From the PPI site: PPI proposes the following changes to the DMCA[...]
Please visit the PPI site for more info. The link is in the article.
Re:E-Commerce has a long way to go
on
Boo No More
·
· Score: 2
Just some prediction
Just some nay-saying.
Now fontier mode does sound very impressive, but what people need most cannot be brought to them via the Internet.
Many of my co-workers buy their food over the Internet. Looking for a home without the use of the Internet is starting to become "old school". And as far as companionship, I can't count the number of couples that I know who met on-line.
What sorts of goods did you have in mind? Cars? Businesses? Office supplies? Sex toys? Contract services? Intellectual property research materials? Financial services?
Let's not start revising history before it's made. The Internet (and commerce thereon) is revolutionizing the way we do things. WILL we all fall back into the same human patterns as before, but with the Internet as a piece of our lives? Of course; that's what humans do. But it's also what humans did as a result of the industrial revolution.
This has to be one of the 10 best Slashdot posts I've seen. If I hadn't squandered my points, I'd be moderating it up myself (er, of course, now I've posted to this thread...)
To bad there isn't a "Funny, Informative, Interesting, Underrated, Insightful" option.
Re:E-Commerce has a long way to go
on
Boo No More
·
· Score: 2
Civil unrest, environmental catastrophe, and ethnic cleansing? You may like it, but I'm heading to my cabin in Montana!
Of course. Civil unrest is ongoing everywhere in the world with hot-spots popping up at the rate of at least one per day.
Catastrophe... how about the fires in the southwest? Of course, someone will pull a major boner in the world of e-commerce, and some group of elderly won't get their medicine. It will happen. Legislation will be proposed. Nothing will happen. Happens in every industry.
Ethnic clensing? Well, if you think of the traditional hacker (MIT definition) as the native Internet inhabitant....
Yes, frontier economies lead to frontier lifestyles. The bad bank-robbers and pirates and stage-holdups will happen. The 50s were a time of tremendous prosperity, but also of massive coruption and misuse. This is human nature, and hiding away from civilization is not really a very good solution.
E-Commerce has a long way to go
on
Boo No More
·
· Score: 3
There will be ups and downs, but the overall internet-retail market will have years of growth ahead of it. It will begin to mature, and that means some of the frontier nature will be gone, but it will certainly continue to be profitable and a growth industry.
I keep trying to explain to people that our economy right now is in "frontier" mode. For examples of what to expect, look to the international expansion of the US economy post-war (both world wars, really); the beginning of the industrial age; the western expansion of the US, etc.
The company can require that you keep the changes proprietary unless you do them on your own time. The GPL is a set of rules for how you treat people to whom you distribute. If the company tells you not to distribute this new version, then they are well within their rights to do so (unless your contract says otherwise).
That said however, I suggest you talk to the folks that you work for, explain that this is your program from prior employment and that you want to contribute this work back just as you contributed it to them in the first place. If they don't like that sort of reasoning, then I suggest you don't want to work for them, and it's a very good market for coders right now....
Raging seems to be heavily influenced by Google. I wonder if there was a deal between the two that fell through, or if AltaVista simply wanted some of the "all we want is a search engine" market....
Either way, google is just want I need, and all I have on my home page.
To clarify, the format I'm proposing is to facilitate relays (such as gnut) using the extra info as searchable items. This means that you can find a review on a Web site; think: hey, this is cool and cut/paste the MD5 checksum into your gnutella client search facility to find the right file. You may not get it from the site that the reviewer did, but you know the MD5 is the same (after you double-check, which your client could potentially to for you, automatically). Also, the optional fourth part allows the reverse (looking up the reviews for a file that you just got).
Note: I'm not proposing an MP3, MPG, AVI, EXE, ZIP or any other format replacement. Whatever you're shipping can go in the third part, unchanged (using the appropriate MIME type). This just allows for tagging of files in searchable and reviewable ways that protect privacy at the same time.
