the theory.
I agree that not programming at all is a bit extreme -- one would probably need to get a mathematics PhD a few years back who was interested in some area of math theory that applied nicely to CS to be a "CS professor" that literally has never programmed, but CS is not about programming. You can pick up a good understanding of programming with texts and the Internet in short order. CS is a much, much larger field.
The professors, surely, for not being adept. But the institution that hired them, as well, for hiring retards. The schools that gave these people their degrees and doctorates in the first place, as well. How about their secondary schooling? That's at fault as well, for not teaching them (at least) how to think critically (deductive logic) and learn on their own. I'd partially blame the plethora of students that tend to go to 4-year schools for education nowadays: it's turned your average university into simply a festering wound for these magots to crawl around in, get drunk, etc. - as opposed to an institution of higher learning that has prestigous requirements, schools their students well, and turns out a very high percentage of leaders. The excess of required programs that students are required to take are utter garbage, things that should have been learned in high school (first and second year english spring to mind).
Frankly, it's not possible to truly do well, to excel, if you do nothing but sit in your seat and slowly absorb what's being said. The curriculum is provided to ensure that you know a bare minimum of what's involved. A *bare minimum*. It's assumed that if you want to really understand things, you will go out and read about them yourself. Every time I've run out and read up on a subject, I've found that professors are delighted to talk about their area in office hours.
Frankly, if you want a good argument for requiring one or two English classes for a CS degree, I think you need look no further than Slashdot. (Though, honestly, most problems here seem to be grammatical, not issues with being able to read and analyze literature, which is most likely what an introductory college-level English course will be on).
When I was attending university, I was delighted the first semester I took three CS courses simultanously. Until, that is, they started to weigh down heavily on me. I lost my interest in independent CS work.
On the other hand, I think that taking the one history course and one public policyish course that I tried in university was a fantastic idea. I really enjoyed them -- perhaps because they were different, but because I was in a new field, and learning new things rapidly, for the first time in a while.
No, the problem is that the garage door manufacturers, much like certain large OS vendors, did not consider eventual security concerns in the least when building their product.
If they *did*, they'd be doing RSA, or at least have a shared secret.
The code to my garage is encoded using a 10-bit switch. Since anybody could open my garage in a reasonable period of time by simply trying all possible codes, I don't don't keep valuable items in my garage unless they are secured through some other means (i.e: vehicle key).
Frankly, I suspect that any kind of sane thief would just break a window...
Nobody said that REing a product was illegal.
The idea is that a couple of ambitious companies tried getting themselves an artificial monopoly by claiming that people producing compatible devices were bypassing copy protection schemes and were therefore criminally liable.
The REing itself was never claimed to be illegal.
"Article 9. 2. Copyright protection shall extend to expressions and not to ideas, procedures, methods of operation or mathematical concepts as such."
Uh, huh.
And then RSA got its patent. Hmm.
Windows paths must describe a path through a filesystem that is a tree -- i.e. you can reach all points, but only by traveling through a single path.
:-) )
On a *IX or Mac OS system, this is not the case.
I'm using the terms "graph" and "tree" from a CS point of view (though I should have said "connected", not "fully connected".
No. NTFS supports hardlinks. I haven't heard of it supporting symlinks.
Shortcuts are not symlinks. They cannot be used by file-level access calls -- open() cannot contain a shortcut in a path. The only thing in Windows that groks shortcuts is Explorer (and Explorer-provided widgets, like the open and save dialogs). I can't just replace my id1/foobar directory with a symlink.
If you run around replacing directories with shortcuts, almost all Windows programs will just break. On the other hand, doing so is feasible with both aliases and symlinks on classic Mac OS and UNIX.
The only time Windows programs can use symlinks that I'm aware of is when they're running in WINE.
Microsoft is extremely twitchy about touching their kernel. All sorts of functionality that should have been put in the kernel has had halfassed or no equivalents in Explorer. Really awful file locking workarounds and a lack of symlinks are my two should-be-done-properly pet peeves.
Depends on your frame of reference. Could be just two axes. :-)
The paint might muck that up, though.
It'd be a hell of a calculation.
I'm sorry, but you are wrong. While I can't comment about whether Bin Laden is Evil according to your definition (probably not - most religious fanats won't be), Saddam Hussein clearly is.
We are talking about a man who needs no further justification for his actions than "That's what I want to do" and "That's what will promote me".
You do realize that if you ran a PR campaign and isolated the least moral and put up your more negative assumptions about motivations on just about anyone (yes, including Bush), you can make them look quite "evil".
