No that was 2.4.0test9-pre9. The current testing kernel is 2.4.0test9. You can't read anything in the version numbers though. It wouldn't surprise me if there was quite a long 2.4.0test series before a very short 2.4.0pre series.
I don't presume to know Linus' mind by my impression is that he went to the test series to try to push things along. I'm not convinced it really worked though - after all we did switch VMs during that time. Still it's looking good.
Check out this IIS setup and this TUX setup. The hardware is identical in each case with the exception of the SCSI controllers, and that shouldn't matter because in the IIS case the data set is small enough to fit in memory. The TUX setup is 2.5 times faster despite having a larger dataset.
They do: wap.domain.tld. wap belongs at the front, just like ftp and www. Google, for example, does this (wap.google.com).
What you seem to want is better WAP support from sites. That's a different issue and making wap a TLD won't help that issue as no one will force companies to register *.wap or to use it even if they do register it.
com.microsoft.www makes a lot more sense.
I really don't see why. Maybe it makes more sense to programmers and other people used to dealing with hierarchical namespaces in that order, but when most people go looking for MS I bet they think "I'm look for Microsoft, which is a company" (i.e. microsoft.com) and not "I'm looking for a company which is called Microsoft" (i.e. com.microsoft).
While the hierarchy for domain names may look like a file system to us most people don't make that connection. As for fitting with Java's name space, who gives a rat's arse?
The problem with this is consumer conscience. While I agree that there should be more TLDs, most internet users are doing good to remember a 2nd level domain with.com (possibly.org or.net).
Yep. One of things I like is that if I'm looking for company I can be fairly confident that the address is www.companyname.com. (I was actually asked this in a survey one: "How would you find Sony on the Web? I'd go to www.sony.com. Duh.) It's going to be bad enough having to guess between www.companyname.com, and www.companyname.biz, without throwing www.companyname into the mix. Oh well, there's always google.
You're argument is based on the assumption that you need DeCSS in this process - you don't. Up till now everyone has assumed that people will want to keep the video in MPEG2 format. After all why re-encode in something else when it'll only lower the quality? MPEG4 is the answer - because the result is 10-15 times smaller. If you're going to re-encode it then you don't need an MPEG2 stream so you don't need DeCSS - any decoder software will do. You just need a means of capturing the decoder's output.
The result is that the MPAA's assertion that the DeCSS case is about piracy is gone. It was never true anyway but now any fool can see it for what it is - a smokescreen. So this actually helps the DeCSS defendants.
The thing is that you have to decode the DVD to get to the MPEG2 stream. So you still need a decoder. But it does reduce the importance of DeCSS. Here's why:
The MPAA and DVDCCA have assumed that people will want to make perfect copies of DVDs. To do that you need to decode the DVD (using DeCSS, for example), and burn it onto something (i.e. you need a DVD burner and a big hard drive). MPEG4 invalidates that assumption because people are willing to trade a small drop in quality for a much more managable file size. Since you're re-encoding the video you don't need DeCSS - any DVD player will do because you can capture the data after the MPEG2 decoding (e.g. with a hacked video driver).
This really highlights a flaw in the CSS system - it only protects the MPEG2 signal. If you don't need a bit perfect MPEG2 copy it's worthless as a protection scheme.
Actually it probably helps the DeCSS defendants. You'll notice that DeCSS is conspiculously absent from the requirements to copy a DVD. Indeed, it's not even being done under Linux. Kinda shoots down the argument that DeCSS was a pirates tool doesn't it? The pirates have a far better tool - MPEG4.
DVDs store about 8GB of data for 100 minutes of movie with MPEG2. Using MPEG4 allows you to put the same movie into 700MB of data. Which means that every Joe with a CD burner can copy it (bit of an exageration - the hardware requirements for encoding are steep. Perhaps Intel finally have an application for the 1.5GHz P4?). Incase you haven't noticed CDRs and CD burners are vastly cheaper than writable DVDs and DVD burners. This is the MPAA's worst nightmare. It also has implications for the DVD marketplace. I for one will be thinking again about whether I need a DVD player.
But the technology mentioned isn't new - MPEG4 has been around for a while in the form of Divx:-). But this is the first review I've seen from a major site like Tom's.
Using a cookie system is a reasonable idea, provided it can be shut off. But it's a far greater privacy invasion than anything else around at the moment because they are able to build a database of a far larger proportion of your online activity than they could with other methods (such as the web banner cookies).
Also, it's always going to be very, very difficult for a search engine to anticipate what you want. It's better to explicitly tell it. To use your example, it would be better for you to just search for "big black albini" than to try to have the search engine guess that's what you wanted based on previous searches for "albini" and a current search for "big black". Otherwise the search will be less effective when you want something new. Using your example again, if you type "big black" when you really mean "big black boobs" then you don't want the search engine to assume that your previous search on "albini" has any relevance.
People have been trying for years to get natural language interfaces and other ways of being imprecise with computers working and so far the results are generally mediocre. Why do we want to try to rehash the same ideas in the context of the search engine?
