Where does it mention any speeds??? It talks about accelerations!
If the impact only lasts for 1 msec and in this time it goes from 8.33 m/s to 0 m/s you already have your 850G. In normal gravity it picks up this speed in less than a second. So, pretty good for normal handling accidents (dropping a notebook on a carpet floor) but easy to exceed by throwing it out of a window on a concrete floor.
I bought a centrino notebook last year only because Intel promised there would be support "soon". One year later, there still is nothing working available. Does writing a driver take that long for them or were they simply lying back then?
I also bought a 54bit PCMCIA card in the meantime. In at most another year this notebook will be obsolete anyway and guess what I will NOT buy next...
Last march I bought a Centrino notebook, only because chipset support seemed to be there, powerstep looked about to be implemented AND wireless LAN drivers were promised to be released very soon by Intel. Now, almost a year later, still no wireless driver and they still say "real soon now". I guess I am too gullible. Oh, I also believed them that their crap keeps cool. Even at 600MHz (instead of 1300) and doing nothing this thing gets freaking hot and makes lots of noise. I am MUCH happier with my Crusoe (Toshiba Libretto) notebook. I guess my next one will be an Efficeon.
That's what I am doing. I did not buy Intel desktop CPUs/boards for years now. Don't believe the myths regarding instabiliy or heat. They are not true. And the price is much lower at same performance. Last year I made the mistake to buy a Centrino notebook. Still waiting for the promised "soon to come" WLan support. And this thing gets really hot doing nothing. Yes, cpufreq and cpudynd are running and it helped a little but the cooler still turns on a lot (still when doing nothing). So, compared to my Transmeta notebooks, it gets freaking hot. Since AMD notebooks are not any better in his regard, my next notebook will have an Efficeon.
There is no Drunken Master 2 release available in Europe. I do not want the changed US version. HkFlix seems to sell a good Asian version but for a rather steep price (+ transport + import tax). And maybe HKFlix will not be able to sell it much longer now? A German importer offered it a few months ago but when I ordered it he said his Asian distributor suddenly no longer had it?!?!
Btw, there is a great Drunken Master release availabe in Europe from www.hongkonglegends.co.uk.
She is not against patents. She pretends not to propose US-like software patents and that this is the reason she is for the new law. Actually, by doing this, software patents will be legalized while they are not actually allowed right now.
Somebody in the Linux DVB group was so kind to point out yesterday that there is a TV channel with live feeds of such discussions on http://europa.eu.int/comm/ebs/reception_en.html. I watched the discussion on software patents this morning. I have to say that without knowing the real backgrounds you could really fall for the arguments of the proponents. No, they do not want laws like the US, they want to restrict software patents. I don't know how they can keep straight faces saying that. Luckily, some other speakers did not fail to mention that software patents are actually against the law and the new proposal would legalize those patents. Restrictions to the new law (like those discussed on Slashdot yesterday) are not needed. Simply do not pass the new law and enforce the existing ones.
The so-called harrassments by citizens also were described quite differently depending on the side the speaker was on. The speakers for patents pretended to be offended by such wild actions. The speakers against rather described them as reasonable concerns by middle sized businesses who fear for their survival.
"Linus aeh Linux" was also mentioned. But one speaker talked about "Unix, which is free"?! Hmmm, let's not let Darl hear about this.
Nokia has been bulding TVs and (first analog and then) digital satellite receivers for a LONG time. I think this even started before their phone business.
The Linux DVB API was developed by them about 3 years ago and they funded implementations of it for the most popular cards here in Europe. The API (after some changes) basically is still in use and part of the 2.6.x kernel.
This is also not the first digital PVR box developed by Nokia. Not all of them made it to the market but at least one is available here for months now.
No, they only sell the rights to use it in binary form. So, they are not releasing their IP under the GPL, but they are explicitly infringing on the copyrights of all other Linux contributors. Can I sue them for this here in Germany?
On the other hand, since here is no SCO IP in Linux which does not already fall under the GPL (as contributed by the old SCO/Caldera), they could later argue they did not violate the GPL or other copyrights but rather sold those suckers hot air. After all, they sell you the rights to use "the SCO IP" in binary form. If there isn't any, you just paid them for nothing but the GPL was no violated. This could of course still be called fraud.
I would not generalize this that much. The same German law which this time was used exactly as it is intended to also is abused very often. There are lawyers who are specialized in this and just try to contruct false cases of unfair competition to collect fees. The case against SCO is one of the few instances which make me believe there is some good in this law.
