When I saw that this had been posted on slashdot, I thought "Ah! So MIT has finally come through and put lecture notes on the Web". But no, it all still looks more or less as it the last time I checked, several weeks ago.
Doubtless some of the courses have enough information available to make them useful, but the status of the courses that I would be interested in (mostly math/physics) is still essentially useless. To take a typical example, Quantum Physics II, which would be a good refresher for me, has only "Syllabus, Calendar, Assignments".
So high marks go to MIT for pioneering the concept, but not even a passing grade for execution if they aren't going to do the job properly. I guess I have to say that I find it a little bothering that I have seen lots of praise for MIT for doing this, but no criticism of the amount of information that they're actually putting online. I do recognise that to do the job properly is a daunting task, and maybe the lack of content is merely becuase it's such early days, but I wish I could feel more confident that a lot of the "courses" aren't simply going to consist of a shell with a few web pages of minimal content.
I always hear the Daleks refered to like that - watching behind the couch. Has anyone here ever been scared of the daleks?
Not only did I indeed watch them from behind the sofa, but I also had nightmares about them.
FWIW, this was during their very first appearance. Those early years were pretty terrifying to young kids. There certainly wasn't anything else remotely as scary as Dr. Who broadcast during children's TV. (Which I suppose invites responders to reply with comments along the lines of: "Oh yes there was, Zebedee from The Magic Roundabout was just as scary, and what about John Noakes trying to make something -- anything -- on Blue Peter?".)
I don't suppose that anyone is really interested, but this does give me the opportunity to say that my novels have all been written using absolutely the best tools for the job: a real editor (in my case, mostly VEDIT Plus under Windows, although I also did one with emacs) and Plain Tex. Yes, that's Plain Tex, not LaTeX.
I remain firmly convinced that the combination of a powerful editor and Plain TeX cannot be beat.
The problem, though, is that nowadays publishers more and more demand manuscripts in the form of M$ Word files, which frankly sucks. I am measurably less productive under Word than I am with the combination of (editor + Plain TeX), and I suspect that the same would be true of most authors who are technically competent.
I expect that within the next year or so, most ISPs will simply block all the P2P stuff to avoid the legal hassles.
And I expect that "within the next year or so" plus a couple of months, P2P protocols will be designed to be indistinguishable from VPN telecommuting. Its easy to see how ISPs can win the current P2P battles; it's less clear how they can hope to win the war.
This has nothing to do with compilers. This is a a function of the linker, which does indeed do exactly what you want: it makes sure that only the functions you actually use (either directly or indirectly) are inserted into the executable. So, in your example, strcat() will be loaded, as will the other string functions that it relies on (and the ones that those rely on, etc.). But a string function that does not get called by strcat() or its descendents will not be put into the executable. And, in particular, all these functions get loaded only once.
In other words, this all works exactly the way that you say that it should work.
At least, that's way classical edit-compile-link-load-run languages work. Now once you start to talk about languages and implementation that subvert this process (often in the name of some sort of efficiency), then all bets are off and you can easily end up with the archetypal 11MB Java "Hello World" program.
All computers will come equipped from the manufacturer with adware installed.
Some of them already do. I needed a new XP machine recently, and, because of their sort-of support of Linux, I decided to buy a cheap HP. When I got it home from the store and turned it on, I was horrified at the amount of preloaded adware, spyware and utter junk. The realization that a non-geek would be stuck with this stuff made me sick. Even a reasonably geeky person like me is still figuring out what can be safely deleted more than a week after installing the new machine.
I suppose that we'll end up paying a premium for boxes that don't have all that garbage installed on them. (After all, I imagine that they pay HP to put their junk on there, so they are really subsidising my purchase.)
Neither MCI or any other carrier is routing their calls via "the Internet".
Some of them are. I was at a presentation at Spring VON (Voice On the Net) conference where an international telephony company described exactly how they do indeed use the public Internet for routing their VoIP traffic.
Of course, this is not to say that MCI is doing this. But it certainly isn't true that no one is doing it.