"Evil" is a pretty poorly defined concept, but in general, I've found that when you actually know and talk to someone, even someone widely disliked, their viewpoint isn't all that unreasonable. They're just another human with a different background and somewhat different set of goals. Be it Bill Gates, Fidel Castro, Richard Nixon, or George Bush, I suspect that friends and family can appreciate their own position.
Furthermore, Saddam is *not* particularly religious -- as a matter of fact, Saddam is also supposed to be on the al Queda shit list because of the fact that he has seriously secularized an Islamic country.
The point being, this wasn't a random computer store - he brought it there because he knew I worked there and knew we did good work. I'm sure he just forgot, or didn't know how to turn it off.
[shrug] Okay, fair enough. That wasn't clear to me in the initial post.
I'd care if I was taking my machine in for repair because it's incredibly rude to impose that sort of thing on others. There's a difference between leaving your porn in a folder somewhere when it goes into the shop, and having it be your desktop and screensaver.
[sigh] This is where we get into gender differences. Imposition wasn't at all the issue I thought you were bringing up -- rather, the possibility that he might be shamed.
I suppose the closest equivalent I could think of would be if I ran into a woman's computer with pornography of men, or perhaps a guy with gay porn on his computer. I just...I can't say I'd feel imposed upon. I'd probably get a chuckle out of it, but not feel uncomfortable.
To be honest, the tint of dishonesty is what bothers me more. Obviously, we try to lie to each other, to present a different image from what we really are. People clean up the house when they have relatives coming over, job interviewees wear formal business attire for interviews (even if they never would on the job), and generally sexuality is Not Spoken Of. It kind of bothers me -- I wish everyone could be entirely open and honest -- if they happen to find tentacle porn appealing, it'd be nice if folks could admit it. I've had a couple friends that are gay, and the degree of net emotional agony they go through with having to deal with hiding the fact much of the time seems like a mind-bogglingly awful social issue.
It may be that two hundred years from now, folks will be as far beyond us today regarding honesty in sexuality as we are beyond Victorian times and the kind of hidden social issues that Freud turned up.
A windows fan claims Longhorn is/will be better than Panther. Will we run a follow-up story when Apple fans claim in this slashdot story that Panther is in fact better than Longhorn?
To Slashdot's credit and my surprise, people actually aren't biting the troll. There are a lot of related comments, but very few comments along the lines of "MS sucks and the stuff from other people is better."
By the way, Longhorn does exist, inasmuch as leaked builds can be found on the Internet.
I agree that there will probably be a Windows OS (or suite of Windows OSes released within two years of Longhorn's projected ship date).
I'm somewhat dubious that the content of those OSes is going to be necessarily all that similar to what's being distributed today as Longhorn.
Remember what Windows XP looked like (okay, I admit that this is a UI issue) in the Whistler betas. Nothing at all like what shipping Windows XP looks like.
The whole task-based business is coming back to life *again*?
Sigh.
It was supposed to be big with Apple's OpenDoc. Neat research idea that didn't map very well to the actual metaphors in use in their UI. Microsoft tried doing it with OLE. The idea is that you have one big monolithic application tied into the OS that can do everything you'd possibly want using components.
And now, some guy is still harping on the "task-based approach". Urgh.
However, it seems like one of the big things in Longhorn will be the WinFS--which I understand to the the database-as-a-filesystem.
(a) Not a new idea. MS will be the first to try to put it into a popular OS, though.
(b) Apple has done similar before, though you'd be unlikely to guess it from the interface -- they had the desktop file, a constantly OS-maintained index of the filesystem.
(c) There are a number of technical and user interface issues with this approach -- there's a reason folks standardized on a hierarchical system. We'll see what happens, though.
Incidently, Windows is the only popular OS that still only supports a tree-style filesystem. Classic Mac OS, modern Mac OS, BSD and Linux all support any fully connected graph you might desire, thanks to symlinks (and on classic Mac OS, aliases).
d) This isn't actually all that new even from a UI perspective -- think of using Apple's Find File. Perhaps you toss in a few more search parameters to get at more metadata and data. The real difference is that traditionally, you must *also* assign a file a position in a hierarchical filesystem (though your hierarchical filesystem could potentially consist of just a single directory node with lots of files in it, a la the My Documents MS approach).