This is a long time coming. Back in the early '90s I was saying that it would be a long time (perhaps as much as 10 years) before the information industries of music, books, research, advertising, etc. started to notice that their world had ended and the Internet had taken over. This war has only just begun, let's hope it's not going to be a bloody one (a vain hope, I'm sure as RIAA will not stop before SOMEONE is sent to jail as an example).
I've just started to check out Gnutella, not because it's really that interesting a service to me, but because the idea of a truely distributed network is intriguing. I think I would have done it differently, and indexing/reviewing is going to have to be the next big advent there, but it certainly has the chance of being to the next generation of Internet/Web usage what gopher was to the pre-Web days (which, if you don't reacall was HUGE on the relatively small scale of the Internet back then).
As a help to growth, I propose a new file format. Simply a MIME document with the extension ".gtl" (itself having the mime-type x-application/gnutella-catalog), which contains three required parts. The first is text/plain (description) the second is text/plain and contains one line that has an MD5 checksum of the third part (used as a key for refering to this document) the third part is of any format, but application/pgp is specifically allowed for (the content: signed and optionally encrypted). If the content is signed, the signature is that, not of the content creator or distributor, but of a review/indexing site that has verified and cataloged the content, perhaps also providing a public forum for user-feedback regarding the content. An optional fourth part would contain a text/plain one-entry-per-line list of URLs where the content is indexed and/or reviewed.
This is the same Stallman who thinks that Linux should be called GNU/Linux but that all of the other contributors (e.g. MIT/X, BSD, etc) should be ignored in the naming?
Let's not try to take the moral high-ground by quoting Stallman; it's always a bad plan.
I just wanted to say thanks to the folks as Slashdot. Too many times we've been quick to slam them for posting something that turned out to be rumor, so I wanted to step up and say that it's good to see the research and source checking happen before the article hits the homepage. Slashdot may indeed be growing up into a mature (and already quite powerful) news outlet, which we will be able to rely on for Stuff That Matters well into the oughts.
A number of posters have said (seriously and jokingly) that the TLD namespace should be opened up to any names. This would be a massive mistake, given the way things are currently set up.
First off, you increase the problem drastically. What is.aclu? Is it THE ACLU? Nope, they've got aclu.org. Must be someone else. Ah, Albert Clemens Lucifer Ulbritch.... The namespace is broken up a little right now..net and.com are not synonymous, but there is not enough distinction (NSI's fault, really)..org is clearly distinguished, and a lot of companies do not bother buying.org because they understand the meaning of it, and it holds no value for them.
I really think that.per should exist for individuals..shop isn't so bad, but there will be a lot of companies that will buy a.shop just because they might want to someday sell something.:-(
The real solution is to start charging big money for domain names (first create a cheap.per). If a domain costs $1000/yr to register, most companies will think twice about picking up 100 of them. This seems high, but think about the cost of a sign for your business. What about a new office? Enough computers to run a small department? Domain names are insanely cheap, and the price needs to go up to reflect the value of domain names as a commodity.
Area 51 is still used for test flights, but as I understand it, all of the really interesting stuff (in terms of the latest and greatest prototype planes) is now elsewhere (including Denver, where high-altitude tests take place).
But, still. As a former sysadmin for IKS (a gimp-like programmers library from the late 80s) I can still appreciate good sattelite images. I still chuckle at the story someone told of showing off the project's demo where a de-classified image of a plane was enhanced to the point that you could make out the numbers on the wing. They were showing this at a trade show, and some air-force type comes by and his jaw hits the floor. Apparently he was the one who de-classified the image;-)
Soon thereafter they went back to using some playmate from the 70s....
There are basically two kinds of software users out there: the ones who really care about security and the ones that don't. The ones that do can be recognized by trappings like: no software gets installed without the software security staff reading the results of an independant security audit based on access to the source; no users are allowed to install anything, only security admins can authorize that; It's someone's full time job to stay on top of security notices and updates/patches; etc.
The type that does not care will often protest that they do. However, they then turn around and say things like "we'll just use XYZ vendor's product because they're a large stable company that, it would seem, would have no reason to wish us ill."
