I don't really see why a microkernel can not have access to userland buffers and a monolithic kernel can.
Maybe QNX 4 (different from Neutrino, which is the current QNX) was not very smart, I don't know. And I personally am not hot on message passing either, but QNX remains a good example of message-passing microkernels being to perform relatively good.
But when you say: Because the driver is a user process, it simply can't do the kind of memory mapping that a Unix kernel can do to grab userland data
I would have to disagree. The kernel could easily provide services that would provide memory mapping to the driver. I don't see why not. In fact, with proper protection mechanisms in place there's no reason why any user app should not be able to figure out real memory addresses. I don't think that would be ugly.
The problem is that x86 just wasn't designed for microkernels (or operating systems in general, it seems)
I can smell a flamebait when I read one. Sorry, but that statement is plain silly. ia32 has (as you asked earlier in an other comment) excellent features to support a microkernel (or any OS), such as multiple levels of privileges, extensive protection mechanism and relatively fast context switching.
A system call (which is essentially nothing more than a jmp) takes 40 times longer than a regular function call (on my PII 300 anyway).
A jmp?? Don't you want it to return??? Linux uses a software INTERRUPT to do system calls (bad decision in my opinion, ia32 provides fine call-gates that are a lot faster).
A context switch (which microkernels do tons of)
A microkernel does not have to do tons of context switches. I think what you are talking about is message-passing kernels. A microkernel does not have to based on message passing. It can use calls, and in fact the ia32 architecture lends itself very nicely to switch between privilege levels quickly, thereby providing protection that a monolythic kernel lacks.
The prove that a well designed microkernel can be VERY fast is QNX.
Yes, I understand. And I mean that even if an admin is needed on the ground, the airlines won't put up with it. You just have to understand how airplane turnarounds work,- there is NO time to fsck around with systems like this. A system should work UNATTENDED for weeks, if not months.
1 802.11 accesspoint to serve, say, 300 passengers? Remember that their system actually has to scale all the way up to at least 500 passengers.
I know it's hard to believe now, but when you design a system like this you have to go for worst case,- when installed this has to work also 10 years from now.
I don't know that considering the physical space and number of users (say 500) 802.11 is the answer. I doubt it will work very well under these conditions though.
needs no admin except when it lands The airlines will NEVER buy into that. The only thing that 'flies' is 'needs NO admin'. fullstop.
Believe me, this environment requires solutions that you don't come up with in 5 seconds.
I work in the industry, and in fact our product would probably benefit (as in, sold more) if Connexion was available.
I have to say though that it is at the moment mainly vapor. (their demo link uses a satellite dish the size of Washington). I've read a bunch of their documents, and it's surprising how much time they spent on describing silly details and being very vague about how to actually solve the real problems.
Their biggest problem though: they have a.com business plan. As in, it don't make no sense. To sum it up: we are going get this fast pipe to the airplane and then we are all going to be rich. It sounds an awful lot like the in-seat airphones fiasco in the making. (for those who don't know it: these things have only COST money, which was carried by GTE & AT&T)
Again, I would love to have a high speed connection to the plane, but there are many problems to overcome. On the less-technical side for example: tech support. Take an office with 300 people all connected to the internet. What kind of staff is needed to support that? Who's going to do that in the airplane? I can guarantee you that it's not going to be the flight attendants. Especially in the US where their union will scream bloody murder over just the slightest increase in workload.
Sorry to be so negative, but the combination of Boeings bureaucracy and a.com business plan just doesn't sound good...
What the hell would a fast connection be good for unless you wanted to use p2p or other networking apps? Seriously, Normal webbrowsing, chatting, and email can be done flawlessly with a 56K modem. All the other things that people do with the internet use broadband.
I have yet to see a real legal reason for broadband
Uhm, I think you are wrong.
Once you are used to a 1.5Mb link, web-browsing is no longer a treat on a dialup. I was forced to use a dialup a couple of weeks ago and it was a rude awakening. You should try it. It's the inevitable 'no way back':o)
Regardless, I download datasheets for chips a lot. > 1 MB files are fairly normal for a 300 page PDF and I download quite a few. I wouldn't want to wait 56K times.
I also used to download the Microsoft DDKs etc. which where 30-40MB each. Those become tricky at a minimum on dialups. I know downloads from the Evil Empire don't count, but quite a few Open Source projects out there are similarly sized.
