Perhaps it will not beat the Athlon clock by clock, but it will start with 1.5
GHz und will scale well beyond 2 Ghz this will make it the performance leader for some time.
Huh. By the time it ships AMD will be shipping K7's at the same clock speed, but with much better throughput. And sledgehammer is coming sooner than you think. --
I've noticed various weird display problems on Windows and Linux systems when the display resolution is smaller or larger than the system the
application was developed on. Plus, some of the operating system provided graphics and fonts do not scale well.
Are there any fixes for this problem? What happens when we get 4000x3000 displays?
Yes, of course there are fixes. First, it's possible to do a decent job of scaling *anything*, even bitmap fonts and pictures, if you take a halfway decent approach to the problem. Hint: do you know what a sinc filter is? If not, you better find out fast, or you have no business writing a bitmap stretcher.
Second, a lot of the formats that are now becoming popular are inherently resolution-indenpendent, for example, any lossy image compression format - jpg (DCT), fractal encoding;-), wavelets, whatever. Outline fonts too.
We need to carry this kind of resolution-independent design all the way through the entire system - from Web design to automatic screen geometry to font rendering. A huge task, but it's underway.
Probably the biggest obstacle is entrenched page/screen designs that were for some incomprensible reason, done with the assumption that screen resolution would never change. For a good example of this, just go to Yahoo. Notice how it's increasingly taking on the appearance of a postage stamp in the middle of your screen.
(Note: screen resolution doesn't just *increase* over time, it decreases too. If you don't believe me, check out IBM's Linux-on-a-wristwatch.)
Even in the face of thoughtlessness on the part of web designers, we can patch up the resulting stupidity with scaling techniques. Check out Andy Hertzfeld's Eazel project and you will see that somebody out there is actually thinking about this.
--
I recently became annoyed at gcc for failing to do the right thing with this expression:
(unsigned int) ((unsigned long long) x / (unsigned int) y)
It should produce a 64/32 divide (on 32 arch); instead it produced 64/64, which is not only a lot slower, but isn't even supported in the kernel. Arrgh. --
Personally, I think it's a waste of time and space to squeeze a Linux kernel on a PDA.
What part of 'Free as in Speech' do you not understand?
checking the entire flash memory would take a long time
There are a whole host of new filesystems on the way that eliminate fsck. In the meantime, using FAT (!) is a reasonable compromise, because while it is slow and stupid, it is also rugged in the face of sudden interruptions (guess why it *had* to be).
Furthermore, what's the point of shoving a server operating system onto a handheld organizer?
Hint: don't install the server, you probably won't need it. You can probably get away with not loading Samba or Wine as well.:-p --
Hint: If you are trying to get open source developers to write stuff for your product, don't give out anything in Microsoft Word format! (Use html. Use pdf. Anything)
And could you tell us something about the licence, please? --
Why open source it? Because coda is about to replace it.
At UKUUG this year, Owen LeBlanc, a Coda expert if there ever was one, said
"if you have a small
number of users and a relatively small amount of data, then Coda may be just what you need". I also seem to recall him saying he thought AFS is pretty darn nice. He'd be the one to know. --
If only Sony were smart enough to put this chip in their regular Vaios, they wouldn't be able to keep them on the shelves.
Right on the money. I'm in the "desktop replacement" market. I'm just about to buy a Sony VAIO F560 but I'd change my plans in an nanosecond if I could get the same machine with a power-sipping Transmeta chip, and thereby get more than an hour of battery-powered compute time. Better, give me a laptop with *two* Transmeta chips. Can anyone think of a single reason why they wouldn't want dual Transmeta's in a laptop? --
...the slow release cycle is slowly stripping debian marketshare to the tpoint where none of this will matter. I can't see debian being relevant in two years
at this pace.
Don't be silly. Exactly because of its conservativism and slow release cycle, Debian is made-in-heaven as a base distribution for other distributions. For example, Corel, Storm Linux and many others.
It doesn't hurt that you can be *sure* that all the code in Debian is free, as in both beer and speech. --
Can this be done with minimal, or *no* loss of quality? IOW, can ogg represent the exact same DCT parameters, block sizes, etc, as mp3 does? Because it would just be that bit of icing on the cake if I would wave my wand and have all my mp3's magically become oggs. --
One thing the current release of Vorbis does not have is channel coupling (like mid-side stereo, although we'll be doing it differently). Beta 1
and beta 2 actually include multiple totally separate channels. The fact that we equal and better mp3's quality missing this huge piece is
exciting.