And on a complete tangent, it is rather ironic that this story was posted while I was (and still am) on a conference call with an FBI representative dealing with CALEA issues related to VoIP. It's worth remembering that while most of us don't like regulation, the fact is that telcos are required to provide certain features (such as 911, E911 and wiretap support) for which it is not at all obvious whether they apply to VoIP. I for one am grateful that some of those regulations are in place, and would be somewhat concerned to be in an all-VoIP world where I could dial 911 and not be guaranteed that the service provider would do its darndest to route that call to the local emergency dispatcher.
But I digress....
Re:Sounds of the plasma wind
on
Goodbye, Galileo
·
· Score: 3, Interesting
Now that the audio is no longer slashdotted, I can hear this particular example and agree that this is the wideband audio downconverted. Almost all the audio recordings from Voyager were generated from the instrument operating in channelized mode, although they did do a few in this same wideband mode.
A similar technique used at Earth would produce very similar results, and would not need to be downconverted, because of the weaker field here.
At one time there was a very cool audio of ring-plane crossing from Voyager 2 at Saturn (from the Planetary Radio Astronomy experiment, which was sort of the higher-frequency brother of the Plasma Wave experiment), but I doubt that that is available any more.
Anyway, probably most slashdotters agree that all of these spacecraft have done some pretty cool things.
Next up... Cassini
Sounds of the plasma wind
on
Goodbye, Galileo
·
· Score: 5, Interesting
listen to Ganymede's eerie sounding plasma wind
The reason that it sounds so "eerie" is because it is recorded with a receiver whose channels are harmonically related. A true wideband recording would sound quite different. This is true of the similar Voyager plasma recordings as well.
It is indeed good news that one of the bigger vendors is putting a Linux box in their catalog... However, just giving a blank box and a CD set is not the way it should be done.
And putting it in a catalogue is hardly the same thing as actually allowing the box to be sold at retail stores. As far as I'm concerned, Linux won't have started to make meaningful inroads into the desktop market until I know that I can tell a friend to walk into CompUSA and be I can be reasonably certain that once inside he will be able to buy a functional Linux box.
What HP is doing now is simply PR. I'm very disappointed in them after the optimism caused by the initial announcement. Imagine if HP sold their Windows boxes only via mail order and without the OS installed. They wouldn't sell very many, would they?
And it would be much better if the notice was at the top, on a bright red background, like a MEGA-SPONSORED link, so that noone could miss it.
Hear, hear! Google is so good at its job that I usually only ever look at the first two or three links, and it's very rare that I actually have to scroll down to the bottom of the page, especially since by default I use the advanced search and list 100 links per page.
I just did the experiment of searching for "kazaa lite", and I can guarantee that I would never have seen that notice hiding at the bottom of the 100 links on the first page if this slashdot story hadn't been posted.
Worse, I am hardly likely to spot any similar notices when I do other searches, since I'm not likely to take the trouble to page down to the bottom on the offchance that there's a notice there. Ever since its inception, google has seemd to do everything right. But this seems to be clearly wrong. If they really want to remove the links, I concur that they should put the fact that they've tampered with the results in some prominent place where even bozos like me are guaranteed to see it.
How did you file a complaint? The page for filing complaints doesn't even do anything, it just says that telemarketers must stop calling you on October 1st and that complaints may begin being filed at that time.
Maybe you missed the fact that this was someone who violated the Colorado no-call list? So I filed a complaint (well, I tried to file a compaint) at www.coloradonocall.com.
I am on the Colorado list. Not long ago I got a call and, since I had nothing much better to do, talked to the telemarketer nicely to extract as much info as possible (since it's my experience that as soon as you take a combative approach, they hang up before you've got enough information to report them).
I then went to the Web site to report them. The Web site makes it clear that the whole do-not-call system only works properly if violators are reported. So I went through a few pages of filling out forms with all the tedious details of the call. Then I hit the "submit" button and get a "your submission could not be processed" error.
OK, thinks I. This is because the morons expect me to be using IE. So I went through it all again using IE instead of Firebird. Same thing.
So I send them an e-mail at the mailto address, telling them that I wanted to report a violation and was unable to do so because the web site repeatedly gave me an error when trying to process the information.
I never heard anything from them.
I'm not sure what to conclude from this story. But I ended up being even more ticked off at the state government than I was at the telemarketer. And that's a pretty high threshold to reach.