I can think of tons of other stuff that I can get legally and legitimatly. That's what I pay my ISP for. If you feel you pay yours for your needs, read the contract: if they don't live up to it, don't whine,- cancel or sue.
I frankly could NOT give a rats-ass if ISPs throttle P2P software. Do you really want me to believe you guys are using it legitimately? Do you REALLY want me to believe that mostly everything on there does NOT violate a copyright of some sort?!
I totally believe in freedom (of speech), and as such I totally hate the DMCA, RIAA AND MPAA.
But fuck it, MY internet connection gets slow because of people exchanging software (music, computer, whatever) illegaly. And my prices don't drop or my ISP goes out of business.
I don't think ISPs have the right to block just anything the want, but you sure make their case a lot more palettable when you don't use the internet responsibly. You can cry bloody murder about people taking away your ability to get your MP3s, but in the meantime your behaviour hurts everyone.
Obviously anyone posting so quickly after the article appeared didn't read it. It's impossible to sit down behind a keyboard after such a mercyless spanking.
Seriously, this is the best interview I've ever read. As a foreigner (alien) living in the US there's little political stuff I could do, but sponsoring the EFF is something EVERYONE can do. (and obviously not only in the interest of the US,- hint!)
Whoever hasn't heard of the plot by now probably has been living under a rock.
Yup, I would consider living anywhere outside the US living under a rock too, if I where American.
Because I wouldn't even know there WAS anything outside the US! Sjeesj, WHEN are you guys going to learn that everything that's tradition in the US is not always known to people from/in other countries?
I'm from Holland (since not too long live in the US, actually) and had never heard of the book until my (US) wife bought it. I would like to see the movie, and yes, he would have spoiled it.
Congratulations, you are getting the point. You may advance to level 2 soon.
You still haven't figured out your real task though.
Hint: How unrealistic do we have to make this monopoly before you start taking action?! If you think you are going to solve this one by joining the others here in bitching & complaining, you are dead wrong.
I suggest hijacking the 'magic lantern' with all evidence pointing towards Redmond,- that might get you somewhere. Think 'front page/.' kind of style. Good luck,- we'll be reading about you.
Uhm, no. The way the BIOS is fetched from flash is pretty much depending on where the code jumps. (Believe me, I wrote the bootcode to replace the BIOS on a system that we developed,- I hate calling ours a BIOS because it doesn't contain any of the legacy crap).
When a system boots, the RAM isn't even working. It is set up by the bootcode (BIOS), which only after that setup copies itself to RAM (shadowing) and disables the actual fetching from the flash chip.
So if your BIOS code jumps (which is actually the first instruction in most any BIOS, since it starts @ FFFFFFF0 (hex)), an other part of the flash is fetched.
No only that, he certainly does not over-state the difficulty of doing this stuff. This is NOT for the electronics-hobbiest. Espcially putting on the ROM socket is not easy (without shorting out something;o))
I have been working with the exact same device and would totally recommend an emulator. It's not cheap, but you are sure to screw up some Flash chips. I use the promice, which works well (although the software is, like with all of these emulators, braindead).
Also regarding the following quote:
My guess is if this information is true, there is a PIC 16LC63A microcontroller right next door to the FLASH...I wouldn't be surprised if the PIC microcontroller monitors the FLASH to make sure the right sequence of operations happens, and halts the processor if something is amiss.
See, if I where to design something like that, (wait a second,- I am in fact designing something like that, it's a little different though) I would do the following: Program a private key and decryption algo in the PIC, then encrypt the BIOS image using the companion public key. The PIC then decrypts the data comming out of the flash before it hits the Southbridge (I guess you could also use a symetric like 3DES algo with just a private key,- there are a couple of ways to go about it, really). One of the problems is that the code is not read sequentialy, but that would be easy to handle with some decent programming.:o)
They probably own an IEEE assigned range (which is the proper way to obtain MAC addresses for Ethernet interfaces see this link) and their software responds based upon that knowledge.
So the software really doesn't send a DHCP respond for only one MAC, but a range.
FYI: Diamond Multimedia owns 00-90-00 (which allows them to have 16,777,216 unique addresses).