Doesn't seem that way to me. These guys are from Hardin MD. The directory in the graph and mentioned most often in the text is MedWebPlus. The only time the mention themselves it is to say:
While the Hardin MD pages kept about the same average (10th)...
The problem is that they're trying to draw conclusions from the tracking of only 3 sites - their own (which hasn't changed), Yahoo's (which went up), and MedWebPlus (which went down). That's not statistically valid, and blaiming the decline of MedWebPlus' rankings on the Yahoo alliance is extremely dubious IMO.
I had a quick look myself using "hematology" as they did in the original survey. It looks a lot less concerning than they make out - there still seem to be plenty of directory sites. Perhaps they're suggesting that the MedWebPlus site has been singled out for negative treatment? Seems a bit paranoid.
One interesting result from my test is that the dmoz Open Directory page is ranked higher than Google's own copy. At least they don't seem to be favouring their own pages.
Finally, here's a tip for anyone concerned about this alleged favouring of Yahoo pages: exclude Yahoo from the search by using the advanced search options, or by including "-site:yahoo.com" in the query terms.
Yep. I just can't work out why these people are so stupid as to think they can shut it down. Are they so arrogant that they believe they can shove the genie back into the bottle? Did they learn nothing from the prohibition? Haven't they ever seen Star Wars? The more they try to stamp this sort of thing out the more people will object and rebel. The Motley fool article mentioned here a while back had it right - the music publishers are obsolete (in fact at this point they're leeches).
You've missed my point completely, well done! When IBM comitted to Monterey Caldera wasn't in the picture, and IBM was only just starting to look at Linux. Now Caldera aren't going to be interested in continuing the project as they are a Linux company so IBM has lost it's partner. IBM are also heavily into Linux (including the Trillian project to port Linux to IA-64) so probably aren't much interested in it either.
You really could see this comming. With Caldera buying bits of SCO IBM has lost their partner in Monterey and it makes little sense for them to create another UNIX by themselves. They've already got AIX and as partners in the Trillian project they'll have a Linux IA-64 solution as well. IBM's complete adoption of Linux may have killed this project regardless of SCO's sale, but I suspect that was the final nail in the coffin.
I think you're better off pointing out the verifiable inaccuracies in Moody's article. Don't sound like a Linux zealot, sound like a reader who's concerned about the facts. The source of the data is http://www.securityfocus.com/vdb/stats.ht ml. Moody has added the RedHat numbers to the Linux aggregate numbers (which already include the RedHat numbers).
It must almost border on libel. It looks like Moody has deliberately lied about the numbers. Here's the quote:
Windows NT totaled 99 new vulnerabilities on the BugTraq list. (So far in 2000, the count stands at 37.) This looks like an alarmingly high number in comparison with Solaris' 34 or NetBSD's 10, but it is significantly less than the 122 racked up by Red Hat and the other Linuxes (their 2000 count stands at 47).
The SecurityFocus stats page clearly shows RedHat's '99 vulnerabilities as 38 - less than 40% of WinNT's. So where did the 122 come from? Moody added RedHat's 38 to the Linux Aggregate of 84. He's done the same for this year's numbers (RedHat's count for this year is 17, and the total for Linux is 30 not 47). But the Linux Aggregate already includes the 38 RedHat vulnerabilities and it clearly states that in the preface on the page - Moody is either an incompetent researcher or he is deliberately counting vulnerabilities twice in order to discredit RedHat. I'd be consulting a lawyer about the possibility of a libel suit if I were them.
I notice both MacJunkie and AppleInsider have been told to remove the pictures by Apple's Legal Deparment. I wonder if the speed with which they've done this means someone in a significant position in Apple reads/.? I also wonder if it means those pictures are more accurate than we've been given them credit for? One thing that has occured to me is that the pictures may be a fake made by someone who has seen the real thing.
The power button (the 'third screw') would never be on the top of a machine.
Oh really? Ever seen an IBM Netfinity 3500? Power button on the top, right at the front on the right hand side. Just where you put your hand when ejecting a CD. And it's sensitive too. They had to issue a patch - a small plastic ring that sticks on and means you need to turn it on and off with a pen.
And before you say "Apple would never be so stupid" I'll say "hockey puck mouse".
Slashdot Mirror
No that was 2.4.0test9-pre9. The current testing kernel is 2.4.0test9. You can't read anything in the version numbers though. It wouldn't surprise me if there was quite a long 2.4.0test series before a very short 2.4.0pre series.
I don't presume to know Linus' mind by my impression is that he went to the test series to try to push things along. I'm not convinced it really worked though - after all we did switch VMs during that time. Still it's looking good.
Check out this IIS setup and this TUX setup. The hardware is identical in each case with the exception of the SCSI controllers, and that shouldn't matter because in the IIS case the data set is small enough to fit in memory. The TUX setup is 2.5 times faster despite having a larger dataset.
TPC-W isn't very useful right now as there has been exactly one result submitted.
WTF is wrong with wap.slashdot.org? WAP is a protocol - it belongs at the front.
What you seem to want is better WAP support from sites. That's a different issue and making wap a TLD won't help that issue as no one will force companies to register *.wap or to use it even if they do register it.
You forgot .beer.
