Just to follow up on it, I read through the thread and found that Foxconn linked to a page on Microsoft's site which supposedly explains ACPI compliance. Interestingly enough, that page refused to display on anything but IE. *Sigh*
Isn't it, though? What does it really mean to be ACPI compliant? I would imagine that it might just mean that the BIOS contains well-formed tables, and in that case this motherboard probably is ACPI compliant. Either way, I don't really think ACPI compliance is specified to mean that the board must work with all operating systems.
Especially considering how ACPI doesn't really seem to be the greatest standard so far conceived, and on top of that is an Intel-Microsoft cooperative product, I just wouldn't bet my life and limb on its specification being sound.
HP has given me boxes that size for 4 screws in a plastic bag, wrapped in foam. Repeatedly.
If they do this so regularly as your comment and many others seem to suggest, I just cannot help but wonder: How do things go so wrong to begin with? Also, one would think that the errors of their ways ought to be completely obvious to anyone involved, so whyever do they not fix it?
It's a stupid question to begin with. The question itself seems to assume that all of human civilization will fail at some point and that we will have to go trough a new dark age. I find myself very skeptical to that. We don't have to worry about warning people 10,000 years from now -- that's for the people 9,950 years from now to think of. We, on the other hand, will just have to think of warning people 50 years from now.
There's no obvious reason to assume that whatever government the future holds will just magically forget the nuclear dumps and stop upgrading them.
All this is of course dependent on us not going through some technological singularity or anything like it under 10,000 years. What's even to say that "humans" 10,000 years from now will even have organic bodies that would be damaged by radioactivity?
Of course, Splashtop is extremely fast, but depending on what exactly he wants to do with these installation, a stock Debian Etch installation might function just as well. It usually boots in 20-30 seconds, which is certainly fast compared to many others, but an order of magnitude slower than Splashtop. So, it really depends on just how fast he wants them to boot.
Even if it was probably meant that way by the submitter, you don't necessarily have to interpret that sentence in the way that Windows cause headaches for its users, but rather for every one else.
See, I haven't used Windows myself for at least five years now, and yet it still gives me headaches. Disregarding headaches caused by peoples' dysfunctional Windows PCs that I have to fix every now and then, the greatest source of headaches is that people disregard that there are other systems. Such as hardware manufacturers, web designers, people sending you MSO files and assume you can read them, and so on. I don't think I have to create a complete list for you to get the picture.
Therefore, Apple climbing ever higher on the usage charts is occasion to celebrate indeed. Any dent that can be made to the monoculture is a good dent.
It is a corallary, because a large part of Linux's challenges to adoption is due not to Windows itself, but rather the monoculture that is the computing environment of today. In other words, the fact that the monoculture is made up Windows is irrelevant; the relevant part is that there is a monoculture. If OSX usage rises high enough, that monoculture will dissipate, and other operating systems, including but not necessarily limited to Linux, will also have an easier way in.
A lot of people here on Slashdot care about the adoption of Linux, so it is relevant to the discussion.
Speaking of which - how would WinFS and ZFS compare?
Much like chalk and cheese. WinFS is(/was) *not* a filesystem, it's a database/metadata layer that sits between the filesystem (NTFS) and the applications.
Indeed. I find it well acceptable that a random slashdotter wouldn't know this, but it disturbs me all the more that the author of TFA says the same in a authoritative voice.
In fact, apart from a few obvious points (like Windows needing better CD burning tools, or that it might not need to be released in umpteen editions), all his points are more or less completely bogus:
20. Modularized OS
Yes, that's easy to say, in about the same way that it's easy to say that "Windows 7 should be better than Vista". He does not appear to have any clear idea in what manner Windows should be modularized or even why. In fact, Windows is probably far more modular than he actually realizes -- it's just that Microsoft only releases complete builds of it.
19. XP Virtual Machine
I don't even know where to start on this. Lots of people seem to think that all Microsoft would have to do is throw an instance of XP into a VM and everything would be fine. It doesn't take much to realize the stupidity of that idea. To begin with, it is obvious that they can't make the instance of XP completely virtualized, because then it would have access to none of the actual hardware, which would result in large outcries of people like gamers or anyone who expects to use any piece of userspace software that comes with a piece of hardware they bought (like synchronization software for their PDA:s or whatever).
So, they'd need to allow API calls through to the real operating system running on the real hardware. Once they've wrapped everything that needs to be wrapped for any real level of hardware compatibility, they're back to square 1, only with a useless CPU virtualizer in between.