On a related subject, I've been looking for a domain name that is a) easy to remember and b) does not generate a zillion hits if you type the name in a search engine. (and c) is not a silly long string of words).
It's funny how most people thing that common word domains are valuable, but forget that if you have a name that, when typed into a search engine, jumps out as the only result is pretty valuable too. Especially if it sounds like it is spelled.
Maybe not the best example, but since the 4 letter TLD's are practically all gone, I was going to register duxo.com. Unfortunately one of the many domain hogs got it the day I was going for it.:o(
I got an other one though, but it's not up yet so I won't tell what it is!;o)))
that don't make sense: (from the article)
A chip without a clock would be about as useful as a page of text without any space between the letters
Actually, it's about as useful as a page of text that only exists when you have your eyes closed.
Re:How fast compared to ATA-100?
on
Firewire and Linux?
·
· Score: 2, Informative
FireWire is 400mbs where ATA100 is 100
Right, the good ole Bytes vs. Bits swapping. Firewire is 400 Mega-bits-per-second.
ATA-100 is 100 Mega-BYTES-per-second. E.g. twice as fast as Firewire.
In either case, you would be hard pressed to find a drive that is capable of media-transfer rates to fill the bandwidth available.
more of a drain on the CPU
Horseshit,- both use PCI devices that use Bus-Master DMA. Setting up an ATA interface to do a transfer is very simple and does NOT take a lot of CPU at all. I have no experience with Firewire drivers, but I'm guessing that it takes more to manage a Firewire controller.
One thing I'll say about XP (and this is grandly off-topic, but I got Karma to spare;o)): ClearType kicks ass on LCD's.
Strange that it is not enabled by default (I guess it may not look that good on a CRT), so if you have an XP box and haven't enabled it yet, try it. It's under Display Properties, Appearance, Effects...
Unfortunately no Open Source alternative for this yet... And unfortunately no time to create one either...:(
I hummbly appologize for such a pro-M$ comment here. At least the box dual-boots Redhat.
They claim these cards would be 'voluntary', much as the act of leaving your home or purchasing groceries are voluntary activities.
* tsssj, tsssj, lights his flame thrower *
Sounds good at first glance. But this is utter horsehit. The comparison is rediculous and is designed to do nothing else but make us all feel warm and fuzzy about this thing. Yet what on earth could leaving your home _possibly_ have in common with getting a National ID card?
Not that I expected Ellison to say something meaningful. He's a master of coming up with statements like these.
If they are going to sell this to the american public with these kind of arguments, I'd be very worried. It's not that the card couldn't be a good idea, it's just the fact that they are obviously not level with us about their true intentions.
* foop... flame thrower off *
about this whole Transmeta thing is the level of speculation and un-clearness.
Talk all the shit you want about Intel, but I can tell you that I'm working on a board right now that uses a Mobile Celeron Mobile 400A: http://developer.intel.com/design/mobile/datashts/ 28365403.pdf. The datasheets says thermal power 10.1 Watt max. Well, we never _ever_ get that high. Also, the newer 500 Mhz ultra low power is 8 Watt max, 5 Watt under more normal conditions.
The thing is that TM _never_ published said figures (quickly: what's the MAX Watts a TM CPU can draw?), because supposedly all that we need to know is the power required to decode a DVD. Well, today that happens largely by the VGA controller now, doesn't it?
What suprises me even more is that Torvalds, if anyone, should know that using the simple HLT instruction in the idle thread, makes any Intel (or AMD) CPU draw a lot less power.
Even on paper I don't see the advantage of the TM CPU's. And I really hoped they would, believe me...
That would be like giving sight to the blind. I've yet to come up with wording to describe what I think when I use Windows. let alone a gesture that would do suffice.;o)
Slashdot Mirror
I don't really see why a microkernel can not have access to userland buffers and a monolithic kernel can.
Maybe QNX 4 (different from Neutrino, which is the current QNX) was not very smart, I don't know. And I personally am not hot on message passing either, but QNX remains a good example of message-passing microkernels being to perform relatively good.
But when you say: Because the driver is a user process, it simply can't do the kind of memory mapping that a Unix kernel can do to grab userland data
I would have to disagree. The kernel could easily provide services that would provide memory mapping to the driver. I don't see why not. In fact, with proper protection mechanisms in place there's no reason why any user app should not be able to figure out real memory addresses. I don't think that would be ugly.