e) I'm remembering the last time the database research people at Microsoft convinced everyone that going all database would be a great idea. MS SQL, for a period of time, used tables internally for *everything*. Performance sucked, but they did it anyway. It's sexy from a theory point of view, because it simplifies things. Unfortunately, it throws out a lot of area-specific design knowledge that's been built up over the years. We know a lot about how to do a good filesystem, and there are features that apply nicely to filesystems that are less convenient with a traditional database. It's going to be tough to make a better system by throwing everything out.
f) I've heard ominous rumblings about WinFS being removed from Longhorn. It may or may not be living up to internal expectations.
g) Anytime something like this is announced *this* far in advance and isn't getting shown off in final form, it means that the promises frequently come from the research people. Research people have all sorts of rosy views about their own work, plenty of pet ideas, and may not have spent a long time doing usability tests. (This comes from one of those research people.) I wouldn't get excited about this any more than I would the frequent announcements on Slashdot about nonexistent new storage technologies ("in four years, we're all going to be using five terabyte 1 cubic centimeter Jell-O blobs to store our data!"). Yeah...come back when you have something shipping instead of a bunch of theoretical maximum numbers coming from a research team.
When you don't like a company, their methods, or their products, a free-market way of dealing with this would be to simply buy the products of their competitors. Is something like this not possible with employment as well?
Unfortunately, I don't believe so.
The simple free market economic model that is taught in Microeconomics I makes a number of assumptions that don't hold here. The first is that of perfect information. We have enough trouble getting accurate and trustable information to the public in the first place when it comes to products. Because of NDAs, the traditional secrecy surrounding salaries, and the fact that workers are generally only very familiar with their own company, people tend to not have a information basis for making comparative decisions about a company. Plus, as I pointed out, senior management exists in a position that is very easy for them to abuse, and because it is a relatively small number of people relative to the rest of the people in a company, usually isn't enough to sink the company.
People indirectly punish this behavior -- they will buy cheaper from another company that does not do the same thing -- but because upper managmeent is a small group, unit price generally isn't impacted all that much.
Finally, it's very difficult to differentiate workers on skill period -- workers want to positively represent themselves, worries about lawsuits prevent many past employers from describing an employee's past performance, and a bos only sees a portion of what an employee does. Finally, while it might be easy to rate a bricklayer (only a few metrics to gauge job performance on and all metrics are pretty visible -- you lay N bricks per hour and your laid bricks deviate from an ideally laid brick in M ways), CEO behavior tends to be very situation-specific, and difficult to judge.
The way I see it, the first step is to let the game companies sell the ROMS in whatever protected and limited fashon they are comfortable with, and then after they see a consumer demand for it, market forces would start letting competitive companies offer better EULAs
The main problem with this idea is that owners of copyrighted material have a natural monopoly over that material. Nobody else can *offer* that material but them, so there is no competition. You can get Phantasy Star from but one company.
There *are* multiple RPGs that you might play instead, but if you're, say, a Mega Man fan, Mario is not a simple substitute, you know?
I agree with your points, though I think that your claim that it's necessarily because someone else "worked harder" isn't necessarily true. Wanting to be the top dog is completely disconnected from how hard other people worked.
Spot on.
:-) ) works with the next closest field. He deals with knowledge about *reality* given some set of (usually) precisely defined axioms. His work is inherently relevant to reality, but not necessary particularly useful.
The entire point of mathematics is to produce perfect but abstract models. In some cases, these models are produced with no consideration for actual application. This is the job of the mathematician. He sets down a set of axioms, and granted those axioms, makes known-true claims about models. His claims are not "true" because they have anything to do with reality, but because they follow a commonly accepted set of axioms. The logician is a mathematician. Technically a statician is as well, but since statistics are usually used in an applied manner (a la science), it tends not to be found in math departments.
The philosopher (well, there are a lot of types of philosopher, but I like just going for the analytic ones and dumping all those ethicists
The scientist (distinct from *all* these) produces knowledge about what models currently are *believed to best apply* to reality given a (not precisely defined, but generally accepted among most folks) set of axioms. The scientist is not interested in absolute truth -- he deals with our current best guesses about reality.
A "pure" scientist is still part of this process...he just proposes models for use in representing reality that may not be easily testable, and may be "better" than other models only more subtle rules, like Ockham's Razor.
If you are a smaller ISV, (Independent Service Provider, in case you didn't know)
:-) ) but doesn't sell the hardware platform that the software runs on. Adobe's a good example of an ISV.
The acronym and the words don't match up.
Independent Software (occasionally Service or Solution) Vendor might be better.
An ISV is just a company that sells software (or services or solutions
And the poor schmucks that put up movies need to learn to use BitTorrent.