18. New UAC In theory UAC was a great idea.
UAC was never a great idea, neither in practice nor in theory. All it can do is tell the user that some program needs more privileges, but it can't actually know why, and neither can the user. It just gives a false sense of security and is probably designed to push the blame onto the users and everything which has already been discussed to death here on Slashdot.
17. Gaming Mode
I would much like to know what services he thinks are detrimental to gaming performance. I'd be very surprised if he'd find even one which would increase gaming performance, by any metric, by even a percent when turned off. The indexer service in Windows Vista might be an exception, but even then, I don't see a reason for an entire "Gaming Mode" just to turn it off.
13. WinFS
Yes, here it was; the author showing off his perfect cluelessness by claiming that WinFS would be an "NTFS replacement". He also makes the completely unbased claim that, and I quote, "the relational database structure should enhance overall system performance". However a relational database would actually be faster than a filesystem is completely beyond me.
9. Program Caching
He's actually trying to blame users' (understandable) ignorance about virtual memory and block device caching on Microsoft. While I might agree that Windows' Program Manager may be a bit too dense about what it actually displays, it's not as if there even exists a metric that Microsoft could choose to display that would make users content.
6. Barebones Kernel
This idea has been thrown around by Microsoft, specifically âMinWinâ(TM). Allowing the user to choose between this and the default kernel could potentially allow older systems (i.e. XP based) to run the new OS with decent performance levels.
SPF is great. It isn't a total solution, and there are negatives, but it certainly is better than the anyone is anyone free for all.
Actually, SPF sucks really badly, and if you've turned it on, you should turn it off immediately. The reason SPF sucks so bad is that it unconditionally and helplessly breaks all forms of forwarding and mailing lists. As such, it is indeed worse than anyone is anyone free for all.
Instead, you should support DKIM, which solves the same problem without those bugs.
I have no idea what you're talking about here. DKIM breaks neither mailing lists nor forwarding, and mailing lists are not supposed to strip the DKIM header (if they did, that would get the mail rejected for sure).
DKIM ensures the integrity of the message and that it, at some point, passed through the MTA for the domain in the "From" header. It does that by signing the body of the message along with a set of headers. As long as the mailing list is RFC 2821/2822 compliant and only adds new headers to the top of the message (only headers below the DKIM header are part of the signature), its integrity under DKIM is unaffected.
Fedora has had that feature for a while now (since 7 or 8, I think). They call it "spins" of Fedora, where you can create your own install or live CD with the package selection you want.
I can't say I can tell you right off the bat how to do the same thing with Ubuntu, but given that it's Debian-based, I can't imagine it being terribly hard.
Speaking of which, though; does anyone know what the memory column in Windows' task manager really means? That is, in terms of what it measures. Is it resident pages? Private pages exclusively? The process' virtual size?
I know that Java programs usually show up as taking >100 MB in Windows's task manager, but in Linux/Unix, top makes it clear that that's just the virtual size, and that the actual resident size is usually closer to 20 MBs (it seems that Sun's Java runtime maps a huge heap with lazy page allocation), which leads me to believe that the task manager might actually be reporting the virtual size, which, while interesting in its own right, doesn't really tell very much about the actual memory usage of a process.
Yes, one you don't have to download, and one you do.
On Linux: Basket Notepads
I was thinking cat on Linux. You know, like:
$ cat >>~/notes Contract number: 14732871-5
^D $
Some may accuse me of being overly simplistic, but I'm not convinced that one would actually need much more than that to take notes. And, if one needs a more structure, it can be done with such mad features as a hierarchical filesystem. One even gets such features as full searchability (grep -ri ~/notes contract).
I'm not sure if I can agree. Even if you have 10 MB of cache, it would only take about a millisecond to fill it with PC5300 RAM, which is only 1/20 of a timeslice, and it's not as if the CPU stands still during that time. And that's assuming that the cache has been completely invalidated since the last timeslice, which is rather unlikely (especially there are many CPUs (or cores) and the system has a smart scheduler).
What kind of tea did you spill on it that actually ruined it?
For the record, it was Oolong tea.
Model Ms are nearly indestructible.
Yes, I was quite surprised myself, because it turned out that it had manage to erode one of the metal conductors in the key matrix to the point where it was no longer closed. I still don't really get how it could get that badly affected by a little tea.
Unfortunately, I could only actually see that once I had dismantled the keyboard, which required destroying up to 50 points where the plastic holding plate for the actual spring actuators was riveted to the backing metal plate by melting. I decided that it was worth paying $70 for a new keyboard from Unicomp (which was USB and had hyper buttons, to boot) rather than trying to re-fasten it.