This makes me feel warm and fuzzy all over.
For those who don't get it yet, please read the writing on the wall: The old king of web music, mp3, isn't dead yet but he's dying. The new king is born, and his name is Ogg.
The old king is dying of complications from a disease brought on by misuse of intellectual property.
--
Windows is everywhere. Like it or not, Windows does a much much better job with games and multimedia right now than linux.
Rewrite the sentence this way:
DOS is everywhere. Like it or not, DOS does a much much better job with games and multimedia right now than Windows.
This simple fact kept Dos alive as the premier gaming platform on PC's for a few years. But not forever. Eventually, the problems with Windows were fixed and the writing was on the wall for Dos gaming.
The writing is already on the wall for Windows. Linux has made tremendous strides ahead as a gaming platform in the last year. Are there any signs this momentum will slow down? No, none. In fact, expect it to accelerate, as many game developers now 'get' this simple fact: it's awfully tough to compete against "free and open". Windows may not be washed up as a gaming platform today, but 3 years from now people will wonder why they ever bothered.
--
USB support. Yeah, it's there, it works OK, but it's a sweet fuggin pain in the ass to get working. In windows; I plug my rio/joystick whatever in, it's detected, the driver
prompted or in most cases automatically set up, and then through the magics of DirectX, all games see it. Linux doesn't have anything even CLOSE right now. Although -
the underlying USB stuff is slick - the intergration into the desktop (Gnome, etc) isn't there yet.
Your experience differs markedly from mine. Presumeably you are using the USB backport into the 2.2 series. Fair enough. That's the version gamers *should* be using, but let me tell you, USB is a pretty darn smooth in the 2.4.0-test series, soon to be the new stable series.
My experience with USB under Linux: plug it in and it works. In fact, for me it works *a damn sight better* than Windows. Under windows, my usb web cam runs for about 10 minutes before the picture vanishes, then the only option to get it going again is reboot. Not very useful. Under Linux it never stops (well hardly ever - if I leave it overnight with the lights out it gets confused, but I can restart it without rebooting). --
The revelation that Oracle hired a group of bungling private dicks to dig
into the finances (and through the trash) of trade groups that support
Microsoft surprised exactly no one.
Give me a break. This "bungling group of dicks" managed to catch Microsoft with its corporate pants down at embarassing times in the recent trial three times or so. Little things like Microsoft presenting testimonials from supposedly independent groups, only to have David Boise immediately follow up with evidence that that those organizations were funded by Microsoft. Guess who dug up the evidence? --
Actually, before I get flamed, let me add... *the stuff they're talking about here* seems to be real voxel rendering. (Marching cubes algorithm, etc.) What is commonly called voxel rendering is *not*. But it's that misnamed technology that's going to take over the world. --
They're actually "surfaces of elevation". But Novalogic, starting with their ground breaking Comanche game, abused the terminology, and called their clever rendering method for surfaces of elevation "Voxel Space[tm]". (They tried to patent it too.) The terminology stuck.
Whatever. In 5 years surfaces of elevation will rule the 3D game world. Call them what you want to.
--
I think the reason that things are progressing so slowly is because what we're doing is REALLY HARD.
Well, that *sounds* convincing, but a quick check of the mailing list shows that there is roughly zero discussion going on at the moment about the crucial issues I mentioned (updating and searching). And roughly the same amount of development. Why do think that is? Sure my post was controversial, but I stand by what I said. I think the real problem is that a lot of developers who have the right stuff to attack and solve these problems are not being well-received by some developers already entrenched in the project. Try it yourself and you'll see what I mean.
Open source developers tend to exhibit well-known cat-like behavior: they do what comes naturally and move on to some other project that can make better use of their talents. If you go back, *way back* through the list archives you'll see exactly what I'm talking about happening over and over again. Don't be distracted by the volume of posts - you have to look closely at the quality to see what I mean. Whatever, I wish you the best of luck. --
This looks like the Cyrix Jalapeno (Which I thought was cancelled...)