We can ignore the second type of company (what the article in question is really talking about) because they will be insecure no matter WHAT they are running. Running MacOS just makes them slightly more secure because that's not what the script kiddies are looking for. If they're going to install Red Hat 6.0 and walk away from it for 2 years, assuming that it's "safe", they'll be wrong. There are numerous security problems with any OS release, and if you don't stay on top of them (and/or find them up front) you're screwed.
In the case of the company that really cares, OSS is much more attractive. For starters, you can re-compile every single binary from source. Second, you have the ability to fix the bugs that your internal security audits fix, while you wait for an official patch. For mission-crittical software, this can be immesurably important.
I remember dealing with a market-data vendor when I was working for a medium-sized financial firm. They wanted to pump their proprietary protocol over the Internet and through our firewall, so I said "no problem, just a) give us your source so that we can perform a security review or provide us with the results of an independant security review in writing." They litterally laughed in my face. Needless to say, their data feed did not happen.
Not everyone cares this much, but when you do, Open Source solves a lot of problems and makes many things much easier.
Yes and no. I agree with you that a lot of companies will go under, but perhaps not nearly as many as with a hard good such as airlines. Remember, the cost of production is low, so the value is in the code, processes and markets. This makes every company that does ok into an attractive buyout target for a traditional business looking to get in on the action.... Look for lots more consolidation later this year.
Yes, but if you plot Linux stocks vs tech IPOs for the last year, you will find a very strong corelation. LNUX was a supply/demand phenomenon. Now that the "players" are out, you will see it stabelize a bit and likely even go up as the market eases back into the frontier-driven climb.
I saw the VHS in a local store and was tempted, but I just don't want to own any more video tapes. They're big, the're clunky, they break and the picture quality sucks (though, I hope the DVD-compression hardware gets better, I'm sick of seeing squares in dark backgrounds).
Look, a lot of people were announcing an NT security hole. Slashdot reported it too. Now, I agree that Slashdot should have a team of investigative reporters who have the tecnical credits to figure out if this is true or not, but that's because I have a very different vision of what Slashdot should be than, say, CmdrTaco. I don't begrude him his site as it is, but feel it would be much more useful as a validating filter on the poor high-tech reporting that goes on in other outlets.
The story is still up in the air as far as I'm concerned. One guy (who, BTW was not the original discoverer of the exploit) is reporting that Microsoft doesn't think there's an exploit.
I want to see some people grab the exploit script (it's on the real bugtraq) and run it against some test servers with valid permissions. Does it work? How invalid do the permissions have to be? Does the Microsoft documentation lead you down the road of "invalid permissions" for settting up virtual hosts?
Many questions need to be answered before this case is closed....
So, in case you haven't red the bug report, the specific password in question is "Netscape engineers are weenies!"
Oh, I love Microsoft's well-developed sense of responsibility and mature approach to the market:-)
So I guess people are backing off because you have to have publishing rights, but the ugly part is that you only have to have publishing rights to one of the virutual hosts on a server to get all of the.asp(s) from it.
I'll have to peruse the Ars Technica comments to see why they don't consider this a back-door.
A lot of people here are talking about how Linux was overvalued or how Techs were overvalued. I hate to break it to you, but that's a very nearsighted point-of-view. The reality is that the Market was overvalued and the Market corrected.
The good news is that technology stocks weren't strong because it was a fad. Technology stocks were strong because there was a ton of money coming into the market trying to take advantage of what is obviously a major shift in the way humans interact with eachother and in the infrastructure required to do it. That sort of upheval is what leads to frontier markets like the international trade situation after WWII or the gold-rush (which was not so much about gold as it was about settling half of a continent, in terms of investment and return).
This period of history will have its winners and its loosers, but I don't think that this correction will signal the last run-up or the last drop. We're in for a wild and bumpy ride for at least the next five years.
Note to Taco: you should be interested. The stock market is perhaps the single capitolist endevor that most resembles open-source, and as such it tends to demonstrate a lot of the pit-falls that we have to look forward to. How so? Well, can you think of another example where total disclosure is made, value is assessed simply on results, and each new endevor is free to use the same tactics demonstrated by earlier projects? The businesses themselves are nothing like Open Source, but if you think of each stock as a project on its own....