I've never really understood the problem with creating a more stringent definition of the kilogram. Other SI units are measured in measurable quantities, such as the second being defined in terms of cycles of radiation from Caesium atoms. Why cannot the kilogram just as easily be defined as the mass of a certain number of atoms of one or another kind?
Of course, I'm no experimental physicist, but if I were to guess, I might suggest the fact that the binding energy (and thus the mass) might change with force-field fluctuations in the vicinity, but I think that problem should be solvable by defining the proper environment for measuring.
What's to stop my friends/enemies from posting pictures of me online?
This might sound weird, but, how about you yourself?
Re:There is only one true keyboard...
on
Review of Das Keyboard
·
· Score: 2, Informative
Unicomp sells a 104-key version of the Customizer that's USB-native. I'm typing on one right now. It's/slightly/ more lightly-built than the Lexmark M 3 feet away and my IBM M at home, but it's much better IMO than a standard kb.
Yes, I have one too. I bought it after I ruined my original Model M by spilling tea on it, and I have to concede that real Model M was better, quality-wise -- in particular, the C key on this board gets stuck in some little plastic detail when depressed from the wrong angle, and it doesn't have removable key caps.
Nevertheless, it is still incomprehensibly much better than any run-of-the-mill rubber dome rubbish, and it sells for lot less than Das Keyboard at $69. If you're not in a position to get your hands on a real Model M, I greatly recommend it.
Ignoring the parent troll for a moment, can someone please show an example of where multiple implementation inheritance is superior (not just equivalent) to multiple interfaces and the composition design pattern?
If you're going to say that, you might as well go as far as saying that inheritance itself has no merit when Java has its interfaces instead. And I'd argue that you'd have a point.
After all, inheritance is just a convenient way of taking functions from some other implementation of an interface and making your own implementation use those same functions. In that regard multiple inheritance makes sense, since you'd be able to pull in code from all over the place easily. But, one might argue that it should be possible (and easy) to do that more explicitly than using inheritance.
Yes, I agree completely. To us Slashdot readers, it might be hard to imagine a world without broadband access to the Internet, but even disregarding the fact that the vast majority of the Earth's population just doesn't have the economy to consider playing around on the Internet, to people who aren't programmers, the Internet is almost exclusively just a way to communicate with other people. And, believe or not, but there are lots of other ways to do that.
To me, 25% seems almost incredibly much. I'd have already been surprised at 10%.
In the font my system's using, they all look completely identical Excepting, of course, that Slashdot apparently doesn't except Unicode from forms. What was this again? 2008?
Or even better, they could use IDN to construct TLDs like.cÎm,.cоm,.cÖ...m or.cám. In the font my system's using, they all look completely identical, give or take an antialiased pixel here or there. If you want a cheat sheet, the first is written with omicron, the second with cyrillic o, the third with an armenian oh, and the fourth with a "small capital O".
Except that Wine doesn't work. Or, if it does, then it's because you have a "paid-for (yeah, right)" Windows install somewhere on your hard drive (go try to run anything beyond Notepad with the wine dlls and weep while I rotflmao). I think you underestimate the progress Wine has made the last year or so. It used to be precisely as you describe, but lately I've discovered a surprising number of Windows programs that work more or less flawlessly on Wine.
Just to list a few, I've been able to play Starcraft, Warcraft III, use Declan's ReadWrite Kanji program and use Blizzard's downloader program for the Starcraft II videos. I've also been able to play MtG Online with only a little tweaking. In contrast, the only program I've tried and not been able to run has been the "School Days" visual novel, and only because it uses WMA.
Slashdot Mirror
The link points here.
Especially considering how ACPI doesn't really seem to be the greatest standard so far conceived, and on top of that is an Intel-Microsoft cooperative product, I just wouldn't bet my life and limb on its specification being sound.
HP has given me boxes that size for 4 screws in a plastic bag, wrapped in foam. Repeatedly.
If they do this so regularly as your comment and many others seem to suggest, I just cannot help but wonder: How do things go so wrong to begin with? Also, one would think that the errors of their ways ought to be completely obvious to anyone involved, so whyever do they not fix it?
There's no obvious reason to assume that whatever government the future holds will just magically forget the nuclear dumps and stop upgrading them.
All this is of course dependent on us not going through some technological singularity or anything like it under 10,000 years. What's even to say that "humans" 10,000 years from now will even have organic bodies that would be damaged by radioactivity?