I sure hope that the national list has a more effective mechanism for reporting offenders.
Changelog is here: http://www3.mandrakelinux.com/en/92beta.php3
That's the changelog for beta2. The story is about RC1.
Re:Truly P2P if SOBIG.G contains the spam message
on
P2P Spam?
·
· Score: 1
The best we can hope for is that ISPs get smart and start blocking SMTP ports on all ip addresses not registered as SMTP servers.
Er... the spam would be sent from ordinary SMTP clients, not servers. Requiring SMTP servers to be registered and blocking port 25 on non-registered boxes, even completely apart from the control-freak aspect of such a worrying suggestion, doesn't seem to me like it would do any good at all.
Maybe if people would get it through their skulls that Windows ships with a BIG WINDOWS UPDATE LINK in the Start Menu for a REASON, and maybe if people would at least check for new, fun things weekly, these viruses wouldn't spread quite so far.
Well, yes. If Windows Update works. Which in my case it doesn't. When I go to the Update site I get a nice little error ("Windows Update has encountered an error and cannot display the requested page."), and no clue as to what to do next. Reporting the problem elicited the usual response: silence. (Yes, I did try their "online support", which sent me in the usual tortuous circles.)
I really was prepared to be impressed with Windows update, since several posts here have extolled its virtues. But it turned out -- at least for me -- to be another disappointment from Microsoft.
By far, hunting down layer after layer of dependency while trying to install software, only to meet conflicts is my biggest problem.
Followed by:
1. the inability to cleanly uninstall software that has been installed with the usual configure/make/make install incantation
2. The lack of something equivalent to the Windows GoBack program (i.e., screw up and put your disk back exactly the way it was ten minutes ago)
3. Drivers that almost but don't quite work properly
4. The current kernel is non-preemptive (at least this annoyance is going away)
Mind you, the list of Windows annoyances would be much longer.
Hmmm... thinks... this being slashdot, the problem with the original question is that anyone who responds with any annoyances will probably get modded down:-)
Taylor says he plans to focus on (and fund) studies that 'will highlight Microsoft's advantages in areas such as security, feature-completeness and total cost of ownership.'"
Perhaps I am old-fashioned (except that I read/.) but I thought that the point of a "study" was to learn something in an objective manner, rather than to find rationalisations that support a pre-determined position.
I have not heard of karamba, but will take a look. Powerpro is an awful lot more than panels/vdesks/shortcuts, though. Just *some* of the extra stuff it does: it builds dynamic menus, can send arbitrary messages to arbitrary windows at arbitrary times (and when particular events happen), expands keystrokes and is thoroughly scriptable. For any given feature, there are doubtless other Windows programs that work as well (but probably not better), but to get the combination in a single, solid package is pretty amazing; and if all the functions can be duplicated under Linux, I have certainly not stumbled across the right program(s).
(Actually, I've never even been able to figure out how to do some of the really simple stuff in Linux that PowerPro does, like having Win+num-pad switch desktops for me. But I digress.)
And I have heard the argument more than once that GoBack can be simulated with CVS and cron jobs, or similar trickery. I am certain that it can't -- although, as you say, that sort of arrangement can probably duplicate XP Restore, which is certainly better than nothing. GoBack is one of those strange programs that, unless you've used it, you really don't recognize how fantastic it is to have every single write-to-disk, no matter what caused it, journalled and undoable.
Anyway, thanks for the heads-up about karamba; I'll take a look at it to see how much of PowerPro it replaces. It's been a while since I tried PowerPro under wine, so maybe it works better under a more recent release of wine than it has in the past; so perhaps PowerPro/wine is a reasonable combination these days.
Oh, an interesting fact about it is if you take a holographic film and cut it in half, because all the information about the image is stored throughout the film, you don't have half a hologram; you have a hologram of the entire object that is half the size of the original. Pretty cool stuff actually.
Just to elaborate: it's also half the resolution of the original. I am sure that you knew that, but the casual reader of your post may wonder what one loses by cutting the hologram in half.
My experience with the Bayesian filtering based Spambayes is extremely good. It is very transparent so that you can see how it is classifying.