Did you see the live-action version of this? It got canceled, but it was the one where the earth will be pretty much dead withing a small number of years unless our heroes leave in an alien-tech ship and find the solution. They have only their skills and a really big gun front-mounted in the nose of the ship that they can only fire in the most dire situations.
Oh sorry, that was Crusade;-)
C'mon, Galen; tell them that we're looking for Trillana!
Would it be possible to write an abstraction layer above Qt/KDE and Gtk/GNOME that allowed you to replace the widget set at compile-time? If so, your installer could detect the libraries that you've got and install that version.
I know GNOME, and I know that just about everything is ultra-generic to the point that it makes sense to use XML descriptions of your user-interface and generate code in whatever language you like. We take the performance hit for internationalization. We take the performance hit for distributed objects. If we could just write an interface layer that abstracts the code the programmer has to write, then we'd be pretty much done, and I can't believe that the performance hit would compare to CORBA.
Let's see. You'd need some common way of getting at objects, but that's easy. Use CORBA, and write an ORBIt-KOM interface. Miguel was saying that this would be hard, but I think with enough programmer eyes on the task, it would be simple.
Of course, in the end this would be the death-knell of KDE. KDE relies on a widget set that's almost entirely free. GNOME relies on one that *is* free. If there's no real difference other than the licensing....
In the end, that's irrelevant. What matters is that people not feel hobbled so that they have to use extinct toolkits to write their apps (I'm sick of Netscape screwing up my theme, and I won't fall for it again).
Linux and *BSD are unusable as an OS design teaching tool due to their complexity.
I don't agree. You could certainly do a "whirlwind tour of the Linux kernel" in a semester and give the students assignments that involve augmenting the kernel in small ways that touch important subsystems. Yes, you can't get your head around the whole system, but that's the nature of large-scale programming, and working in that environment is practically a given in the industry.
The important part is that the students learn why an operating system works the way it does, and how the OS that they are studying has tackled certain problems. I would cover: VM, task/process management, kernel/user space issues in driver writing, writing a loadable module. I think this would give them a good head-start on the lay of the land. If I had time, or as extra credit, I would go into the interface between the core kernel and something like the IP stack. Either that or a few SMP features that would give them a sense of the depth of that abyss.
Now take the student who uses Minix as his/her example. There's no sense of how modern hardware issues are tackled. None of the complexity that stems from supporting every $0.25 ethernet card on the planet. How exactly is that supposed to introduce MODERN OS design. I mean, you could always teach them Tandy Color Computer BASIC/OS. The assembly exists in several well-annotated forms....
Interesting reading different people's comments. I especially love the ones that start: "They broke what made Star Trek great. What I really loved was..." The basic flaw in that line of reasoning is: a) most of you have no idea what it was about Star Trek that made it work for you, or you'd be writing your own series and b) what made it work for you is almost certainly the one thing that someone else hated.
Personally, I think it just doesn't matter. The basic problem that Star Trek has isn't acting or writing or special effects or evil execs. It's us.
The fans are really quite blood-thirsty at times, and that has to be difficult to deal with. Not in terms of ego, or hurt feelings, but in terms of audience. When the most vocal part of your audience decries everything you do, you can't assess what you should do.
This is why the series that succede in the fans eyes are always the new upstarts. They have no expectations, so the fans judge them, pretty much, at face value (as examples: Babylon 5, Farscape, X-Files). What Star Trek needs is to back off and wait a few years, but the franchise can't afford that, so they'll plow forward with what the focus groups tell them is hated least, and we'll get the blandest possible thing with a lot of things blowing up, and just enough skin to keep the football-aint-on-yet, channel-surfing, what-the-hell-is-this-shit-with-the-fucked-up-fore heads crowd happy.
They could try being truely daring, and do something that will piss off half of the fans and galvenize the rest into advocates who will pull in the next generation of fans, but that won't fly with the shareholders.
In the end, Trek is dead because fans and studios can't work together. It's both of us that did it, but we'll keep pointing fingers until the last show spits up blood. Of course, then we'll just say: look what you did....
