RMS and the FSF are a little more strict than that, though - they'll recommend against something whether that comparably good alternative exists or not.
That's incorrect--the analogy was drawn precisely from the Debian situation. The FSF used to recommend Debian because it was the best they could find, even though it wasn't 100% free. Then a 100% free distro became available, so the recommendation switched over.
Yes, that's correct. Animal rights people are happy to learn that vegetarian diets lead to fewer heart attacks, but they'd urge you to stay vegetarian even if it were established that that caused more heart attacks. Of course you're entitled to not take their advice, but asking them to start recommending non-vegetarian diets will take a much deeper and more fundamental argument than "you're not giving people enough options".
If you find those advantages so compelling that you're willing to trade your soul for them, RMS may simply prefer to sadly write you off, than to make concessions of his own soul.
The free software movement is like a group of people who decided to become vegetarian out of ethical concerns about animal rights. Not everybody thinks like them and they're practical enough to understand that. But suppose a Free Vegetable Movement starts a foundation to make vegetarian utensils, publish vegetarian cookbooks and so forth, and get a lot of followers. If non-vegetarians now start also using the recipes, that's fine with them. There's even a splinter "open vegetable movement" of people who don't care about the animal rights issues but have discovered the benefits of eating more vegetables (such as having fewer heart attacks). The OVM may have mixed meat/vegetable diets but the FVM doesn't want to have anything to do with that.
What's happening in these threads sounds to me like non-vegetarians somehow claiming the vegetarian foundation is foolishly restricting people's options because it won't link to restaurants that serve meat dishes, and no longer recommends a particular cookbook with good vegetarian recipes, because that cookbook also has meat dishes and there's now finally a comparably good cookbook which is 100% vegetarian. IMO it would be crazy for the veg foundation to do anything else, given its values. All you can decide is that its values are not your values. Asking them to turn against their very principles by also presenting the "meat option" is ridiculous (do you also ask your xtian church to present the "satan option"?).
They did a lot of work making their cookbooks and recipes what they are, and the changes you're asking for show that you're trying to impose your values on them, not the other way around.
I don't think they could make you show ID every tiem you send a letter. I imagine them making you show ID when you buy stamps. The stamps would each have a bar code (maybe in UV ink so you wouldn't see it under normal light) and they'd scan the bar codes and enter your ID info into a computer when you buy the stamps.
I wonder what they'll do about international mailing coming in to the country. They certainly can't make every country check people's ID's to sell them stamps.
I've sent dozens of fedex packages by going to the fedex office, filling in the carbonless form, and paying the shipping charge in cash. I always used my real name and address for the return address (had no reason to do otherwise), but have never been asked for any kind of ID. I could have written just about anything for a return address and sent it. So, no authentication.
I don't know how UPS does it because I haven't used UPS that way. I've shipped UPS packages from Mailbox Etc. type stores though, and again have never been asked for ID.
Wow, almost as good as an X-box and even cheaper
on
Lindows Webstation
·
· Score: 0, Troll
Now all we need is for someone to come out with a mod chip so we can run Linux on it.
Would an encrypting version of GNU Tar be prior art? I put Blowfish into GNU Tar in the mid 90s and
posted to Usenet about it in 1996 and at various other times. I've
offered to send out copies and a few people have asked for and gotten them. I'd think that constitutes publication.
They had a bogus data compression patent that they successfully sued Microsoft over. The result was that Microsoft bought out Stac for something like $120 million and incorporated the Stac algorithm into Disk Doubler.
These guys are probably hoping for a similar outcome, with a bigger payout.
There's been no evidence stated that Microsoft ever heard of those guys. Almost all software patents are for stuff that's obvious and Microsoft could have developed whatever it was completely independently of whatever that company did. So there's no basis to think that company is entitled to anything. Nobody forced them to do whatever it was that they did. It's better to just not have software patents. Those who find developing without getting patents doesn't pay high enough rewards, are at perfect liberty to do something else instead. They can invest in aerospace or biotechnology or hamburger chains instead of software development. The rest of us active developers will be much happier to be able to get our work done without have the threat of patent litigation over our heads.