Or try popmail. Or Mozilla Mail. Or almost anything else with bayesian filtering. I had stability problems with spambayes so I use popmail on some systems and Moz mail on others. Popmail in particular verges in the magical (for me, anyway).
The real message here, though, is: if you see more than about one spam per week, try bayesian filtering.
I can't live without virtual desktops... The poor man's multi-monitor setup. Barring extensions like litestep, Windows has never done this.
I don't know exactly what you mean by extension as opposed to program, but PowerPro (www.windowspowerpro.com) includes vdesks among its uncountable number of great features. In fact, I find the PowerPro implementation to be considerably superior to, and more powerful than, the KDE version.
In my irrelevant opinion, the two Big Missing Applications in Linux are PowerPro and GoBack. But I expect that I am the only person in the world with that opinion.
Sigh; having praised a couple of Windows-only programs, I suppose I'll be modded down now. So in an attempt to pre-empt that possibility, please note that it's the applications I like, not the OS.
just how many of you out there are actually using a machine that came with Windows, but you never used that copy?
Me, for one.
I have an HP box here still with the original XP disk inside it, but disconnected. The disk that is connected is running Mandrake. The XP on the original disk has never been booted.
I suppose this means that I should try to get a refund. But, frankly, I don't have the stomach for all the pain that would put me through. Having taken a case (unrelated to computing) to the U S Court of Appeals and lost there (after winning at the district level), I am afraid that I have had more than enough of courts for one lifetime.
If it is "true" that Linux has copyrighted material from SCO, could they remove that material or rewrite it so that 2.6.x wouldn't be affected by the licensing system SCO is imposing?
Sure, this is one of several possible things that could happen. But since SCO refuses to tell us exactly what code is infringing on their copyright, it's kind of hard for the developers to know what code they might want to change:-) I think that's the whole point here: SCO keeps saying, "You owe us money but we won't tell you why".
Slashdot Mirror
Doubtless some of the courses have enough information available to make them useful, but the status of the courses that I would be interested in (mostly math/physics) is still essentially useless. To take a typical example, Quantum Physics II, which would be a good refresher for me, has only "Syllabus, Calendar, Assignments".
So high marks go to MIT for pioneering the concept, but not even a passing grade for execution if they aren't going to do the job properly. I guess I have to say that I find it a little bothering that I have seen lots of praise for MIT for doing this, but no criticism of the amount of information that they're actually putting online. I do recognise that to do the job properly is a daunting task, and maybe the lack of content is merely becuase it's such early days, but I wish I could feel more confident that a lot of the "courses" aren't simply going to consist of a shell with a few web pages of minimal content.
Not only did I indeed watch them from behind the sofa, but I also had nightmares about them.
FWIW, this was during their very first appearance. Those early years were pretty terrifying to young kids. There certainly wasn't anything else remotely as scary as Dr. Who broadcast during children's TV. (Which I suppose invites responders to reply with comments along the lines of: "Oh yes there was, Zebedee from The Magic Roundabout was just as scary, and what about John Noakes trying to make something -- anything -- on Blue Peter?".)
I remain firmly convinced that the combination of a powerful editor and Plain TeX cannot be beat.
The problem, though, is that nowadays publishers more and more demand manuscripts in the form of M$ Word files, which frankly sucks. I am measurably less productive under Word than I am with the combination of (editor + Plain TeX), and I suspect that the same would be true of most authors who are technically competent.
And I expect that "within the next year or so" plus a couple of months, P2P protocols will be designed to be indistinguishable from VPN telecommuting. Its easy to see how ISPs can win the current P2P battles; it's less clear how they can hope to win the war.
In other words, this all works exactly the way that you say that it should work.
At least, that's way classical edit-compile-link-load-run languages work. Now once you start to talk about languages and implementation that subvert this process (often in the name of some sort of efficiency), then all bets are off and you can easily end up with the archetypal 11MB Java "Hello World" program.
Some of them already do. I needed a new XP machine recently, and, because of their sort-of support of Linux, I decided to buy a cheap HP. When I got it home from the store and turned it on, I was horrified at the amount of preloaded adware, spyware and utter junk. The realization that a non-geek would be stuck with this stuff made me sick. Even a reasonably geeky person like me is still figuring out what can be safely deleted more than a week after installing the new machine.
