I know that PC-based cards expect some functionality from the BIOS, but couldn't this be emulated with a clever driver?"
"The drivers for a huge amount of cards have already been written for Linux, so it would seem to me that once someone made a port for any driver to Sparc, the same techniques could be used to port all kinds of PCI cards.
well, I wish it was that simple...
but its not quite that easy.
The poster is right on the fact that it requires some functions from the bios, but its not something that can easily be simulated in a driver.
Unless you want to use it as a secondary display card designated strictly to do 3d work, then you would need to have the card itself suport the platform, at a point in the boot process BEFORE linux or solaris or any other OS loads..
What would be required would be an OpenFirmware bios for the card itself, in order to have the hardware suport it..
One possible way, which could be vbery much worth looking into, is cards made for Macs, as almost all PowerMacs use OpenFirmware, so it should theoratically be possible to go and use a mac card, and simply use the already-available Linux drivers for it.
But other than either using a Mac card, it would be very dificult to getting it to work on any sparc system, unless someones got enough time on their hands to port the BIOS of the card itself to Openfirmware... (which has been done before, there were 2 attempts at porting the nVidia TNT and TNT2 chipset's BIOSes to trhe Mac, but they were never stable enough, untill after 3DFX had released the mac drivers for their cards, making it more or less a moot point)
Just give it a higher priority, not the highest priority.
I made the mistake one time in class to set a counter program into realtime in NT, not even the mouse could get enough of a timeslice to register.
the counter sure as hell was speeding along tho;)
just make sure to set it to high, or +1 or +2, and not realtime
(not sure what priority levels are available in Win9x)
"I think one of the most egregious parts of the new APSL is the clause preventing you from making modifications for your own use without sending them in to Apple"
Not to nitpick, but have you ever actually READ the version 1.2 of the APSL? in it it clearly states that its free to use as you see fit for personal use, or internally for R&D purposes WITHOUT notifying apple of your modifications
if you wanna take the code, and hack it around to have darwin working on your power mac 6100*, go ahead, the code is there, tweak and hack away.
if you wanna make it work for one of the Motorola PowerMax 6000**'s you got off of ebay last year, go right ahead.
As long as your not putting it in a commercial production environment, or selling a product based on the code and your modifications to it, Apple doesnt give a rats ass about it.
no, its not as open and free as the BSD license.
but remember, you said you want the code to hack around, to change things in your environment...
GO RIGHT AHEAD
the APSL isnt stopping you. do as you see fit with the code, provided you dont release it in a production or commercial environment.
and if you were to do so, all you have to do is send your mods over to Apple, and if they think its something worthwhile including in the main darwin product, then they will do so.
Just send them notification.
and if you were to use linux for your project which you were selling, remember, you have to give your source code away anyway.
the only diference is that your sending your changes to one entity under the APSL, and to any joe hacker schmoe that wants to look at your code under the GPL.
*The powermac 6100 is the first powerpc-based mac, but since its based on the NuBus architecture, its pretty much incompatibvle with darwin, OSX, and pretty much anything except MKLinux
**The PowerMax 6000 was a CHRP (common Hardware Reference Platform) machine, the first chrp machine introduced to the general public (Macworld Boston 97, I had the pleasure of playing with one at the floor) but was only made as a few hundred prototypes, and killed shortly thereafter after Apple killed cloning and refused to allow motorola to sell them
and completely OT--- why does the Extrans mode (allowing HTML tags to the text) NEVER work?
it might be possible to do what you are thinking by setting up a few small devices around the yard that would emit a radio signal, thereby telling the robot its relative position in the yard.
i know there are systems for yards with a similar setup. although the sgnals are used not to control or give a position, but to keep a dog within the specified area (it will shock the dog if it tries to leave):P
Farallon made Timbuktu Pro for the Macs, and as of version 2.x, they also made it for windows, thereby enabling what i believe is for the first time cross-platform remote administration (this was back in early 90's)
Very handy tool, and one of the only ones available for a mac for a long time
Im not trying to say that they dont work, i had NetBSd running on my centris at one point in time, but the simple fact that they are so intenet on porting it to everything imaginable "Of Course it runs NetBSD" naturally means that a lot of new advances dont get put in because resources are more or less spread out along an insane number of platforms that is being suported..