The problem is that x86 just wasn't designed for microkernels (or operating systems in general, it seems)
I can smell a flamebait when I read one. Sorry, but that statement is plain silly. ia32 has (as you asked earlier in an other comment) excellent features to support a microkernel (or any OS), such as multiple levels of privileges, extensive protection mechanism and relatively fast context switching.
A system call (which is essentially nothing more than a jmp) takes 40 times longer than a regular function call (on my PII 300 anyway).
A jmp?? Don't you want it to return??? Linux uses a software INTERRUPT to do system calls (bad decision in my opinion, ia32 provides fine call-gates that are a lot faster).
A context switch (which microkernels do tons of)
A microkernel does not have to do tons of context switches. I think what you are talking about is message-passing kernels. A microkernel does not have to based on message passing. It can use calls, and in fact the ia32 architecture lends itself very nicely to switch between privilege levels quickly, thereby providing protection that a monolythic kernel lacks.
The prove that a well designed microkernel can be VERY fast is QNX.
Yes, I understand. And I mean that even if an admin is needed on the ground, the airlines won't put up with it. You just have to understand how airplane turnarounds work,- there is NO time to fsck around with systems like this. A system should work UNATTENDED for weeks, if not months.
1 802.11 accesspoint to serve, say, 300 passengers? Remember that their system actually has to scale all the way up to at least 500 passengers.
I know it's hard to believe now, but when you design a system like this you have to go for worst case,- when installed this has to work also 10 years from now.
I don't know that considering the physical space and number of users (say 500) 802.11 is the answer. I doubt it will work very well under these conditions though.
needs no admin except when it lands
The airlines will NEVER buy into that. The only thing that 'flies' is 'needs NO admin'. fullstop.
Believe me, this environment requires solutions that you don't come up with in 5 seconds.
but don't have your hopes up yet.
.com business plan. As in, it don't make no sense. To sum it up: we are going get this fast pipe to the airplane and then we are all going to be rich. It sounds an awful lot like the in-seat airphones fiasco in the making. (for those who don't know it: these things have only COST money, which was carried by GTE & AT&T)
.com business plan just doesn't sound good...
I work in the industry, and in fact our product would probably benefit (as in, sold more) if Connexion was available.
I have to say though that it is at the moment mainly vapor. (their demo link uses a satellite dish the size of Washington). I've read a bunch of their documents, and it's surprising how much time they spent on describing silly details and being very vague about how to actually solve the real problems.
Their biggest problem though: they have a
Again, I would love to have a high speed connection to the plane, but there are many problems to overcome. On the less-technical side for example: tech support. Take an office with 300 people all connected to the internet. What kind of staff is needed to support that? Who's going to do that in the airplane? I can guarantee you that it's not going to be the flight attendants. Especially in the US where their union will scream bloody murder over just the slightest increase in workload.
Sorry to be so negative, but the combination of Boeings bureaucracy and a
What the hell would a fast connection be good for unless you wanted to use p2p or other networking apps?
:o)
Seriously, Normal webbrowsing, chatting, and email can be done flawlessly with a 56K modem. All the other things that people do with the internet use broadband.
I have yet to see a real legal reason for broadband
Uhm, I think you are wrong.
Once you are used to a 1.5Mb link, web-browsing is no longer a treat on a dialup. I was forced to use a dialup a couple of weeks ago and it was a rude awakening. You should try it. It's the inevitable 'no way back'
Regardless, I download datasheets for chips a lot. > 1 MB files are fairly normal for a 300 page PDF and I download quite a few. I wouldn't want to wait 56K times.
I also used to download the Microsoft DDKs etc. which where 30-40MB each. Those become tricky at a minimum on dialups. I know downloads from the Evil Empire don't count, but quite a few Open Source projects out there are similarly sized.
I can think of tons of other stuff that I can get legally and legitimatly. That's what I pay my ISP for. If you feel you pay yours for your needs, read the contract: if they don't live up to it, don't whine,- cancel or sue.
Oh boy, this is not going to be popular.
I frankly could NOT give a rats-ass if ISPs throttle P2P software. Do you really want me to believe you guys are using it legitimately? Do you REALLY want me to believe that mostly everything on there does NOT violate a copyright of some sort?!