I suppose that we'll end up paying a premium for boxes that don't have all that garbage installed on them. (After all, I imagine that they pay HP to put their junk on there, so they are really subsidising my purchase.)
Some of them are. I was at a presentation at Spring VON (Voice On the Net) conference where an international telephony company described exactly how they do indeed use the public Internet for routing their VoIP traffic.
Of course, this is not to say that MCI is doing this. But it certainly isn't true that no one is doing it.
And on a complete tangent, it is rather ironic that this story was posted while I was (and still am) on a conference call with an FBI representative dealing with CALEA issues related to VoIP. It's worth remembering that while most of us don't like regulation, the fact is that telcos are required to provide certain features (such as 911, E911 and wiretap support) for which it is not at all obvious whether they apply to VoIP. I for one am grateful that some of those regulations are in place, and would be somewhat concerned to be in an all-VoIP world where I could dial 911 and not be guaranteed that the service provider would do its darndest to route that call to the local emergency dispatcher.
But I digress....
A similar technique used at Earth would produce very similar results, and would not need to be downconverted, because of the weaker field here.
At one time there was a very cool audio of ring-plane crossing from Voyager 2 at Saturn (from the Planetary Radio Astronomy experiment, which was sort of the higher-frequency brother of the Plasma Wave experiment), but I doubt that that is available any more.
Anyway, probably most slashdotters agree that all of these spacecraft have done some pretty cool things.
Next up... Cassini
The reason that it sounds so "eerie" is because it is recorded with a receiver whose channels are harmonically related. A true wideband recording would sound quite different. This is true of the similar Voyager plasma recordings as well.
And putting it in a catalogue is hardly the same thing as actually allowing the box to be sold at retail stores. As far as I'm concerned, Linux won't have started to make meaningful inroads into the desktop market until I know that I can tell a friend to walk into CompUSA and be I can be reasonably certain that once inside he will be able to buy a functional Linux box.
What HP is doing now is simply PR. I'm very disappointed in them after the optimism caused by the initial announcement. Imagine if HP sold their Windows boxes only via mail order and without the OS installed. They wouldn't sell very many, would they?
Hear, hear! Google is so good at its job that I usually only ever look at the first two or three links, and it's very rare that I actually have to scroll down to the bottom of the page, especially since by default I use the advanced search and list 100 links per page.
I just did the experiment of searching for "kazaa lite", and I can guarantee that I would never have seen that notice hiding at the bottom of the 100 links on the first page if this slashdot story hadn't been posted.
Worse, I am hardly likely to spot any similar notices when I do other searches, since I'm not likely to take the trouble to page down to the bottom on the offchance that there's a notice there. Ever since its inception, google has seemd to do everything right. But this seems to be clearly wrong. If they really want to remove the links, I concur that they should put the fact that they've tampered with the results in some prominent place where even bozos like me are guaranteed to see it.
Maybe you missed the fact that this was someone who violated the Colorado no-call list? So I filed a complaint (well, I tried to file a compaint) at www.coloradonocall.com.
I then went to the Web site to report them. The Web site makes it clear that the whole do-not-call system only works properly if violators are reported. So I went through a few pages of filling out forms with all the tedious details of the call. Then I hit the "submit" button and get a "your submission could not be processed" error.
OK, thinks I. This is because the morons expect me to be using IE. So I went through it all again using IE instead of Firebird. Same thing.
So I send them an e-mail at the mailto address, telling them that I wanted to report a violation and was unable to do so because the web site repeatedly gave me an error when trying to process the information.
I never heard anything from them.
I'm not sure what to conclude from this story. But I ended up being even more ticked off at the state government than I was at the telemarketer. And that's a pretty high threshold to reach.
I sure hope that the national list has a more effective mechanism for reporting offenders.
That's the changelog for beta2. The story is about RC1.
Er... the spam would be sent from ordinary SMTP clients, not servers. Requiring SMTP servers to be registered and blocking port 25 on non-registered boxes, even completely apart from the control-freak aspect of such a worrying suggestion, doesn't seem to me like it would do any good at all.