No, I'm not trying to troll here, but is it truly possible to avoid having the linux kernel end up forking somehwhere along the line?
Right now you have linux running on a power mac, on an ultra sparc, on a 386, on an SGI, on a cisco router, on IBM mainframes, on tiny embedded projects...
You have one kernel, yet everyone is trying to push it every which way imaginable..
If you try to make the code trully able to take advantage of big iron, you make it so it cant really run on a 386. You keep it optimized for your 386, you wouldnt consider putting it anywhere near your mainframe without major code changes, ala what IBM is doing.
yet by keeping these things away from the kernel, it will basically force forking in the not too distant future...
i guess Linus might be tryin to avoid a situation like NetBSD that runs on everything imaginable, but doesnt really doa good job on any.,.
but if he is, with all the diferent directions Linux is headed, forking is just inevitable.
Actually, the RAID included in the Abit boards and almost all other IDE RAID boards are BIOS-based solutions..
they are still softare, just at a lower level (BIOS vs. Driver) but as far as true hardware raid, only one company makes a microcontroller-based IDE RAID card, which, at the price of high end adaptec raid cards, isnt too lovely an idea...
OMFG, i havent laughed this much in such a long time...
thank you
that was just perfect
:')
ROTFLMAO
please mod this up, it deserves at least a 5 funny...
But then again, you also have to take into account that each of those 10 G4 macs also have 10 300W power supplies, each generating its own ammount of heat, whereas one or 2 300w PS could take care of thsoe 10 G4 drive bays. And you also have several other components making heat, and wasting more electricity...
and getting back to the point of my original post, you also have to remember, even if it is more heat concentrated in that one cabinet with a couple of dozen of these drive bay computers, its still gonna be cheaper to cool that 1 cabinet then it is gonna to cool off an entire large room of huge computers.
Damn, thats gotta be the lowest UID ive ever seen around here on/.
damn.
Your old school;-)
</OT>
but getting back to topic, the added benefit that this also provides, versus a completely passive backplane solution, is that while you can dedicate a rack case to a backplane with a few of the controller cards, with these, you can also put a little server in the spare drive bays of your other servers, and anywhere that it might fit, since it isnt requiring its own case like a backplane solution.
Considering the lovely condition of state poewr problems here in california, this could be a great solution if implemented.
Since these machines are small, and are more or less fuill-featured computers, this could make building a server farm a lot cheaper and less power hungry, since one could have each of these set up as its own server, feeding off of a single rackmount cases's power supply, instead of having each server having its own oversized powersupply, as seen in SO many installations.
and at this size and power usage, it would also cut down on AC costs dramatically, as you can now fit several dozen computers in a space that you could possibly fit only maybe 10, therefore reducing the necessary cooling costs.
--warning--beowulf comment follows---
Now, where can i get a beowulf cluster of these?;-)
Ive seen the M$ optical trackballs around before, but one thing thats always come to my mind, and still does:
is there ANY benefit whatsoever from having a trackball be optcall, since its still relying on a ball?
i thought the major benefir of going towards the optical neutered mice was that you didnt have the probs associated with its one-balled brethren, the dust and all...
with a trackball, that is kinda pointless...
even the name implies a ball. =P
or is this simply a marketing tactic for M$, to use its latest technology and marketing gimicks to a market that it has no point being in?
while i agree that Intel chipsets do have SOME advantages, its kind of questionable the way you phrased your post, considering that it reads exactly like an intel PR ad.
The true benefits of intel chip[sets, are, in fact, the integration that they have.
by throwing toguether really crappy on-board video and sound, and maybe a NIC too, they are able tomake really cheap motherboards, while cutting the cost of a video card out the door for the system builder..
allowing them to either make more profits or pass the savings on to the consumer.