No, I wasn't aware, so thanks, that was helpful. I've been looking at the arch webpage for the past couple minutes. I'll look at it some more, but it's a poor sign that I -still- haven't figured out what arch does that CVS doesn't (other than handle changesets and maybe have a better branching system than CVS). I also don't grok what it means to have a distributed repository. That doesn't make sense to me. There has to be some authoritative version of the tree somewhere. Of course individual maintainers should have their own copies of the tree to bang on, and maybe those can be used as mirrors for other maintainers to check out from, but I don't see why that's a distributed repository.
I continue to not understand what's so hard about writing these things. I also wonder why Arch is written in C, which these days I think of as a high level assembly language.
I have to be clear, I've never been an expert at this stuff. I've done basic setup of CVS and RCS and used them at work, and another work project used Perforce so I've used that for a while, but I probably just used basic functionality of each system and not the advanced stuff if that really exists. Still, from what I've seen flipping through the manuals for these programs, not counting GUI stuff that I didn't use, it looked like it would have been pretty simple to just bang out the code to implement any of them and do a much better job of concurrent access control, by just using a transactional database to handle the checkins. So I'm wondering the same thing as before, which is what the big deal is.
No offense but that just seems like a content-free answer. Re your points,
1) precision: every serious database, financial application, avionics code, medical instrumentation program, or anything else doing something critical faces the same reliability constraints and then some. Source control is nothing special. And reliable programs in all those application areas are being written all the time. Believing that there's something unusually difficult about implementing source control reliably compared with those other areas is ridiculous.
2) Sure, there's No Right Answers, but I asked for specific things where CVS delivers a Wrong Answer and haven't gotten any except for CVS's woeful inability to atomically commit and roll back groups of changes. So I have to repeat my question, what else is missing from CVS? You haven't named a single SPECIFIC thing that Bitkeeper does better than CVS does. I'm sure there are some, and I'm asking what they are.
3) There is no rule 3
4) That's just like any other area of programming, so you'll have to do better than that answer. With today's hardware I wouldn't make all the same choices CVS's designers made, however, I don't remember seeing any serious CVS performance failures even for quite large projects.
5) Again, I know CVS is being used for that purpose. Is it doing such a bad job? What do the other programs do that's better than how CVS does it? Please be SPECIFIC.
There's BitKeeper, Perforce, Subversion, and several other free and proprietary source control programs that I know of. I saw part of a panel about them at CodeCon and the main thing the implementers agreed on was that it's hard to write these things and get all the details right. I never could understand why. Whenever I've asked anyone what these fancy source control programs do that CVS doesn't, the only specific answer I've gotten is they allow committing and backing out groups of changes all at once instead of one at a time.
Sure, that's a valuable feature that I've missed in CVS, but come on people, if that's the only thing missing from CVS, why not just add it? If there's some implementation reason that can't be done in the CVS/RCS framework, how big a deal is it to even rewrite CVS from the ground up, totaly abandoning the RCS underpinnings but keeping the same command set (write the new code to do what the old documentation says) and implementing change groups? Use a transactional SQL database to manage issues of concurrent access and what the heck is the rest of the problem? CVS isn't THAT complicated.
I don't doubt I'm missing something, but I wish someone would spell out precisely what it is.
Remember Eldred vs. Ashcroft? Which branch of government do you think Ashcroft worked for at the time? Hint, he was Attorney General, which is an executive post, not legislative. And in that role he (represented by the Solicitor General, the guy who actually argues for the government in the Supreme Court) went all out defending the Mickey Mouse Protection Act. They were probably required to put up some kind of defense, but it could have been a token one (as sometimes happens when they realize the government position is actually wrong) and Eldred might have won.
And what about those judges who actually decided the case (making "case law")? They're not legislators either! They're the judicial branch, and how do they get on the bench? Right, they are appointed by (ta-da) the president! The president's court appointments will have far-reaching effects not only on copyright but on every other branch of law. And Clinton, for all the good stuff he did (and there was quite a lot), was basically a tool of Hollywood as far as copyright was concerned. Bush doesn't appear to be any better.
It is exceedingly in our interest to get a president who has his act together about copyright and other legislative issues. They have a tremendous amount of influence in every area.
I haven't downloaded that 18 MB zip file but the blurb on the download html page makes it sound like the zip contains binaries only. There is an accompanying offer to distribute source code on physical media for the cost of copying, saying you need to send contact and billing info to somewhere, and that the offer is valid for 3 years, language taken direct from the GPL. Basically they are being in-your-face about doing the absolute minimum that they can to conform to the letter of the GPL.