Still, by the time this hits mass production it will be incredibly out of date. Standard for the recent Cyrix processors (Ironic really, because back in the 486 days they were
bigger than AMD).
I remember very similar things being said about the first Japanese 64K DRAMs, which were made by simply reproducing the mask of 16K DRAMs four times. Lame, huh?
Now look at the balance of power in the DRAM market. But actually, I think that *we, the geeks* are the big winners when more players come to the processor party, and frankly I don't give a damn on which side of the Pacific the balance of power lies. --
Freenet appears to be quite a bold project. However it already faces competition from the most common distributed file sharing services:
Napster
Gnutella
Yes, and Freenet development proceeds at a snail's pace compared to either of the above. I know why too: some of the early developers in the project are more interested in preserving their own exalted positions in the project that in letting the design/development process move forward effectively. There is more 'blocking' going on in the project than actual code development. I don't doubt it will eventually get where it's trying to go, but the question is: how many other projects with the same goals are going to get there first? I mean, come on, it's been months, and still no way to update information? Or search it? Give me a break, those problems are not insoluable.
Ian is pretty cool, and especially, his philosphy is right on the money. His original paper is a mighty good read. But it takes more than philosophy to build a killer software app. Please note, this isn't a troll or a flame, it's because I actually care about the project. Get your act together guys, bury the egos, and you'll have a worldbeater. --
Of all the slashdot editors, I enjoy reading your posts the most. They tend to be the most thoughtful, and have the fewest spelling mistakes. Frankly, I'm *amazed* at the number of Slash editorial posts that have typos, wordos, thinkos, etc. in them. Better add a preview button to the the story submit queue.
But I digress. Timothy, whoever you are, keep it up. --
Slashdot Mirror
Perhaps it will not beat the Athlon clock by clock, but it will start with 1.5 GHz und will scale well beyond 2 Ghz this will make it the performance leader for some time.
Huh. By the time it ships AMD will be shipping K7's at the same clock speed, but with much better throughput. And sledgehammer is coming sooner than you think.
--
In case that's what you were thinking.
conjuring up warm fuzzy images for countless investors, as well as limitless mock fodder for folks like me
*** hates it when Slash editors don't do their research
--
I've noticed various weird display problems on Windows and Linux systems when the display resolution is smaller or larger than the system the application was developed on. Plus, some of the operating system provided graphics and fonts do not scale well. Are there any fixes for this problem? What happens when we get 4000x3000 displays?
;-), wavelets, whatever. Outline fonts too.
Yes, of course there are fixes. First, it's possible to do a decent job of scaling *anything*, even bitmap fonts and pictures, if you take a halfway decent approach to the problem. Hint: do you know what a sinc filter is? If not, you better find out fast, or you have no business writing a bitmap stretcher.
Second, a lot of the formats that are now becoming popular are inherently resolution-indenpendent, for example, any lossy image compression format - jpg (DCT), fractal encoding
We need to carry this kind of resolution-independent design all the way through the entire system - from Web design to automatic screen geometry to font rendering. A huge task, but it's underway.
Probably the biggest obstacle is entrenched page/screen designs that were for some incomprensible reason, done with the assumption that screen resolution would never change. For a good example of this, just go to Yahoo. Notice how it's increasingly taking on the appearance of a postage stamp in the middle of your screen. (Note: screen resolution doesn't just *increase* over time, it decreases too. If you don't believe me, check out IBM's Linux-on-a-wristwatch.)
Even in the face of thoughtlessness on the part of web designers, we can patch up the resulting stupidity with scaling techniques. Check out Andy Hertzfeld's Eazel project and you will see that somebody out there is actually thinking about this.
--
--
Personally, I think it's a waste of time and space to squeeze a Linux kernel on a PDA.
:-p
What part of 'Free as in Speech' do you not understand?
checking the entire flash memory would take a long time
There are a whole host of new filesystems on the way that eliminate fsck. In the meantime, using FAT (!) is a reasonable compromise, because while it is slow and stupid, it is also rugged in the face of sudden interruptions (guess why it *had* to be).
Furthermore, what's the point of shoving a server operating system onto a handheld organizer?
Hint: don't install the server, you probably won't need it. You can probably get away with not loading Samba or Wine as well.
--
Hint: If you are trying to get open source developers to write stuff for your product, don't give out anything in Microsoft Word format! (Use html. Use pdf. Anything)
And could you tell us something about the licence, please?