That is the reason why the duron, even as cheap and superior as it is, hasnt found much of a home, since the motherboards available to suport it are still so expensive..
and the one area where intel chipsets ruled above VIA for a LONG time, and still slightly do to this day, is memory bandwidth, which has tradtionally been somewhat of a strong point for intels chips..
And as far as the hub architecture goes, there are 2 main diferences between that and the northbridge/southbridge concept:
ONe: theres a special bus now connecting the 2 parts of the chipset toguether instead of the PCI bus
two: Intel gave it a diferent name.
the idea is still the same, however.
If you have some time to spare, and some are in the mood of a nice decent project, you might want to look into building your own box and putting it in the car, powering all the media you want, with a keypad and a small lcd display...
I have seen several of sites come and go, but the one with the more useful information, and the one im basing my system on the most (My 87 Plymouth Sundance has no sound right now, so im building a small box to handle CDs/MP3s/AM/FM and whoknows whtelse im in the mood for) is the InMotion car mp3 player http://www.jarcom.com/inmotion/
the site hasnt been updated since early summer, but the setup and schematics are still valid..
definately worth a look.
Also, if anyone else has done this, are there any sugestions/things to avoid in doing this?
-Joel
Well its probably cause every piece of "first evidence" got sucked into the black hole they were studying
;-)
or maybe it just found itself into earth's own little black hole, my closet
:P
What do you see in the future for *BSD, with the huge amount of popularity that linux keeps on receiving, not to mention attention, esp. from our buddy Bill Gate$...
Do you think it will remain the strong, viable but simply less popular free OS it is now, hiding behind the limelight of linux, or will it come up in popularity, esp with the codebase for Apple's Darwin, which is all BSD based?
Sorry to nitpick, but the first power-pc based mac was released on may 1994, and that was the Power Mac 6100, which was soon followed by the 71 and 8100 machines...
it was still a few more years till the 9500/120 came out (late 95)..
so i dont think your 9500/200 is QUITE that old yet
:P
Im not familiar with the palm software, as I havent had the pleasure of owning one. (but am willing to try, free donations accepted;-)
But wouldnt it be possible to run 2 copies of the sopftware, installed at diferent locations, and having them both run simultaneously, as diferent processes, each one with its own configuration settings?
like i said, im not familiar with it, but its worth a shot
g'luck
One detail I meant to add, With a new BIOS revision, most motherboards should work perfectly with this chip, just not out of the box, not for a while..
which will definately make it even less atractive for the do-it-yourselfer then it already is, but it should still be a cheap, atractive option for low-powered notebooks and subnotebooks, and the appliance market.
One important thing that most people have forgotten with the new CYrix's, you cant just pop them into any motherboard made for a P3 or celeron.
Well, technially you could, but one thing that has to be present in a motherboard for it to start up, is the processor's microcode in the BIOS.
that is why new BIOS revisions are always coming out saying "now supports this speed celeron or P3" or whatnot. Its not because the motherboard was previously incapable of handling it, but there was no micorocode instructions stored in the BIOS for that revision of the CPU.
Which is exactly why some motherboards needed new BIOS revisions to suport the P3 when it came out, or the CuMines..
they were mainly the same chip, but with diferent microcode, so it needed those instructions before it could properly POST.
Would you have access to a net connection and a burner?(long with a pile of blanks?)
If so, one cd, a net connection, and you can get access to everything else you might need, just download and burn;-)
just a thought
I can see one reason why you would want to do this, although i have no idea if that is why he would need to pull it off a DB..
in case there are several machines running apache, and any settings have to be changed, if they can collect those settings from a datavase automatically once the setting is changed in the database, it would be a lot easier then going through each one and running the script, and making sure all changes were applied.
just an idea..
Slashdot Mirror
I know that PC-based cards expect some functionality from the BIOS, but couldn't this be emulated with a clever driver?"