Slashdot Mirror
They aren't saying modify the DMCA
From the PPI site: PPI proposes the following changes to the DMCA[...]
Please visit the PPI site for more info. The link is in the article.
Just some prediction
Just some nay-saying.
Now fontier mode does sound very impressive, but what people need most cannot be brought to them via the Internet.
Many of my co-workers buy their food over the Internet. Looking for a home without the use of the Internet is starting to become "old school". And as far as companionship, I can't count the number of couples that I know who met on-line.
What sorts of goods did you have in mind? Cars? Businesses? Office supplies? Sex toys? Contract services? Intellectual property research materials? Financial services?
Let's not start revising history before it's made. The Internet (and commerce thereon) is revolutionizing the way we do things. WILL we all fall back into the same human patterns as before, but with the Internet as a piece of our lives? Of course; that's what humans do. But it's also what humans did as a result of the industrial revolution.
This has to be one of the 10 best Slashdot posts I've seen. If I hadn't squandered my points, I'd be moderating it up myself (er, of course, now I've posted to this thread...)
To bad there isn't a "Funny, Informative, Interesting, Underrated, Insightful" option.
Civil unrest, environmental catastrophe, and ethnic cleansing? You may like it, but I'm heading to my cabin in Montana!
Of course. Civil unrest is ongoing everywhere in the world with hot-spots popping up at the rate of at least one per day.
Catastrophe... how about the fires in the southwest? Of course, someone will pull a major boner in the world of e-commerce, and some group of elderly won't get their medicine. It will happen. Legislation will be proposed. Nothing will happen. Happens in every industry.
Ethnic clensing? Well, if you think of the traditional hacker (MIT definition) as the native Internet inhabitant....
Yes, frontier economies lead to frontier lifestyles. The bad bank-robbers and pirates and stage-holdups will happen. The 50s were a time of tremendous prosperity, but also of massive coruption and misuse. This is human nature, and hiding away from civilization is not really a very good solution.
There will be ups and downs, but the overall internet-retail market will have years of growth ahead of it. It will begin to mature, and that means some of the frontier nature will be gone, but it will certainly continue to be profitable and a growth industry.
I keep trying to explain to people that our economy right now is in "frontier" mode. For examples of what to expect, look to the international expansion of the US economy post-war (both world wars, really); the beginning of the industrial age; the western expansion of the US, etc.
The company can require that you keep the changes proprietary unless you do them on your own time. The GPL is a set of rules for how you treat people to whom you distribute. If the company tells you not to distribute this new version, then they are well within their rights to do so (unless your contract says otherwise).
That said however, I suggest you talk to the folks that you work for, explain that this is your program from prior employment and that you want to contribute this work back just as you contributed it to them in the first place. If they don't like that sort of reasoning, then I suggest you don't want to work for them, and it's a very good market for coders right now....
Raging seems to be heavily influenced by Google. I wonder if there was a deal between the two that fell through, or if AltaVista simply wanted some of the "all we want is a search engine" market....
Either way, google is just want I need, and all I have on my home page.
To clarify, the format I'm proposing is to facilitate relays (such as gnut) using the extra info as searchable items. This means that you can find a review on a Web site; think: hey, this is cool and cut/paste the MD5 checksum into your gnutella client search facility to find the right file. You may not get it from the site that the reviewer did, but you know the MD5 is the same (after you double-check, which your client could potentially to for you, automatically). Also, the optional fourth part allows the reverse (looking up the reviews for a file that you just got).
Note: I'm not proposing an MP3, MPG, AVI, EXE, ZIP or any other format replacement. Whatever you're shipping can go in the third part, unchanged (using the appropriate MIME type). This just allows for tagging of files in searchable and reviewable ways that protect privacy at the same time.
This is a long time coming. Back in the early '90s I was saying that it would be a long time (perhaps as much as 10 years) before the information industries of music, books, research, advertising, etc. started to notice that their world had ended and the Internet had taken over. This war has only just begun, let's hope it's not going to be a bloody one (a vain hope, I'm sure as RIAA will not stop before SOMEONE is sent to jail as an example).