Given that OpenTV was previously in breach of the GPL and therefore had had its rights terminated, that they're now distributing the GPL'd stuff indicates that they came to an agreement with the FSF and the the FSF restored their rights. Personally I think FSF should have leaned on them a little harder and insisted on online source distribution before agreeing to restore rights, but that's just me. I do hope someone gets the source distribution and puts it up for download somewhere.
As much as I admire Eldred and Lessig, I have to say I don't like their proposal, which is a form of surrender.
The proposal is anti-constitutional because the whole point of the "limited times" language in the Constitution's copyright clause is to get the work out into the public domain while it's still worth something. If only worthless works (or works worth only $1 to the copyright holder) are supposed to enter the public domain, then there never would have been any controversy, and the battles over the Statute of Anne (the Eldred vs. Ashcroft of 18th century England, fought vigorously by the Stationers' Guild which was the publishing cartel of the time) would never have been fought.
Of course the SCOTUS decision in Eldred is even more anti-constitutional, but I'd rather that Lessig and Eldred seek a direct remedy, in the form of legislation or future court action. Eldred lost in the SCOTUS by a 7-2 margin, the same margin that upheld the WW1 anti-sedition law in a different free speech case, Abrams vs. US (1919) (we remember Abrams today mostly because of the "clear and present danger" test from Oliver Wendell Holmes' dissent). That law made it a felony to publish anything that criticized the U.S. Government. It wasn't until 1969, 50 years later, in the anti-Vietnam-war protest case Brandenburg vs. Ohio, that the anti-sedition law was overturned. Public awareness and opposition to government control had by then improved to the point that the SC had to disavow its previous mistake. The same thing can potentially happen in copyright reform and that's a goal to work toward.
The other argument againt the "PD Enhancement act" is that it's ineffectual. It's intended to free abandoned works, those whose copyright holders can't be located or can't be bothered. But at current long-term interest rates (6.25%) the value today of a $1.00 payment to be made 50 years in the future is less than 5 cents. That suggests that the same types of agencies that now act as clearinghouses for music licensing (jukebox royalties and so forth) are likely to just add another service to their offerings: pay us the $1.00 today with the catalog info for your literary copyright, and we'll send it in with your renewal paperwork 50 years from now, even if you're dead and gone. So the agency gets 95 cents profit on that present-day dollar, the work still never enters the public domain, and the Public Domain Enhancement Act merely accomplishes yet more transfer of wealth to the IP cartel. Certainly every big commercial publisher, the entities who pushed through the Sonny Bono term extension (Mickey Mouse protection act) in the first place, would never let any copyright lapse. Only noncommercial or ephemeral works, or maybe the occasional small press publication, would ever enter the public domain, again enlarging the disparity between the authors who actually create literary works and the media conglomerates which merchandise them.
Lessig and Eldred have put a lot of work and thought into the cause of the public domain and deserve our admiration; however, in my opinion, this is not one of their better-considered ideas.
It sounds pretty similar to what Smale and others proved decades ago, that the Simplex algorithm's running time is polynomial except on a measure-zero set of problem instances.
Smale, Blum et al have a really good book called "Complexity and Real Computation" which analyzes the Simplex Method and other algorithms at some length. It tries to build up a theoretical foundation for complexity analysis of numerical algorithms (where the objects being computed are real numbers), just like traditional complexity theory (with Turing machines) analyzes algorithms which compute on discrete sets (e.g. bits).
Face recognition is one of the TIA components Tien lists. I see a lot of folks wearing surgical masks to ward off SARS infection. I wonder if those masks could stay in style after the SARS panic tapers off, and whether there are some that cover enough of the face to defeat face recognition cameras.
IpsissimusMarr already asked that question but I figured it needs a new thread heading. That this didn't come out earlier, even as leaks, is suspicious. Xanadu-xtroot's explanation (they were waiting til their lawyers got the letter phrased absolutely right, that the assertions were correct, etc) reaches for plausibility but isn't entirely convincing.
How the heck did SCO buy "Unix" without buying the copyrights? Why have they been in discussion (dispute?) between SCO and Novell for the past several months? Novell's letter has qualifiers like "to our knowledge" when it says SCO doesn't own the copyrights. It sounds like the Novell-SCO agreement has been flawed all along and nobody knows what the real situation is.
I think there are yet more layers to this madness waiting to be unpeeled.