"The drivers for a huge amount of cards have already been written for Linux, so it would seem to me that once someone made a port for any driver to Sparc, the same techniques could be used to port all kinds of PCI cards.
well, I wish it was that simple...
but its not quite that easy.
The poster is right on the fact that it requires some functions from the bios, but its not something that can easily be simulated in a driver.
Unless you want to use it as a secondary display card designated strictly to do 3d work, then you would need to have the card itself suport the platform, at a point in the boot process BEFORE linux or solaris or any other OS loads..
What would be required would be an OpenFirmware bios for the card itself, in order to have the hardware suport it..
One possible way, which could be vbery much worth looking into, is cards made for Macs, as almost all PowerMacs use OpenFirmware, so it should theoratically be possible to go and use a mac card, and simply use the already-available Linux drivers for it.
But other than either using a Mac card, it would be very dificult to getting it to work on any sparc system, unless someones got enough time on their hands to port the BIOS of the card itself to Openfirmware...
(which has been done before, there were 2 attempts at porting the nVidia TNT and TNT2 chipset's BIOSes to trhe Mac, but they were never stable enough, untill after 3DFX had released the mac drivers for their cards, making it more or less a moot point)
Just give it a higher priority, not the highest priority. ;)
I made the mistake one time in class to set a counter program into realtime in NT, not even the mouse could get enough of a timeslice to register.
the counter sure as hell was speeding along tho
just make sure to set it to high, or +1 or +2, and not realtime
(not sure what priority levels are available in Win9x)
"I think one of the most egregious parts of the new APSL is the clause preventing you from making modifications for your own use without sending them in to Apple"
Not to nitpick, but have you ever actually READ the version 1.2 of the APSL? in it it clearly states that its free to use as you see fit for personal use, or internally for R&D purposes WITHOUT notifying apple of your modifications if you wanna take the code, and hack it around to have darwin working on your power mac 6100*, go ahead, the code is there, tweak and hack away. if you wanna make it work for one of the Motorola PowerMax 6000**'s you got off of ebay last year, go right ahead.
As long as your not putting it in a commercial production environment, or selling a product based on the code and your modifications to it, Apple doesnt give a rats ass about it.
no, its not as open and free as the BSD license. but remember, you said you want the code to hack around, to change things in your environment...
GO RIGHT AHEAD
the APSL isnt stopping you. do as you see fit with the code, provided you dont release it in a production or commercial environment.
and if you were to do so, all you have to do is send your mods over to Apple, and if they think its something worthwhile including in the main darwin product, then they will do so. Just send them notification.
and if you were to use linux for your project which you were selling, remember, you have to give your source code away anyway.
the only diference is that your sending your changes to one entity under the APSL, and to any joe hacker schmoe that wants to look at your code under the GPL.
*The powermac 6100 is the first powerpc-based mac, but since its based on the NuBus architecture, its pretty much incompatibvle with darwin, OSX, and pretty much anything except MKLinux
**The PowerMax 6000 was a CHRP (common Hardware Reference Platform) machine, the first chrp machine introduced to the general public (Macworld Boston 97, I had the pleasure of playing with one at the floor) but was only made as a few hundred prototypes, and killed shortly thereafter after Apple killed cloning and refused to allow motorola to sell them
and completely OT--- why does the Extrans mode (allowing HTML tags to the text) NEVER work?
it might be possible to do what you are thinking by setting up a few small devices around the yard that would emit a radio signal, thereby telling the robot its relative position in the yard. :P
i know there are systems for yards with a similar setup. although the sgnals are used not to control or give a position, but to keep a dog within the specified area (it will shock the dog if it tries to leave)
Farallon made Timbuktu Pro for the Macs, and as of version 2.x, they also made it for windows, thereby enabling what i believe is for the first time cross-platform remote administration (this was back in early 90's)
Very handy tool, and one of the only ones available for a mac for a long time
Im not trying to say that they dont work, i had NetBSd running on my centris at one point in time, but the simple fact that they are so intenet on porting it to everything imaginable "Of Course it runs NetBSD" naturally means that a lot of new advances dont get put in because resources are more or less spread out along an insane number of platforms that is being suported..