I totally believe in freedom (of speech), and as such I totally hate the DMCA, RIAA AND MPAA.
But fuck it, MY internet connection gets slow because of people exchanging software (music, computer, whatever) illegaly. And my prices don't drop or my ISP goes out of business.
I don't think ISPs have the right to block just anything the want, but you sure make their case a lot more palettable when you don't use the internet responsibly. You can cry bloody murder about people taking away your ability to get your MP3s, but in the meantime your behaviour hurts everyone.
That's why I say I don't give a damn.
Obviously anyone posting so quickly after the article appeared didn't read it. It's impossible to sit down behind a keyboard after such a mercyless spanking.
Seriously, this is the best interview I've ever read. As a foreigner (alien) living in the US there's little political stuff I could do, but sponsoring the EFF is something EVERYONE can do. (and obviously not only in the interest of the US,- hint!)
Whoever hasn't heard of the plot by now probably has been living under a rock.
Yup, I would consider living anywhere outside the US living under a rock too, if I where American.
Because I wouldn't even know there WAS anything outside the US! Sjeesj, WHEN are you guys going to learn that everything that's tradition in the US is not always known to people from/in other countries?
I'm from Holland (since not too long live in the US, actually) and had never heard of the book until my (US) wife bought it. I would like to see the movie, and yes, he would have spoiled it.
Congratulations, you are getting the point. You may advance to level 2 soon.
/.' kind of style. Good luck,- we'll be reading about you.
You still haven't figured out your real task though.
Hint: How unrealistic do we have to make this monopoly before you start taking action?! If you think you are going to solve this one by joining the others here in bitching & complaining, you are dead wrong.
I suggest hijacking the 'magic lantern' with all evidence pointing towards Redmond,- that might get you somewhere. Think 'front page
Although I largely agree with you (I mean, 512MB of RAM is now 40$ @ Frys), an other reason for keeping the kernel small is cache.
L2 cache is still fairly small, and the more you can do inside it the better. The performance hit of a cache miss is bigger than you think.
But, yeah, if the reason would be to keep it running on 10-20 year systems, that's wrong.
Uhm, no. The way the BIOS is fetched from flash is pretty much depending on where the code jumps. (Believe me, I wrote the bootcode to replace the BIOS on a system that we developed,- I hate calling ours a BIOS because it doesn't contain any of the legacy crap).
When a system boots, the RAM isn't even working. It is set up by the bootcode (BIOS), which only after that setup copies itself to RAM (shadowing) and disables the actual fetching from the flash chip.
So if your BIOS code jumps (which is actually the first instruction in most any BIOS, since it starts @ FFFFFFF0 (hex)), an other part of the flash is fetched.
No only that, he certainly does not over-state the difficulty of doing this stuff. This is NOT for the electronics-hobbiest. Espcially putting on the ROM socket is not easy (without shorting out something ;o))
:o)
I have been working with the exact same device and would totally recommend an emulator. It's not cheap, but you are sure to screw up some Flash chips. I use the promice, which works well (although the software is, like with all of these emulators, braindead).
Also regarding the following quote:
My guess is if this information is true, there is a PIC 16LC63A microcontroller right next door to the FLASH...I wouldn't be surprised if the PIC microcontroller monitors the FLASH to make sure the right sequence of operations happens, and halts the processor if something is amiss.
See, if I where to design something like that, (wait a second,- I am in fact designing something like that, it's a little different though) I would do the following: Program a private key and decryption algo in the PIC, then encrypt the BIOS image using the companion public key. The PIC then decrypts the data comming out of the flash before it hits the Southbridge (I guess you could also use a symetric like 3DES algo with just a private key,- there are a couple of ways to go about it, really). One of the problems is that the code is not read sequentialy, but that would be easy to handle with some decent programming.
In case you have been living under a rock, JBoss is an Open Source Java Application Server (J2EE)
That's right. Don't you just love it when people think the entire universe revolves around what they are doing.
But both you and I will be flaimed for being so ignorant. Even though it's pretty busy under this rock.
They probably own an IEEE assigned range (which is the proper way to obtain MAC addresses for Ethernet interfaces see this link) and their software responds based upon that knowledge.