I guess if the sending domain doesn't return an RMX record, the receiving domain should treat the mail as authorized. The most commonly forged addresses are from AOL, Yahoo, Hotmail, etc. So if those big domains supply RMX, that should get rid of a lot of those forgeries. Of course it just means the forgers will switch to smaller domains that don't serve RMX...
The RMX record can return any IP addresses that it wants, the receiving machine just does a DNS lookup on the originating address and makes sure that IP is authorized to send mail. Read the RFC for more details.
You ask a good question and the answer lies in why we (well, some of us) go see such movies in the first place. We go not because they're deeply enlightening (read a book if you want that) or because they're that they get your pulse racing and palms sweaty (getting laid more often does a better job of that). Rather, we go see those things because they're part of our surrounding culture and if we don't see any of them it turns us into weirdos and recluses. When our friends groan "why does it have to be snakes?" we won't know what they mean. But if those movies didn't exist, we really wouldn't be any worse off, as we'd find plenty of other things to keep ourselves entertained and occupied.
So let those studios shut down, or (more likely and more desirably) go back to making smaller-budget films that can be profitable without needing to rake everyone over the coals and trying to extend copyright legislation into our bedrooms. I'd even stop boycotting DVD purchases if that happened (I rent them so the rental store gets the money instead of the MPAA).
The point about the big-theater experience and the expense of home theaters is valid, but it's not that big a deal to most people. I don't have an actual home theater and am a little extreme in not even having a TV set. But watching rented DVD's on my computer with small external amplified speakers works quite well, it's at least as good as watching on a TV. And as computer LCD projectors get cheaper, a real home theater-like experience gets both more affordable and less imposing on your living space (you can just have a little projector and roll-up screen that you can put away when not in use, instead of a huge rear-projection set that fills your room).
Slashdot Mirror
Yes, that's correct. Animal rights people are happy to learn that vegetarian diets lead to fewer heart attacks, but they'd urge you to stay vegetarian even if it were established that that caused more heart attacks. Of course you're entitled to not take their advice, but asking them to start recommending non-vegetarian diets will take a much deeper and more fundamental argument than "you're not giving people enough options".
The free software movement is like a group of people who decided to become vegetarian out of ethical concerns about animal rights. Not everybody thinks like them and they're practical enough to understand that. But suppose a Free Vegetable Movement starts a foundation to make vegetarian utensils, publish vegetarian cookbooks and so forth, and get a lot of followers. If non-vegetarians now start also using the recipes, that's fine with them. There's even a splinter "open vegetable movement" of people who don't care about the animal rights issues but have discovered the benefits of eating more vegetables (such as having fewer heart attacks). The OVM may have mixed meat/vegetable diets but the FVM doesn't want to have anything to do with that.
What's happening in these threads sounds to me like non-vegetarians somehow claiming the vegetarian foundation is foolishly restricting people's options because it won't link to restaurants that serve meat dishes, and no longer recommends a particular cookbook with good vegetarian recipes, because that cookbook also has meat dishes and there's now finally a comparably good cookbook which is 100% vegetarian. IMO it would be crazy for the veg foundation to do anything else, given its values. All you can decide is that its values are not your values. Asking them to turn against their very principles by also presenting the "meat option" is ridiculous (do you also ask your xtian church to present the "satan option"?). They did a lot of work making their cookbooks and recipes what they are, and the changes you're asking for show that you're trying to impose your values on them, not the other way around.
Interesting post, thanks.
I wonder what they'll do about international mailing coming in to the country. They certainly can't make every country check people's ID's to sell them stamps.
I've sent dozens of fedex packages by going to the fedex office, filling in the carbonless form, and paying the shipping charge in cash. I always used my real name and address for the return address (had no reason to do otherwise), but have never been asked for any kind of ID. I could have written just about anything for a return address and sent it. So, no authentication.
I don't know how UPS does it because I haven't used UPS that way. I've shipped UPS packages from Mailbox Etc. type stores though, and again have never been asked for ID.
Now all we need is for someone to come out with a mod chip so we can run Linux on it.
There's also a Usenet thread about encrypting archive programs including some modified Zip programs.
who is a Slashdotter. Trippi posted in the thread about Dean guest blogging for Lawrence Lessig a week or so ago.
These guys are probably hoping for a similar outcome, with a bigger payout.