No, I'm not trying to troll here, but is it truly possible to avoid having the linux kernel end up forking somehwhere along the line?
Right now you have linux running on a power mac, on an ultra sparc, on a 386, on an SGI, on a cisco router, on IBM mainframes, on tiny embedded projects...
You have one kernel, yet everyone is trying to push it every which way imaginable..
If you try to make the code trully able to take advantage of big iron, you make it so it cant really run on a 386. You keep it optimized for your 386, you wouldnt consider putting it anywhere near your mainframe without major code changes, ala what IBM is doing.
yet by keeping these things away from the kernel, it will basically force forking in the not too distant future...
i guess Linus might be tryin to avoid a situation like NetBSD that runs on everything imaginable, but doesnt really doa good job on any.,.
but if he is, with all the diferent directions Linux is headed, forking is just inevitable.
Actually, the RAID included in the Abit boards and almost all other IDE RAID boards are BIOS-based solutions..
they are still softare, just at a lower level (BIOS vs. Driver) but as far as true hardware raid, only one company makes a microcontroller-based IDE RAID card, which, at the price of high end adaptec raid cards, isnt too lovely an idea...
OMFG, i havent laughed this much in such a long time...
thank you
that was just perfect
:')
ROTFLMAO
please mod this up, it deserves at least a 5 funny...
But then again, you also have to take into account that each of those 10 G4 macs also have 10 300W power supplies, each generating its own ammount of heat, whereas one or 2 300w PS could take care of thsoe 10 G4 drive bays. And you also have several other components making heat, and wasting more electricity...
and getting back to the point of my original post, you also have to remember, even if it is more heat concentrated in that one cabinet with a couple of dozen of these drive bay computers, its still gonna be cheaper to cool that 1 cabinet then it is gonna to cool off an entire large room of huge computers.
Damn, thats gotta be the lowest UID ive ever seen around here on
damn.
Your old school
</OT>
but getting back to topic, the added benefit that this also provides, versus a completely passive backplane solution, is that while you can dedicate a rack case to a backplane with a few of the controller cards, with these, you can also put a little server in the spare drive bays of your other servers, and anywhere that it might fit, since it isnt requiring its own case like a backplane solution.
Considering the lovely condition of state poewr problems here in california, this could be a great solution if implemented.
;-)
Since these machines are small, and are more or less fuill-featured computers, this could make building a server farm a lot cheaper and less power hungry, since one could have each of these set up as its own server, feeding off of a single rackmount cases's power supply, instead of having each server having its own oversized powersupply, as seen in SO many installations.
and at this size and power usage, it would also cut down on AC costs dramatically, as you can now fit several dozen computers in a space that you could possibly fit only maybe 10, therefore reducing the necessary cooling costs.
--warning--beowulf comment follows---
Now, where can i get a beowulf cluster of these?
Does this mean that all the spam i sign up for with my "other" e-mail billg@microsoft.com is gonna bounce back?
:(
how sad
Ive seen the M$ optical trackballs around before, but one thing thats always come to my mind, and still does:
is there ANY benefit whatsoever from having a trackball be optcall, since its still relying on a ball?
i thought the major benefir of going towards the optical neutered mice was that you didnt have the probs associated with its one-balled brethren, the dust and all...
with a trackball, that is kinda pointless...
even the name implies a ball. =P
or is this simply a marketing tactic for M$, to use its latest technology and marketing gimicks to a market that it has no point being in?
while i agree that Intel chipsets do have SOME advantages, its kind of questionable the way you phrased your post, considering that it reads exactly like an intel PR ad.
The true benefits of intel chip[sets, are, in fact, the integration that they have.
by throwing toguether really crappy on-board video and sound, and maybe a NIC too, they are able tomake really cheap motherboards, while cutting the cost of a video card out the door for the system builder..
allowing them to either make more profits or pass the savings on to the consumer.