--
Why open source it? Because coda is about to replace it.
At UKUUG this year, Owen LeBlanc, a Coda expert if there ever was one, said "if you have a small number of users and a relatively small amount of data, then Coda may be just what you need". I also seem to recall him saying he thought AFS is pretty darn nice. He'd be the one to know.
--
It prolly means our fearless leader will soon be our *rich* fearless leader.
--
Half a meg of binary goop!
--
Gosh, this looks a lot like source code: HALlib README mga.cpp mga_bios.h mga_dga.c mga_driver.c mga_macros.h mga_shadow.c mga_video.c mga_wrap.h Imakefile client.h mga.h mga_dac3026.c mga_dri.c mga_driver.c~ mga_map.h mga_storm.c mga_warp.c mgareg_flags.h Makefile clientlx.c mga_arc.c mga_dacG.c mga_dri.h mga_hwcurs.c mga_reg.h mga_ucode.h mga_wrap.c
It didn't build for me because I don't have the XFree 4.0.x headers, but....
*** Tough Love pinches himself & still doesn't quite believe he's not dreaming
--
Gnome provides all of the look, none of the feel
This is why we need experienced people like Andy Hertzfeld in the loop. Cool by itself is not enough.
Gnome moderation radicals: please don't mistake this for an attack.
--
After all, what works for Corel it should work for Microsoft, right??
Let's think about how to put some one-way signs on this bridge pointing *our* way.
--
If only Sony were smart enough to put this chip in their regular Vaios, they wouldn't be able to keep them on the shelves.
Right on the money. I'm in the "desktop replacement" market. I'm just about to buy a Sony VAIO F560 but I'd change my plans in an nanosecond if I could get the same machine with a power-sipping Transmeta chip, and thereby get more than an hour of battery-powered compute time. Better, give me a laptop with *two* Transmeta chips. Can anyone think of a single reason why they wouldn't want dual Transmeta's in a laptop?
--
...the slow release cycle is slowly stripping debian marketshare to the tpoint where none of this will matter. I can't see debian being relevant in two years at this pace.
Don't be silly. Exactly because of its conservativism and slow release cycle, Debian is made-in-heaven as a base distribution for other distributions. For example, Corel, Storm Linux and many others.
It doesn't hurt that you can be *sure* that all the code in Debian is free, as in both beer and speech.
--
Can this be done with minimal, or *no* loss of quality? IOW, can ogg represent the exact same DCT parameters, block sizes, etc, as mp3 does? Because it would just be that bit of icing on the cake if I would wave my wand and have all my mp3's magically become oggs.
--
One thing the current release of Vorbis does not have is channel coupling (like mid-side stereo, although we'll be doing it differently). Beta 1 and beta 2 actually include multiple totally separate channels. The fact that we equal and better mp3's quality missing this huge piece is exciting.
This makes me feel warm and fuzzy all over.
For those who don't get it yet, please read the writing on the wall: The old king of web music, mp3, isn't dead yet but he's dying. The new king is born, and his name is Ogg.
The old king is dying of complications from a disease brought on by misuse of intellectual property.
--
Windows is everywhere. Like it or not, Windows does a much much better job with games and multimedia right now than linux.
Rewrite the sentence this way:
DOS is everywhere. Like it or not, DOS does a much much better job with games and multimedia right now than Windows.
This simple fact kept Dos alive as the premier gaming platform on PC's for a few years. But not forever. Eventually, the problems with Windows were fixed and the writing was on the wall for Dos gaming.
The writing is already on the wall for Windows. Linux has made tremendous strides ahead as a gaming platform in the last year. Are there any signs this momentum will slow down? No, none. In fact, expect it to accelerate, as many game developers now 'get' this simple fact: it's awfully tough to compete against "free and open". Windows may not be washed up as a gaming platform today, but 3 years from now people will wonder why they ever bothered.
--
USB support. Yeah, it's there, it works OK, but it's a sweet fuggin pain in the ass to get working. In windows; I plug my rio/joystick whatever in, it's detected, the driver prompted or in most cases automatically set up, and then through the magics of DirectX, all games see it. Linux doesn't have anything even CLOSE right now. Although - the underlying USB stuff is slick - the intergration into the desktop (Gnome, etc) isn't there yet.