Of course, Splashtop is extremely fast, but depending on what exactly he wants to do with these installation, a stock Debian Etch installation might function just as well. It usually boots in 20-30 seconds, which is certainly fast compared to many others, but an order of magnitude slower than Splashtop. So, it really depends on just how fast he wants them to boot.
See, I haven't used Windows myself for at least five years now, and yet it still gives me headaches. Disregarding headaches caused by peoples' dysfunctional Windows PCs that I have to fix every now and then, the greatest source of headaches is that people disregard that there are other systems. Such as hardware manufacturers, web designers, people sending you MSO files and assume you can read them, and so on. I don't think I have to create a complete list for you to get the picture.
Therefore, Apple climbing ever higher on the usage charts is occasion to celebrate indeed. Any dent that can be made to the monoculture is a good dent.
Windows firewall sucks at outbound protection, a lot...
Is it only I who get the mental image of the Berlin wall when people speak of outbound "protection"?
Speaking of which - how would WinFS and ZFS compare?
Much like chalk and cheese. WinFS is(/was) *not* a filesystem, it's a database/metadata layer that sits between the filesystem (NTFS) and the applications.
Indeed. I find it well acceptable that a random slashdotter wouldn't know this, but it disturbs me all the more that the author of TFA says the same in a authoritative voice.
In fact, apart from a few obvious points (like Windows needing better CD burning tools, or that it might not need to be released in umpteen editions), all his points are more or less completely bogus:
20. Modularized OS
Yes, that's easy to say, in about the same way that it's easy to say that "Windows 7 should be better than Vista". He does not appear to have any clear idea in what manner Windows should be modularized or even why. In fact, Windows is probably far more modular than he actually realizes -- it's just that Microsoft only releases complete builds of it.
19. XP Virtual Machine
I don't even know where to start on this. Lots of people seem to think that all Microsoft would have to do is throw an instance of XP into a VM and everything would be fine. It doesn't take much to realize the stupidity of that idea. To begin with, it is obvious that they can't make the instance of XP completely virtualized, because then it would have access to none of the actual hardware, which would result in large outcries of people like gamers or anyone who expects to use any piece of userspace software that comes with a piece of hardware they bought (like synchronization software for their PDA:s or whatever).
So, they'd need to allow API calls through to the real operating system running on the real hardware. Once they've wrapped everything that needs to be wrapped for any real level of hardware compatibility, they're back to square 1, only with a useless CPU virtualizer in between.
18. New UAC
In theory UAC was a great idea.
UAC was never a great idea, neither in practice nor in theory. All it can do is tell the user that some program needs more privileges, but it can't actually know why, and neither can the user. It just gives a false sense of security and is probably designed to push the blame onto the users and everything which has already been discussed to death here on Slashdot.
17. Gaming Mode
I would much like to know what services he thinks are detrimental to gaming performance. I'd be very surprised if he'd find even one which would increase gaming performance, by any metric, by even a percent when turned off. The indexer service in Windows Vista might be an exception, but even then, I don't see a reason for an entire "Gaming Mode" just to turn it off.
13. WinFS
Yes, here it was; the author showing off his perfect cluelessness by claiming that WinFS would be an "NTFS replacement". He also makes the completely unbased claim that, and I quote, "the relational database structure should enhance overall system performance". However a relational database would actually be faster than a filesystem is completely beyond me.
9. Program Caching
He's actually trying to blame users' (understandable) ignorance about virtual memory and block device caching on Microsoft. While I might agree that Windows' Program Manager may be a bit too dense about what it actually displays, it's not as if there even exists a metric that Microsoft could choose to display that would make users content.
6. Barebones Kernel
This idea has been thrown around by Microsoft, specifically âMinWinâ(TM). Allowing the user to choose between this and the default kernel could potentially allow older systems (i.e. XP based) to run the new OS with decent performance levels.
I think this point almost takes the price i
SPF is great. It isn't a total solution, and there are negatives, but it certainly is better than the anyone is anyone free for all.
Actually, SPF sucks really badly, and if you've turned it on, you should turn it off immediately. The reason SPF sucks so bad is that it unconditionally and helplessly breaks all forms of forwarding and mailing lists. As such, it is indeed worse than anyone is anyone free for all.
Instead, you should support DKIM, which solves the same problem without those bugs.
DKIM ensures the integrity of the message and that it, at some point, passed through the MTA for the domain in the "From" header. It does that by signing the body of the message along with a set of headers. As long as the mailing list is RFC 2821/2822 compliant and only adds new headers to the top of the message (only headers below the DKIM header are part of the signature), its integrity under DKIM is unaffected.