While the hierarchy for domain names may look like a file system to us most people don't make that connection. As for fitting with Java's name space, who gives a rat's arse?
You're really comparing whatever.com to com/whatever. I don't see the benefits.
The result is that the MPAA's assertion that the DeCSS case is about piracy is gone. It was never true anyway but now any fool can see it for what it is - a smokescreen. So this actually helps the DeCSS defendants.
The MPAA and DVDCCA have assumed that people will want to make perfect copies of DVDs. To do that you need to decode the DVD (using DeCSS, for example), and burn it onto something (i.e. you need a DVD burner and a big hard drive). MPEG4 invalidates that assumption because people are willing to trade a small drop in quality for a much more managable file size. Since you're re-encoding the video you don't need DeCSS - any DVD player will do because you can capture the data after the MPEG2 decoding (e.g. with a hacked video driver).
This really highlights a flaw in the CSS system - it only protects the MPEG2 signal. If you don't need a bit perfect MPEG2 copy it's worthless as a protection scheme.
Actually it probably helps the DeCSS defendants. You'll notice that DeCSS is conspiculously absent from the requirements to copy a DVD. Indeed, it's not even being done under Linux. Kinda shoots down the argument that DeCSS was a pirates tool doesn't it? The pirates have a far better tool - MPEG4.
But the technology mentioned isn't new - MPEG4 has been around for a while in the form of Divx:-). But this is the first review I've seen from a major site like Tom's.
Also, it's always going to be very, very difficult for a search engine to anticipate what you want. It's better to explicitly tell it. To use your example, it would be better for you to just search for "big black albini" than to try to have the search engine guess that's what you wanted based on previous searches for "albini" and a current search for "big black". Otherwise the search will be less effective when you want something new. Using your example again, if you type "big black" when you really mean "big black boobs" then you don't want the search engine to assume that your previous search on "albini" has any relevance.
People have been trying for years to get natural language interfaces and other ways of being imprecise with computers working and so far the results are generally mediocre. Why do we want to try to rehash the same ideas in the context of the search engine?
I had a quick look myself using "hematology" as they did in the original survey. It looks a lot less concerning than they make out - there still seem to be plenty of directory sites. Perhaps they're suggesting that the MedWebPlus site has been singled out for negative treatment? Seems a bit paranoid.
One interesting result from my test is that the dmoz Open Directory page is ranked higher than Google's own copy. At least they don't seem to be favouring their own pages.
Finally, here's a tip for anyone concerned about this alleged favouring of Yahoo pages: exclude Yahoo from the search by using the advanced search options, or by including "-site:yahoo.com" in the query terms.
Actually a law suit is not necessary. It only requires someone to file an application for reexamination pointing out the (documented) prior art.
Yep. I just can't work out why these people are so stupid as to think they can shut it down. Are they so arrogant that they believe they can shove the genie back into the bottle? Did they learn nothing from the prohibition? Haven't they ever seen Star Wars? The more they try to stamp this sort of thing out the more people will object and rebel. The Motley fool article mentioned here a while back had it right - the music publishers are obsolete (in fact at this point they're leeches).
You've missed my point completely, well done! When IBM comitted to Monterey Caldera wasn't in the picture, and IBM was only just starting to look at Linux. Now Caldera aren't going to be interested in continuing the project as they are a Linux company so IBM has lost it's partner. IBM are also heavily into Linux (including the Trillian project to port Linux to IA-64) so probably aren't much interested in it either.
You really could see this comming. With Caldera buying bits of SCO IBM has lost their partner in Monterey and it makes little sense for them to create another UNIX by themselves. They've already got AIX and as partners in the Trillian project they'll have a Linux IA-64 solution as well. IBM's complete adoption of Linux may have killed this project regardless of SCO's sale, but I suspect that was the final nail in the coffin.
I think you're better off pointing out the verifiable inaccuracies in Moody's article. Don't sound like a Linux zealot, sound like a reader who's concerned about the facts.
The source of the data is http://www.securityfocus.com/vdb/stats.ht ml. Moody has added the RedHat numbers to the Linux aggregate numbers (which already include the RedHat numbers).
The SecurityFocus stats page clearly shows RedHat's '99 vulnerabilities as 38 - less than 40% of WinNT's.
So where did the 122 come from? Moody added RedHat's 38 to the Linux Aggregate of 84. He's done the same for this year's numbers (RedHat's count for this year is 17, and the total for Linux is 30 not 47). But the Linux Aggregate already includes the 38 RedHat vulnerabilities and it clearly states that in the preface on the page - Moody is either an incompetent researcher or he is deliberately counting vulnerabilities twice in order to discredit RedHat. I'd be consulting a lawyer about the possibility of a libel suit if I were them.
I notice both MacJunkie and AppleInsider have been told to remove the pictures by Apple's Legal Deparment. I wonder if the speed with which they've done this means someone in a significant position in Apple reads /.? I also wonder if it means those pictures are more accurate than we've been given them credit for? One thing that has occured to me is that the pictures may be a fake made by someone who has seen the real thing.
And before you say "Apple would never be so stupid" I'll say "hockey puck mouse".