That is the reason why the duron, even as cheap and superior as it is, hasnt found much of a home, since the motherboards available to suport it are still so expensive..
and the one area where intel chipsets ruled above VIA for a LONG time, and still slightly do to this day, is memory bandwidth, which has tradtionally been somewhat of a strong point for intels chips..
And as far as the hub architecture goes, there are 2 main diferences between that and the northbridge/southbridge concept:
ONe: theres a special bus now connecting the 2 parts of the chipset toguether instead of the PCI bus
two: Intel gave it a diferent name.
the idea is still the same, however.
If you have some time to spare, and some are in the mood of a nice decent project, you might want to look into building your own box and putting it in the car, powering all the media you want, with a keypad and a small lcd display... /AM/FM and whoknows whtelse im in the mood for) is the InMotion car mp3 player
I have seen several of sites come and go, but the one with the more useful information, and the one im basing my system on the most (My 87 Plymouth Sundance has no sound right now, so im building a small box to handle CDs/MP3s
http://www.jarcom.com/inmotion/
the site hasnt been updated since early summer, but the setup and schematics are still valid.. definately worth a look.
Also, if anyone else has done this, are there any sugestions/things to avoid in doing this?
-Joel
Well its probably cause every piece of "first evidence" got sucked into the black hole they were studying
;-)
or maybe it just found itself into earth's own little black hole, my closet
:P
What do you see in the future for *BSD, with the huge amount of popularity that linux keeps on receiving, not to mention attention, esp. from our buddy Bill Gate$...
Do you think it will remain the strong, viable but simply less popular free OS it is now, hiding behind the limelight of linux, or will it come up in popularity, esp with the codebase for Apple's Darwin, which is all BSD based?
Sorry to nitpick, but the first power-pc based mac was released on may 1994, and that was the Power Mac 6100, which was soon followed by the 71 and 8100 machines...
it was still a few more years till the 9500/120 came out (late 95)..
so i dont think your 9500/200 is QUITE that old yet
:P
Im not familiar with the palm software, as I havent had the pleasure of owning one. (but am willing to try, free donations accepted ;-)
But wouldnt it be possible to run 2 copies of the sopftware, installed at diferent locations, and having them both run simultaneously, as diferent processes, each one with its own configuration settings?
like i said, im not familiar with it, but its worth a shot
g'luck
One detail I meant to add, With a new BIOS revision, most motherboards should work perfectly with this chip, just not out of the box, not for a while..
which will definately make it even less atractive for the do-it-yourselfer then it already is, but it should still be a cheap, atractive option for low-powered notebooks and subnotebooks, and the appliance market.
One important thing that most people have forgotten with the new CYrix's, you cant just pop them into any motherboard made for a P3 or celeron.
Well, technially you could, but one thing that has to be present in a motherboard for it to start up, is the processor's microcode in the BIOS.
that is why new BIOS revisions are always coming out saying "now supports this speed celeron or P3" or whatnot. Its not because the motherboard was previously incapable of handling it, but there was no micorocode instructions stored in the BIOS for that revision of the CPU.
Which is exactly why some motherboards needed new BIOS revisions to suport the P3 when it came out, or the CuMines..
they were mainly the same chip, but with diferent microcode, so it needed those instructions before it could properly POST.
Would you have access to a net connection and a burner?(long with a pile of blanks?) ;-)
If so, one cd, a net connection, and you can get access to everything else you might need, just download and burn
just a thought
Are you sure you wanna use tires as a comparison, with all the firestone stuff going on?
;-)
just a thought.
(but good point tho)
I can see one reason why you would want to do this, although i have no idea if that is why he would need to pull it off a DB..
in case there are several machines running apache, and any settings have to be changed, if they can collect those settings from a datavase automatically once the setting is changed in the database, it would be a lot easier then going through each one and running the script, and making sure all changes were applied.
just an idea..