I can't say I can tell you right off the bat how to do the same thing with Ubuntu, but given that it's Debian-based, I can't imagine it being terribly hard.
13,320k for Word, and 46,716k for Writer.
Speaking of which, though; does anyone know what the memory column in Windows' task manager really means? That is, in terms of what it measures. Is it resident pages? Private pages exclusively? The process' virtual size?
I know that Java programs usually show up as taking >100 MB in Windows's task manager, but in Linux/Unix, top makes it clear that that's just the virtual size, and that the actual resident size is usually closer to 20 MBs (it seems that Sun's Java runtime maps a huge heap with lazy page allocation), which leads me to believe that the task manager might actually be reporting the virtual size, which, while interesting in its own right, doesn't really tell very much about the actual memory usage of a process.
Open Source Onenote?
Yes, one you don't have to download, and one you do.
On Linux: Basket Notepads
I was thinking cat on Linux. You know, like:
Some may accuse me of being overly simplistic, but I'm not convinced that one would actually need much more than that to take notes. And, if one needs a more structure, it can be done with such mad features as a hierarchical filesystem. One even gets such features as full searchability (grep -ri ~/notes contract).
I'm not sure if I can agree. Even if you have 10 MB of cache, it would only take about a millisecond to fill it with PC5300 RAM, which is only 1/20 of a timeslice, and it's not as if the CPU stands still during that time. And that's assuming that the cache has been completely invalidated since the last timeslice, which is rather unlikely (especially there are many CPUs (or cores) and the system has a smart scheduler).
What kind of tea did you spill on it that actually ruined it?
For the record, it was Oolong tea.
Model Ms are nearly indestructible.
Yes, I was quite surprised myself, because it turned out that it had manage to erode one of the metal conductors in the key matrix to the point where it was no longer closed. I still don't really get how it could get that badly affected by a little tea.
Unfortunately, I could only actually see that once I had dismantled the keyboard, which required destroying up to 50 points where the plastic holding plate for the actual spring actuators was riveted to the backing metal plate by melting. I decided that it was worth paying $70 for a new keyboard from Unicomp (which was USB and had hyper buttons, to boot) rather than trying to re-fasten it.
Of course, I'm no experimental physicist, but if I were to guess, I might suggest the fact that the binding energy (and thus the mass) might change with force-field fluctuations in the vicinity, but I think that problem should be solvable by defining the proper environment for measuring.
Does anyone know?
What's to stop my friends/enemies from posting pictures of me online?
This might sound weird, but, how about you yourself?
Unicomp sells a 104-key version of the Customizer that's USB-native. I'm typing on one right now. It's /slightly/ more lightly-built than the Lexmark M 3 feet away and my IBM M at home, but it's much better IMO than a standard kb.
Yes, I have one too. I bought it after I ruined my original Model M by spilling tea on it, and I have to concede that real Model M was better, quality-wise -- in particular, the C key on this board gets stuck in some little plastic detail when depressed from the wrong angle, and it doesn't have removable key caps.
Nevertheless, it is still incomprehensibly much better than any run-of-the-mill rubber dome rubbish, and it sells for lot less than Das Keyboard at $69. If you're not in a position to get your hands on a real Model M, I greatly recommend it.
I just thought you might like to know; you can use ^W to erase a word, rather than ^H to erase each character in it.
Ignoring the parent troll for a moment, can someone please show an example of where multiple implementation inheritance is superior (not just equivalent) to multiple interfaces and the composition design pattern?
If you're going to say that, you might as well go as far as saying that inheritance itself has no merit when Java has its interfaces instead. And I'd argue that you'd have a point.After all, inheritance is just a convenient way of taking functions from some other implementation of an interface and making your own implementation use those same functions. In that regard multiple inheritance makes sense, since you'd be able to pull in code from all over the place easily. But, one might argue that it should be possible (and easy) to do that more explicitly than using inheritance.
To me, 25% seems almost incredibly much. I'd have already been surprised at 10%.
Or even better, they could use IDN to construct TLDs like .cÎm, .cоm, .cÖ...m or .cám. In the font my system's using, they all look completely identical, give or take an antialiased pixel here or there. If you want a cheat sheet, the first is written with omicron, the second with cyrillic o, the third with an armenian oh, and the fourth with a "small capital O".
Just to list a few, I've been able to play Starcraft, Warcraft III, use Declan's ReadWrite Kanji program and use Blizzard's downloader program for the Starcraft II videos. I've also been able to play MtG Online with only a little tweaking. In contrast, the only program I've tried and not been able to run has been the "School Days" visual novel, and only because it uses WMA.