There's been no evidence stated that Microsoft ever heard of those guys. Almost all software patents are for stuff that's obvious and Microsoft could have developed whatever it was completely independently of whatever that company did. So there's no basis to think that company is entitled to anything. Nobody forced them to do whatever it was that they did. It's better to just not have software patents. Those who find developing without getting patents doesn't pay high enough rewards, are at perfect liberty to do something else instead. They can invest in aerospace or biotechnology or hamburger chains instead of software development. The rest of us active developers will be much happier to be able to get our work done without have the threat of patent litigation over our heads.
I continue to not understand what's so hard about writing these things. I also wonder why Arch is written in C, which these days I think of as a high level assembly language.
I have to be clear, I've never been an expert at this stuff. I've done basic setup of CVS and RCS and used them at work, and another work project used Perforce so I've used that for a while, but I probably just used basic functionality of each system and not the advanced stuff if that really exists. Still, from what I've seen flipping through the manuals for these programs, not counting GUI stuff that I didn't use, it looked like it would have been pretty simple to just bang out the code to implement any of them and do a much better job of concurrent access control, by just using a transactional database to handle the checkins. So I'm wondering the same thing as before, which is what the big deal is.
No offense but that just seems like a content-free answer. Re your points,
1) precision: every serious database, financial application, avionics code, medical instrumentation program, or anything else doing something critical faces the same reliability constraints and then some. Source control is nothing special. And reliable programs in all those application areas are being written all the time. Believing that there's something unusually difficult about implementing source control reliably compared with those other areas is ridiculous.
2) Sure, there's No Right Answers, but I asked for specific things where CVS delivers a Wrong Answer and haven't gotten any except for CVS's woeful inability to atomically commit and roll back groups of changes. So I have to repeat my question, what else is missing from CVS? You haven't named a single SPECIFIC thing that Bitkeeper does better than CVS does. I'm sure there are some, and I'm asking what they are.
3) There is no rule 3
4) That's just like any other area of programming, so you'll have to do better than that answer. With today's hardware I wouldn't make all the same choices CVS's designers made, however, I don't remember seeing any serious CVS performance failures even for quite large projects.
5) Again, I know CVS is being used for that purpose. Is it doing such a bad job? What do the other programs do that's better than how CVS does it? Please be SPECIFIC.
Sure, that's a valuable feature that I've missed in CVS, but come on people, if that's the only thing missing from CVS, why not just add it? If there's some implementation reason that can't be done in the CVS/RCS framework, how big a deal is it to even rewrite CVS from the ground up, totaly abandoning the RCS underpinnings but keeping the same command set (write the new code to do what the old documentation says) and implementing change groups? Use a transactional SQL database to manage issues of concurrent access and what the heck is the rest of the problem? CVS isn't THAT complicated.
I don't doubt I'm missing something, but I wish someone would spell out precisely what it is.
And what about those judges who actually decided the case (making "case law")? They're not legislators either! They're the judicial branch, and how do they get on the bench? Right, they are appointed by (ta-da) the president! The president's court appointments will have far-reaching effects not only on copyright but on every other branch of law. And Clinton, for all the good stuff he did (and there was quite a lot), was basically a tool of Hollywood as far as copyright was concerned. Bush doesn't appear to be any better.
It is exceedingly in our interest to get a president who has his act together about copyright and other legislative issues. They have a tremendous amount of influence in every area.
Someone should still order the CD-ROM and make sure all the stuff on it is also in the download.
I haven't downloaded that 18 MB zip file but the blurb on the download html page makes it sound like the zip contains binaries only. There is an accompanying offer to distribute source code on physical media for the cost of copying, saying you need to send contact and billing info to somewhere, and that the offer is valid for 3 years, language taken direct from the GPL. Basically they are being in-your-face about doing the absolute minimum that they can to conform to the letter of the GPL.
Given that OpenTV was previously in breach of the GPL and therefore had had its rights terminated, that they're now distributing the GPL'd stuff indicates that they came to an agreement with the FSF and the the FSF restored their rights. Personally I think FSF should have leaned on them a little harder and insisted on online source distribution before agreeing to restore rights, but that's just me. I do hope someone gets the source distribution and puts it up for download somewhere.