So the software really doesn't send a DHCP respond for only one MAC, but a range.
FYI: Diamond Multimedia owns 00-90-00 (which allows them to have 16,777,216 unique addresses).
On a related subject, I've been looking for a domain name that is a) easy to remember and b) does not generate a zillion hits if you type the name in a search engine. (and c) is not a silly long string of words).
:o(
;o)))
It's funny how most people thing that common word domains are valuable, but forget that if you have a name that, when typed into a search engine, jumps out as the only result is pretty valuable too. Especially if it sounds like it is spelled.
Maybe not the best example, but since the 4 letter TLD's are practically all gone, I was going to register duxo.com. Unfortunately one of the many domain hogs got it the day I was going for it.
I got an other one though, but it's not up yet so I won't tell what it is!
that don't make sense: (from the article)
A chip without a clock would be about as useful as a page of text without any space between the letters
Actually, it's about as useful as a page of text that only exists when you have your eyes closed.
FireWire is 400mbs where ATA100 is 100
Right, the good ole Bytes vs. Bits swapping. Firewire is 400 Mega-bits-per-second.
ATA-100 is 100 Mega-BYTES-per-second. E.g. twice as fast as Firewire.
In either case, you would be hard pressed to find a drive that is capable of media-transfer rates to fill the bandwidth available.
more of a drain on the CPU
Horseshit,- both use PCI devices that use Bus-Master DMA. Setting up an ATA interface to do a transfer is very simple and does NOT take a lot of CPU at all. I have no experience with Firewire drivers, but I'm guessing that it takes more to manage a Firewire controller.
One thing I'll say about XP (and this is grandly off-topic, but I got Karma to spare ;o)): ClearType kicks ass on LCD's.
:(
Strange that it is not enabled by default (I guess it may not look that good on a CRT), so if you have an XP box and haven't enabled it yet, try it. It's under Display Properties, Appearance, Effects...
Unfortunately no Open Source alternative for this yet... And unfortunately no time to create one either...
I hummbly appologize for such a pro-M$ comment here. At least the box dual-boots Redhat.
The question I'm sure you're asking now is Why stop there?
No, actually, it's "Why FULLstop".
Just kiddin.
I'm naming my next game quake3.exe. Anything for a few more FPS. ;o))
They claim these cards would be 'voluntary', much as the act of leaving your home or purchasing groceries are voluntary activities.
* tsssj, tsssj, lights his flame thrower *
Sounds good at first glance. But this is utter horsehit. The comparison is rediculous and is designed to do nothing else but make us all feel warm and fuzzy about this thing. Yet what on earth could leaving your home _possibly_ have in common with getting a National ID card?
Not that I expected Ellison to say something meaningful. He's a master of coming up with statements like these.
If they are going to sell this to the american public with these kind of arguments, I'd be very worried. It's not that the card couldn't be a good idea, it's just the fact that they are obviously not level with us about their true intentions.
* foop... flame thrower off *
about this whole Transmeta thing is the level of speculation and un-clearness.
/ 28365403.pdf. The datasheets says thermal power 10.1 Watt max. Well, we never _ever_ get that high. Also, the newer 500 Mhz ultra low power is 8 Watt max, 5 Watt under more normal conditions.
Talk all the shit you want about Intel, but I can tell you that I'm working on a board right now that uses a Mobile Celeron Mobile 400A: http://developer.intel.com/design/mobile/datashts
The thing is that TM _never_ published said figures (quickly: what's the MAX Watts a TM CPU can draw?), because supposedly all that we need to know is the power required to decode a DVD. Well, today that happens largely by the VGA controller now, doesn't it?
What suprises me even more is that Torvalds, if anyone, should know that using the simple HLT instruction in the idle thread, makes any Intel (or AMD) CPU draw a lot less power.
Even on paper I don't see the advantage of the TM CPU's. And I really hoped they would, believe me...
Finally, Handspring has done it before everyone else
From the website: Coming early 2002
Sounds to me they haven't done anything yet.
Talk about being first: I bought my Nokia Communicator many _many_ years back. (mainly for the l33t telnet client hehe)
Now you can really tell Windows what you think ...
;o)
That would be like giving sight to the blind. I've yet to come up with wording to describe what I think when I use Windows. let alone a gesture that would do suffice.