Some others already commented on the situation with lacking Centrino (power saving and WLAN) support. I did not even get any answer from their "customer support". I'll only buy Transmeta in the future. I just got a Libretto L5 and it is MUCh better regarding heat and noise (none besides the harddisk).
But I also still have the Centrino notebook (Samsung X10). Does anybody know if there are Linux supported mini-PCI WLAN cards which I could use instead of the Intel card?
Hmm, you are right. The VSYNC is a problem. I just looked into the specs again. It says the frame capture region can be extended by setting the EXT_FRAME register bit and setting VDELAY to the minimum value of 2. This implies that there still is a delay between VSYNC and start of capturing.
I only used the mode as an oscilloscope for TV signals. It is very useful to look at special signals in the TV signal. One could e.g. write a software PAL-Plus decoder (for those who do not know this, thats not a scrambling system but kind of an enhanced widescreen format).
For other purposes your idea with the PIC sounds good. But one will probably lose some data during and after each VSYNC.
The Bt848/878 has a raw mode which delivers the pure composite signal. I never tried to use it with non-TV signals but I guess it could work.In this raw mode the chip does not care about horizontal and vertical syncs. You can see all the "good stuff": color burst, front porch, strange "signal modifications" in scrambled channels, etc. When I programmed the Linux drivers I experimented with this mode a little bit.
This does not deliver great resolution (only 8 bit) but goes up to at least 36MHz (PAL frq.*8). With the chips with internal PLL (Bt848A and upwards) you can even change the sampling frequency in small steps.
You can find the Bt848/878 in most cheap TV cards on the market but you might have to make some modifications to the Linux driver to support
the raw mode and non-standard PLL settings.
I see that many people here are complaining about the capabilities and price of the box. For me it is just the kind of STB I was looking for.
Sure, you can build a PC with MPEG2 playback, networking etc. cheaper or for about the same price. But thats not what this device tries to replace, at least I would not buy it for doing this. When I saw the older TV Bokks (you can see it on their web page) I thought this would be great as an STB to stream files and live DVB streams from a server in the basement to other TVs here at home. But, although we have one of those at work to play with, I never saw it being offered for sale anywhere.
The old Bokks has a Geode chip as processor/graphics chip/etc. and a Sigma Designs EM8400 for decoding. I guess the new Bokks is based on the same design. They only added some hardware (like the hard drive) and changed the shape to a brick. The EM8400 is a nice decoder chip. We developed our own driver which is API compatible to our DVB drivers. So, I would not mind playing with one of those Bokkses. Also, if you consider the price of an EM8400 card (e.g. the Netstream 2000), the Bokks does not seem that expensive anymore.
Sure, I would also not mind if the Bokks were a little cheaper. With VAT (AFAIK 25% in Sweden which I guess I will have to pay when ordering from Germany) its over 1100 German Marks + shipping. Maybe they decide to also sell the old version for a lower price. I at least would be interested in it. On the other hand, the disk and nice brick shape makes the new system really portable. And I think the specs mention only 10W power usage. No noise, little heat, sounds good.
Now please everybody buy it so that the price goes down and I can get one:-)
First of all, I was not looking through the 2.3 kernel sources. I was looking through the Nvidia kernel driver sources (more specifically the file nv.c) to see if they could be adapted to work with 2.3 kernels because I could not compile them on my setup (2.3.99-pre6).
And am I happy with them removing the code as soon as possible? Well, of course I am happy with that compared to them not doing anything at all about it. But it does not mean I would not consider taking other steps if "as soon as possible" turns out to be very long or never. An INSTANT SOLUTION for Nvidia would be to take code out of their driver RIGHT NOW and provide a (GPLed) kernel patch with their drivers to include the code in the kernels memory management functions. This code, or at least something with the same functionality, belongs there anyway. Several (GPLed) drivers already use them and have an unnecessary extra copy in their code.
Ohh, and since I think it also was not mentioned before what code we are talking about. Its about the memory pointer conversion routines uvirt_to_pa() up to nv_vfree(). While those are hard to implement any other way, Nvidia did not need to take them line for line with almost the same function names and even including the original comments out of a GPLed file.
Slashdot Mirror
Where does it mention any speeds???
It talks about accelerations!