Well, yes. If Windows Update works. Which in my case it doesn't. When I go to the Update site I get a nice little error ("Windows Update has encountered an error and cannot display the requested page."), and no clue as to what to do next. Reporting the problem elicited the usual response: silence. (Yes, I did try their "online support", which sent me in the usual tortuous circles.)
I really was prepared to be impressed with Windows update, since several posts here have extolled its virtues. But it turned out -- at least for me -- to be another disappointment from Microsoft.
Followed by:
1. the inability to cleanly uninstall software that has been installed with the usual configure/make/make install incantation
2. The lack of something equivalent to the Windows GoBack program (i.e., screw up and put your disk back exactly the way it was ten minutes ago)
3. Drivers that almost but don't quite work properly
4. The current kernel is non-preemptive (at least this annoyance is going away)
Mind you, the list of Windows annoyances would be much longer.
Hmmm... thinks... this being slashdot, the problem with the original question is that anyone who responds with any annoyances will probably get modded down :-)
Perhaps I am old-fashioned (except that I read /.) but I thought that the point of a "study" was to learn something in an objective manner, rather than to find rationalisations that support a pre-determined position.
I have not heard of karamba, but will take a look. Powerpro is an awful lot more than panels/vdesks/shortcuts, though. Just *some* of the extra stuff it does: it builds dynamic menus, can send arbitrary messages to arbitrary windows at arbitrary times (and when particular events happen), expands keystrokes and is thoroughly scriptable. For any given feature, there are doubtless other Windows programs that work as well (but probably not better), but to get the combination in a single, solid package is pretty amazing; and if all the functions can be duplicated under Linux, I have certainly not stumbled across the right program(s).
(Actually, I've never even been able to figure out how to do some of the really simple stuff in Linux that PowerPro does, like having Win+num-pad switch desktops for me. But I digress.)
And I have heard the argument more than once that GoBack can be simulated with CVS and cron jobs, or similar trickery. I am certain that it can't -- although, as you say, that sort of arrangement can probably duplicate XP Restore, which is certainly better than nothing. GoBack is one of those strange programs that, unless you've used it, you really don't recognize how fantastic it is to have every single write-to-disk, no matter what caused it, journalled and undoable.
Anyway, thanks for the heads-up about karamba; I'll take a look at it to see how much of PowerPro it replaces. It's been a while since I tried PowerPro under wine, so maybe it works better under a more recent release of wine than it has in the past; so perhaps PowerPro/wine is a reasonable combination these days.
Just to elaborate: it's also half the resolution of the original. I am sure that you knew that, but the casual reader of your post may wonder what one loses by cutting the hologram in half.
Or try popmail. Or Mozilla Mail. Or almost anything else with bayesian filtering. I had stability problems with spambayes so I use popmail on some systems and Moz mail on others. Popmail in particular verges in the magical (for me, anyway). The real message here, though, is: if you see more than about one spam per week, try bayesian filtering.
I don't know exactly what you mean by extension as opposed to program, but PowerPro (www.windowspowerpro.com) includes vdesks among its uncountable number of great features. In fact, I find the PowerPro implementation to be considerably superior to, and more powerful than, the KDE version.
In my irrelevant opinion, the two Big Missing Applications in Linux are PowerPro and GoBack. But I expect that I am the only person in the world with that opinion.
Sigh; having praised a couple of Windows-only programs, I suppose I'll be modded down now. So in an attempt to pre-empt that possibility, please note that it's the applications I like, not the OS.
Me, for one.
I have an HP box here still with the original XP disk inside it, but disconnected. The disk that is connected is running Mandrake. The XP on the original disk has never been booted.
I suppose this means that I should try to get a refund. But, frankly, I don't have the stomach for all the pain that would put me through. Having taken a case (unrelated to computing) to the U S Court of Appeals and lost there (after winning at the district level), I am afraid that I have had more than enough of courts for one lifetime.
I somehow missed making a connection. This is "News for Nerds" or "Stuff that Matters" exactly how again?
Sure, this is one of several possible things that could happen. But since SCO refuses to tell us exactly what code is infringing on their copyright, it's kind of hard for the developers to know what code they might want to change :-) I think that's the whole point here: SCO keeps saying, "You owe us money but we won't tell you why".