I've just started to check out Gnutella, not because it's really that interesting a service to me, but because the idea of a truely distributed network is intriguing. I think I would have done it differently, and indexing/reviewing is going to have to be the next big advent there, but it certainly has the chance of being to the next generation of Internet/Web usage what gopher was to the pre-Web days (which, if you don't reacall was HUGE on the relatively small scale of the Internet back then).
As a help to growth, I propose a new file format. Simply a MIME document with the extension ".gtl" (itself having the mime-type x-application/gnutella-catalog), which contains three required parts. The first is text/plain (description) the second is text/plain and contains one line that has an MD5 checksum of the third part (used as a key for refering to this document) the third part is of any format, but application/pgp is specifically allowed for (the content: signed and optionally encrypted). If the content is signed, the signature is that, not of the content creator or distributor, but of a review/indexing site that has verified and cataloged the content, perhaps also providing a public forum for user-feedback regarding the content. An optional fourth part would contain a text/plain one-entry-per-line list of URLs where the content is indexed and/or reviewed.
Can anyone else smell a business model, here...?
This is the same Stallman who thinks that Linux should be called GNU/Linux but that all of the other contributors (e.g. MIT/X, BSD, etc) should be ignored in the naming?
Let's not try to take the moral high-ground by quoting Stallman; it's always a bad plan.
I just wanted to say thanks to the folks as Slashdot. Too many times we've been quick to slam them for posting something that turned out to be rumor, so I wanted to step up and say that it's good to see the research and source checking happen before the article hits the homepage. Slashdot may indeed be growing up into a mature (and already quite powerful) news outlet, which we will be able to rely on for Stuff That Matters well into the oughts.
A number of posters have said (seriously and jokingly) that the TLD namespace should be opened up to any names. This would be a massive mistake, given the way things are currently set up.
.aclu? Is it THE ACLU? Nope, they've got aclu.org. Must be someone else. Ah, Albert Clemens Lucifer Ulbritch.... The namespace is broken up a little right now. .net and .com are not synonymous, but there is not enough distinction (NSI's fault, really). .org is clearly distinguished, and a lot of companies do not bother buying .org because they understand the meaning of it, and it holds no value for them.
.per should exist for individuals. .shop isn't so bad, but there will be a lot of companies that will buy a .shop just because they might want to someday sell something. :-(
.per). If a domain costs $1000/yr to register, most companies will think twice about picking up 100 of them. This seems high, but think about the cost of a sign for your business. What about a new office? Enough computers to run a small department? Domain names are insanely cheap, and the price needs to go up to reflect the value of domain names as a commodity.
First off, you increase the problem drastically. What is
I really think that
The real solution is to start charging big money for domain names (first create a cheap
Area 51 is still used for test flights, but as I understand it, all of the really interesting stuff (in terms of the latest and greatest prototype planes) is now elsewhere (including Denver, where high-altitude tests take place).
;-)
But, still. As a former sysadmin for IKS (a gimp-like programmers library from the late 80s) I can still appreciate good sattelite images. I still chuckle at the story someone told of showing off the project's demo where a de-classified image of a plane was enhanced to the point that you could make out the numbers on the wing. They were showing this at a trade show, and some air-force type comes by and his jaw hits the floor. Apparently he was the one who de-classified the image
Soon thereafter they went back to using some playmate from the 70s....
There are basically two kinds of software users out there: the ones who really care about security and the ones that don't. The ones that do can be recognized by trappings like: no software gets installed without the software security staff reading the results of an independant security audit based on access to the source; no users are allowed to install anything, only security admins can authorize that; It's someone's full time job to stay on top of security notices and updates/patches; etc.
The type that does not care will often protest that they do. However, they then turn around and say things like "we'll just use XYZ vendor's product because they're a large stable company that, it would seem, would have no reason to wish us ill."