If the impact only lasts for 1 msec and in this time it goes from 8.33 m/s to 0 m/s you already have your 850G. In normal gravity it picks up this speed in less than a second. So, pretty good for normal handling accidents (dropping a notebook on a carpet floor) but easy to exceed by throwing it out of a window on a concrete floor.
Correct, they are one year too late.
...
I bought a centrino notebook last year only because Intel promised there would be support "soon". One year later, there still is nothing working available. Does writing a driver take that long for them or were they simply lying back then?
I also bought a 54bit PCMCIA card in the meantime. In at most another year this notebook will be obsolete anyway and guess what I will NOT buy next
Last march I bought a Centrino notebook, only because chipset support seemed to be there, powerstep looked about to be implemented AND wireless LAN drivers were promised to be released very soon by Intel. Now, almost a year later, still no wireless driver and they still say "real soon now". I guess I am too gullible.
Oh, I also believed them that their crap keeps cool. Even at 600MHz (instead of 1300) and doing nothing this thing gets freaking hot and makes lots of noise.
I am MUCH happier with my Crusoe (Toshiba Libretto) notebook. I guess my next one will be an Efficeon.
That's what I am doing.
I did not buy Intel desktop CPUs/boards for years now. Don't believe the myths regarding instabiliy or heat. They are not true. And the price is much lower at same performance.
Last year I made the mistake to buy a Centrino notebook. Still waiting for the promised "soon to come" WLan support. And this thing gets really hot doing nothing. Yes, cpufreq and cpudynd are running and it helped a little but the cooler still turns on a lot (still when doing nothing). So, compared to my Transmeta notebooks, it gets freaking hot. Since AMD notebooks are not any better in his regard, my next notebook will have an Efficeon.
There is no Drunken Master 2 release available in Europe. I do not want the changed US version. HkFlix seems to sell a good Asian version but for a rather steep price (+ transport + import tax). And maybe HKFlix will not be able to sell it much longer now?
A German importer offered it a few months ago but when I ordered it he said his Asian distributor suddenly no longer had it?!?!
Btw, there is a great Drunken Master release availabe in Europe from www.hongkonglegends.co.uk.
Wow, people really fall for her double talk ...
She is not against patents. She pretends not to propose US-like software patents and that this is the reason she is for the new law. Actually, by doing this, software patents will be legalized while they are not actually allowed right now.
Somebody in the Linux DVB group was so kind to point out yesterday that there is a TV channel with live feeds of such discussions on http://europa.eu.int/comm/ebs/reception_en.html.
I watched the discussion on software patents this morning. I have to say that without knowing the real backgrounds you could really fall for the arguments of the proponents. No, they do not want laws like the US, they want to restrict software patents. I don't know how they can keep straight faces saying that. Luckily, some other speakers did not fail to mention that software patents are actually against the law and the new proposal would legalize those patents. Restrictions to the new law (like those discussed on Slashdot yesterday) are not needed. Simply do not pass the new law and enforce the existing ones.
The so-called harrassments by citizens also were described quite differently depending on the side the speaker was on. The speakers for patents pretended to be offended by such wild actions.
The speakers against rather described them as reasonable concerns by middle sized businesses who fear for their survival.
"Linus aeh Linux" was also mentioned. But one speaker talked about "Unix, which is free"?! Hmmm, let's not let Darl hear about this.
Get a clue.
Nokia has been bulding TVs and (first analog and then) digital satellite receivers for a LONG time. I think this even started before their phone business.
The Linux DVB API was developed by them about 3 years ago and they funded implementations of it for the most popular cards here in Europe.
The API (after some changes) basically is still in use and part of the 2.6.x kernel.
This is also not the first digital PVR box developed by Nokia. Not all of them made it to the market but at least one is available here for months now.
No, they only sell the rights to use it in binary form. So, they are not releasing their IP under the GPL, but they are explicitly infringing on the copyrights of all other Linux contributors. Can I sue them for this here in Germany?
On the other hand, since here is no SCO IP in Linux which does not already fall under the GPL (as contributed by the old SCO/Caldera), they could later argue they did not violate the GPL or other copyrights but rather sold those suckers hot air. After all, they sell you the rights to use "the SCO IP" in binary form. If there isn't any, you just paid them for nothing but the GPL was no violated. This could of course still be called fraud.
I would not generalize this that much.
The same German law which this time was used exactly as it is intended to also is abused very often.