The proposal is anti-constitutional because the whole point of the "limited times" language in the Constitution's copyright clause is to get the work out into the public domain while it's still worth something. If only worthless works (or works worth only $1 to the copyright holder) are supposed to enter the public domain, then there never would have been any controversy, and the battles over the Statute of Anne (the Eldred vs. Ashcroft of 18th century England, fought vigorously by the Stationers' Guild which was the publishing cartel of the time) would never have been fought.
Of course the SCOTUS decision in Eldred is even more anti-constitutional, but I'd rather that Lessig and Eldred seek a direct remedy, in the form of legislation or future court action. Eldred lost in the SCOTUS by a 7-2 margin, the same margin that upheld the WW1 anti-sedition law in a different free speech case, Abrams vs. US (1919) (we remember Abrams today mostly because of the "clear and present danger" test from Oliver Wendell Holmes' dissent). That law made it a felony to publish anything that criticized the U.S. Government. It wasn't until 1969, 50 years later, in the anti-Vietnam-war protest case Brandenburg vs. Ohio, that the anti-sedition law was overturned. Public awareness and opposition to government control had by then improved to the point that the SC had to disavow its previous mistake. The same thing can potentially happen in copyright reform and that's a goal to work toward.
The other argument againt the "PD Enhancement act" is that it's ineffectual. It's intended to free abandoned works, those whose copyright holders can't be located or can't be bothered. But at current long-term interest rates (6.25%) the value today of a $1.00 payment to be made 50 years in the future is less than 5 cents. That suggests that the same types of agencies that now act as clearinghouses for music licensing (jukebox royalties and so forth) are likely to just add another service to their offerings: pay us the $1.00 today with the catalog info for your literary copyright, and we'll send it in with your renewal paperwork 50 years from now, even if you're dead and gone. So the agency gets 95 cents profit on that present-day dollar, the work still never enters the public domain, and the Public Domain Enhancement Act merely accomplishes yet more transfer of wealth to the IP cartel. Certainly every big commercial publisher, the entities who pushed through the Sonny Bono term extension (Mickey Mouse protection act) in the first place, would never let any copyright lapse. Only noncommercial or ephemeral works, or maybe the occasional small press publication, would ever enter the public domain, again enlarging the disparity between the authors who actually create literary works and the media conglomerates which merchandise them.
Lessig and Eldred have put a lot of work and thought into the cause of the public domain and deserve our admiration; however, in my opinion, this is not one of their better-considered ideas.
Disclaimer: IMHO, IANAL, etc.
Smale, Blum et al have a really good book called "Complexity and Real Computation" which analyzes the Simplex Method and other algorithms at some length. It tries to build up a theoretical foundation for complexity analysis of numerical algorithms (where the objects being computed are real numbers), just like traditional complexity theory (with Turing machines) analyzes algorithms which compute on discrete sets (e.g. bits).
slightly different from Pynchon version...
Kill two birds with one stone.
How the heck did SCO buy "Unix" without buying the copyrights? Why have they been in discussion (dispute?) between SCO and Novell for the past several months? Novell's letter has qualifiers like "to our knowledge" when it says SCO doesn't own the copyrights. It sounds like the Novell-SCO agreement has been flawed all along and nobody knows what the real situation is.
I think there are yet more layers to this madness waiting to be unpeeled.
I guess if the sending domain doesn't return an RMX record, the receiving domain should treat the mail as authorized. The most commonly forged addresses are from AOL, Yahoo, Hotmail, etc. So if those big domains supply RMX, that should get rid of a lot of those forgeries. Of course it just means the forgers will switch to smaller domains that don't serve RMX...
The RMX record can return any IP addresses that it wants, the receiving machine just does a DNS lookup on the originating address and makes sure that IP is authorized to send mail. Read the RFC for more details.
So let those studios shut down, or (more likely and more desirably) go back to making smaller-budget films that can be profitable without needing to rake everyone over the coals and trying to extend copyright legislation into our bedrooms. I'd even stop boycotting DVD purchases if that happened (I rent them so the rental store gets the money instead of the MPAA).
The point about the big-theater experience and the expense of home theaters is valid, but it's not that big a deal to most people. I don't have an actual home theater and am a little extreme in not even having a TV set. But watching rented DVD's on my computer with small external amplified speakers works quite well, it's at least as good as watching on a TV. And as computer LCD projectors get cheaper, a real home theater-like experience gets both more affordable and less imposing on your living space (you can just have a little projector and roll-up screen that you can put away when not in use, instead of a huge rear-projection set that fills your room).