Your experience differs markedly from mine. Presumeably you are using the USB backport into the 2.2 series. Fair enough. That's the version gamers *should* be using, but let me tell you, USB is a pretty darn smooth in the 2.4.0-test series, soon to be the new stable series.
My experience with USB under Linux: plug it in and it works. In fact, for me it works *a damn sight better* than Windows. Under windows, my usb web cam runs for about 10 minutes before the picture vanishes, then the only option to get it going again is reboot. Not very useful. Under Linux it never stops (well hardly ever - if I leave it overnight with the lights out it gets confused, but I can restart it without rebooting).
--
The revelation that Oracle hired a group of bungling private dicks to dig into the finances (and through the trash) of trade groups that support Microsoft surprised exactly no one.
Give me a break. This "bungling group of dicks" managed to catch Microsoft with its corporate pants down at embarassing times in the recent trial three times or so. Little things like Microsoft presenting testimonials from supposedly independent groups, only to have David Boise immediately follow up with evidence that that those organizations were funded by Microsoft. Guess who dug up the evidence?
--
Actually, before I get flamed, let me add... *the stuff they're talking about here* seems to be real voxel rendering. (Marching cubes algorithm, etc.) What is commonly called voxel rendering is *not*. But it's that misnamed technology that's going to take over the world.
--
They're actually "surfaces of elevation". But Novalogic, starting with their ground breaking Comanche game, abused the terminology, and called their clever rendering method for surfaces of elevation "Voxel Space[tm]". (They tried to patent it too.) The terminology stuck.
Whatever. In 5 years surfaces of elevation will rule the 3D game world. Call them what you want to.
--
I think the reason that things are progressing so slowly is because what we're doing is REALLY HARD.
Well, that *sounds* convincing, but a quick check of the mailing list shows that there is roughly zero discussion going on at the moment about the crucial issues I mentioned (updating and searching). And roughly the same amount of development. Why do think that is? Sure my post was controversial, but I stand by what I said. I think the real problem is that a lot of developers who have the right stuff to attack and solve these problems are not being well-received by some developers already entrenched in the project. Try it yourself and you'll see what I mean.
Open source developers tend to exhibit well-known cat-like behavior: they do what comes naturally and move on to some other project that can make better use of their talents. If you go back, *way back* through the list archives you'll see exactly what I'm talking about happening over and over again. Don't be distracted by the volume of posts - you have to look closely at the quality to see what I mean. Whatever, I wish you the best of luck.
--
This looks like the Cyrix Jalapeno (Which I thought was cancelled...) Still, by the time this hits mass production it will be incredibly out of date. Standard for the recent Cyrix processors (Ironic really, because back in the 486 days they were bigger than AMD).
I remember very similar things being said about the first Japanese 64K DRAMs, which were made by simply reproducing the mask of 16K DRAMs four times. Lame, huh?
Now look at the balance of power in the DRAM market. But actually, I think that *we, the geeks* are the big winners when more players come to the processor party, and frankly I don't give a damn on which side of the Pacific the balance of power lies.
--
Freenet appears to be quite a bold project. However it already faces competition from the most common distributed file sharing services: Napster Gnutella
Yes, and Freenet development proceeds at a snail's pace compared to either of the above. I know why too: some of the early developers in the project are more interested in preserving their own exalted positions in the project that in letting the design/development process move forward effectively. There is more 'blocking' going on in the project than actual code development. I don't doubt it will eventually get where it's trying to go, but the question is: how many other projects with the same goals are going to get there first? I mean, come on, it's been months, and still no way to update information? Or search it? Give me a break, those problems are not insoluable.
Ian is pretty cool, and especially, his philosphy is right on the money. His original paper is a mighty good read. But it takes more than philosophy to build a killer software app. Please note, this isn't a troll or a flame, it's because I actually care about the project. Get your act together guys, bury the egos, and you'll have a worldbeater.
--
Of all the slashdot editors, I enjoy reading your posts the most. They tend to be the most thoughtful, and have the fewest spelling mistakes. Frankly, I'm *amazed* at the number of Slash editorial posts that have typos, wordos, thinkos, etc. in them. Better add a preview button to the the story submit queue.
But I digress. Timothy, whoever you are, keep it up.
--