There are lawyers who are specialized in this and just try to contruct false cases of unfair competition to collect fees. The case against SCO is one of the few instances which make me believe there is some good in this law.
Some others already commented on the situation with lacking Centrino (power saving and WLAN) support. I did not even get any answer from their "customer support". I'll only buy Transmeta in the future. I just got a Libretto L5 and it is MUCh better regarding heat and noise (none besides the harddisk).
But I also still have the Centrino notebook (Samsung X10). Does anybody know if there are Linux supported mini-PCI WLAN cards which I could use instead of the Intel card?
An they still list Convergence Integrated Media which, AFAIK, no longer exists.
Hmm, you are right. The VSYNC is a problem. I just looked into the specs again. It says the frame capture region can be extended by setting the EXT_FRAME register bit and setting VDELAY to the minimum value of 2. This implies that there still is a delay between VSYNC and start of capturing.
I only used the mode as an oscilloscope for TV signals. It is very useful to look at special signals in the TV signal. One could e.g. write a software PAL-Plus decoder (for those who do not know this, thats not a scrambling system but kind of an enhanced widescreen format).
For other purposes your idea with the PIC sounds good. But one will probably lose some data during and after each VSYNC.
The Bt848/878 has a raw mode which delivers the pure composite signal. I never tried to use it with non-TV signals but I guess it could work.In this raw mode the chip does not care about horizontal and vertical syncs. You can see all the "good stuff": color burst, front porch, strange "signal modifications" in scrambled channels, etc. When I programmed the Linux drivers I experimented with this mode a little bit.
This does not deliver great resolution (only 8 bit) but goes up to at least 36MHz (PAL frq.*8). With the chips with internal PLL (Bt848A and upwards) you can even change the sampling frequency in small steps.
You can find the Bt848/878 in most cheap TV cards on the market but you might have to make some modifications to the Linux driver to support
the raw mode and non-standard PLL settings.
I see that many people here are complaining about the capabilities and price of the box. For me it is just the kind of STB I was looking for.
:-)
Sure, you can build a PC with MPEG2 playback, networking etc. cheaper or for about the same price. But thats not what this device tries to replace, at least I would not buy it for doing this. When I saw the older TV Bokks (you can see it on their web page) I thought this would be great as an STB to stream files and live DVB streams from a server in the basement to other TVs here at home. But, although we have one of those at work to play with, I never saw it being offered for sale anywhere.
The old Bokks has a Geode chip as processor/graphics chip/etc. and a Sigma Designs EM8400 for decoding. I guess the new Bokks is based on the same design. They only added some hardware (like the hard drive) and changed the shape to a brick. The EM8400 is a nice decoder chip. We developed our own driver which is API compatible to our DVB drivers. So, I would not mind playing with one of those Bokkses. Also, if you consider the price of an EM8400 card (e.g. the Netstream 2000), the Bokks does not seem that expensive anymore.
Sure, I would also not mind if the Bokks were a little cheaper. With VAT (AFAIK 25% in Sweden which I guess I will have to pay when ordering from Germany) its over 1100 German Marks + shipping. Maybe they decide to also sell the old version for a lower price. I at least would be interested in it. On the other hand, the disk and nice brick shape makes the new system really portable. And I think the specs mention only 10W power usage. No noise, little heat, sounds good.
Now please everybody buy it so that the price goes down and I can get one
I want to clear up 2 points made in the article.
First of all, I was not looking through the 2.3 kernel sources. I was looking through the Nvidia
kernel driver sources (more specifically the file nv.c) to see if they could be adapted to work with 2.3 kernels because I could not compile them on my setup (2.3.99-pre6).
And am I happy with them removing the code as soon as possible?
Well, of course I am happy with that compared to them not doing anything at all about it.
But it does not mean I would not consider taking other steps if "as soon as possible" turns out to be very long or never.
An INSTANT SOLUTION for Nvidia would be to take code out of their driver RIGHT NOW and provide a (GPLed) kernel patch with their drivers to include the code in the kernels memory management functions.
This code, or at least something with the same functionality, belongs there anyway. Several (GPLed) drivers already use them and have an unnecessary extra copy in their code.
Ohh, and since I think it also was not mentioned before what code we are talking about.
Its about the memory pointer conversion routines
uvirt_to_pa() up to nv_vfree().
While those are hard to implement any other way, Nvidia did not need to take them line for line with almost the same function names and even including the original comments out of a GPLed file.
Ralph Metzler