We can ignore the second type of company (what the article in question is really talking about) because they will be insecure no matter WHAT they are running. Running MacOS just makes them slightly more secure because that's not what the script kiddies are looking for. If they're going to install Red Hat 6.0 and walk away from it for 2 years, assuming that it's "safe", they'll be wrong. There are numerous security problems with any OS release, and if you don't stay on top of them (and/or find them up front) you're screwed.
In the case of the company that really cares, OSS is much more attractive. For starters, you can re-compile every single binary from source. Second, you have the ability to fix the bugs that your internal security audits fix, while you wait for an official patch. For mission-crittical software, this can be immesurably important.
I remember dealing with a market-data vendor when I was working for a medium-sized financial firm. They wanted to pump their proprietary protocol over the Internet and through our firewall, so I said "no problem, just a) give us your source so that we can perform a security review or provide us with the results of an independant security review in writing." They litterally laughed in my face. Needless to say, their data feed did not happen.
Not everyone cares this much, but when you do, Open Source solves a lot of problems and makes many things much easier.
Yes and no. I agree with you that a lot of companies will go under, but perhaps not nearly as many as with a hard good such as airlines. Remember, the cost of production is low, so the value is in the code, processes and markets. This makes every company that does ok into an attractive buyout target for a traditional business looking to get in on the action.... Look for lots more consolidation later this year.
Yes, but if you plot Linux stocks vs tech IPOs for the last year, you will find a very strong corelation. LNUX was a supply/demand phenomenon. Now that the "players" are out, you will see it stabelize a bit and likely even go up as the market eases back into the frontier-driven climb.
I saw the VHS in a local store and was tempted, but I just don't want to own any more video tapes. They're big, the're clunky, they break and the picture quality sucks (though, I hope the DVD-compression hardware gets better, I'm sick of seeing squares in dark backgrounds).
Look, a lot of people were announcing an NT security hole. Slashdot reported it too. Now, I agree that Slashdot should have a team of investigative reporters who have the tecnical credits to figure out if this is true or not, but that's because I have a very different vision of what Slashdot should be than, say, CmdrTaco. I don't begrude him his site as it is, but feel it would be much more useful as a validating filter on the poor high-tech reporting that goes on in other outlets.
The story is still up in the air as far as I'm concerned. One guy (who, BTW was not the original discoverer of the exploit) is reporting that Microsoft doesn't think there's an exploit.
I want to see some people grab the exploit script (it's on the real bugtraq) and run it against some test servers with valid permissions. Does it work? How invalid do the permissions have to be? Does the Microsoft documentation lead you down the road of "invalid permissions" for settting up virtual hosts?
Many questions need to be answered before this case is closed....
So, in case you haven't red the bug report, the specific password in question is "Netscape engineers are weenies!"
:-)
.asp(s) from it.
Oh, I love Microsoft's well-developed sense of responsibility and mature approach to the market
So I guess people are backing off because you have to have publishing rights, but the ugly part is that you only have to have publishing rights to one of the virutual hosts on a server to get all of the
I'll have to peruse the Ars Technica comments to see why they don't consider this a back-door.
A lot of people here are talking about how Linux was overvalued or how Techs were overvalued. I hate to break it to you, but that's a very nearsighted point-of-view. The reality is that the Market was overvalued and the Market corrected.
The good news is that technology stocks weren't strong because it was a fad. Technology stocks were strong because there was a ton of money coming into the market trying to take advantage of what is obviously a major shift in the way humans interact with eachother and in the infrastructure required to do it. That sort of upheval is what leads to frontier markets like the international trade situation after WWII or the gold-rush (which was not so much about gold as it was about settling half of a continent, in terms of investment and return).
This period of history will have its winners and its loosers, but I don't think that this correction will signal the last run-up or the last drop. We're in for a wild and bumpy ride for at least the next five years.
Note to Taco: you should be interested. The stock market is perhaps the single capitolist endevor that most resembles open-source, and as such it tends to demonstrate a lot of the pit-falls that we have to look forward to. How so? Well, can you think of another example where total disclosure is made, value is assessed simply on results, and each new endevor is free to use the same tactics demonstrated by earlier projects? The businesses themselves are nothing like Open Source, but if you think of each stock as a project on its own....
Did you see the live-action version of this? It got canceled, but it was the one where the earth will be pretty much dead withing a small number of years unless our heroes leave in an alien-tech ship and find the solution. They have only their skills and a really big gun front-mounted in the nose of the ship that they can only fire in the most dire situations.
;-)
Oh sorry, that was Crusade
C'mon, Galen; tell them that we're looking for Trillana!
Would it be possible to write an abstraction layer above Qt/KDE and Gtk/GNOME that allowed you to replace the widget set at compile-time? If so, your installer could detect the libraries that you've got and install that version.
I know GNOME, and I know that just about everything is ultra-generic to the point that it makes sense to use XML descriptions of your user-interface and generate code in whatever language you like. We take the performance hit for internationalization. We take the performance hit for distributed objects. If we could just write an interface layer that abstracts the code the programmer has to write, then we'd be pretty much done, and I can't believe that the performance hit would compare to CORBA.
Let's see. You'd need some common way of getting at objects, but that's easy. Use CORBA, and write an ORBIt-KOM interface. Miguel was saying that this would be hard, but I think with enough programmer eyes on the task, it would be simple.
Of course, in the end this would be the death-knell of KDE. KDE relies on a widget set that's almost entirely free. GNOME relies on one that *is* free. If there's no real difference other than the licensing....
In the end, that's irrelevant. What matters is that people not feel hobbled so that they have to use extinct toolkits to write their apps (I'm sick of Netscape screwing up my theme, and I won't fall for it again).
Linux and *BSD are unusable as an OS design teaching tool due to their complexity.
I don't agree. You could certainly do a "whirlwind tour of the Linux kernel" in a semester and give the students assignments that involve augmenting the kernel in small ways that touch important subsystems. Yes, you can't get your head around the whole system, but that's the nature of large-scale programming, and working in that environment is practically a given in the industry.
The important part is that the students learn why an operating system works the way it does, and how the OS that they are studying has tackled certain problems. I would cover: VM, task/process management, kernel/user space issues in driver writing, writing a loadable module. I think this would give them a good head-start on the lay of the land. If I had time, or as extra credit, I would go into the interface between the core kernel and something like the IP stack. Either that or a few SMP features that would give them a sense of the depth of that abyss.
Now take the student who uses Minix as his/her example. There's no sense of how modern hardware issues are tackled. None of the complexity that stems from supporting every $0.25 ethernet card on the planet. How exactly is that supposed to introduce MODERN OS design. I mean, you could always teach them Tandy Color Computer BASIC/OS. The assembly exists in several well-annotated forms....
Interesting reading different people's comments. I especially love the ones that start: "They broke what made Star Trek great. What I really loved was..." The basic flaw in that line of reasoning is: a) most of you have no idea what it was about Star Trek that made it work for you, or you'd be writing your own series and b) what made it work for you is almost certainly the one thing that someone else hated.
e heads crowd happy.
Personally, I think it just doesn't matter. The basic problem that Star Trek has isn't acting or writing or special effects or evil execs. It's us.
The fans are really quite blood-thirsty at times, and that has to be difficult to deal with. Not in terms of ego, or hurt feelings, but in terms of audience. When the most vocal part of your audience decries everything you do, you can't assess what you should do.
This is why the series that succede in the fans eyes are always the new upstarts. They have no expectations, so the fans judge them, pretty much, at face value (as examples: Babylon 5, Farscape, X-Files). What Star Trek needs is to back off and wait a few years, but the franchise can't afford that, so they'll plow forward with what the focus groups tell them is hated least, and we'll get the blandest possible thing with a lot of things blowing up, and just enough skin to keep the football-aint-on-yet, channel-surfing, what-the-hell-is-this-shit-with-the-fucked-up-for
They could try being truely daring, and do something that will piss off half of the fans and galvenize the rest into advocates who will pull in the next generation of fans, but that won't fly with the shareholders.
In the end, Trek is dead because fans and studios can't work together. It's both of us that did it, but we'll keep pointing fingers until the last show spits up blood. Of course, then we'll just say: look what you did....