It's a fair question. There is obviously some competion from MySQL and PostgreSQ since the people using them are not paying CAL's (Client Access Licenses) to Oracle. What Oracle can charge for is where their code gives value over MySQL and PostgreSQL. Now, the databases do not live in an isolated world. Integration with other software is important and we see a lot of that in OSS. PHP works like out of the box with MySQL and PostgresSQL. With Oracle and Informix, there some work to do before you get it to run. Zope, python and mysql, I haven't tried but I guess it's a nobrainer.
If this integration in OSS continues, products like Oracle and Informix will find it harder to provide added value since integration with other products will eat in to that and the OSS competition will seem even stronger.
So, the difference between OSS:ing and simply adapting pricing to competion is to get into the OSS integration framework. If developers start to think equaly about PHP+MySQL as PHP+Oracle, more people will knock on Oracle door to ask how to solve various database related problems. This is allready a large revenue stream for Oracle (I guess) and the question companies like Oracle have to answer is wheather to climb onto the chimney of what _may_ be a sinking boat or redefine their business into service and consulting and try to make a living out of that.
With negative index of refraction you are able to recreate the source field at some other "image" position. That means no resolution limits set by the wavelength of the light and and focusing power (usually called numerical aperture) set by the lens. Note that the negative refraction lens conjugates the phase of the wave. For a plane wave, it just looks like the wave reverses direction at the other side of the (flat) lens. For a spherical wave originating at some point, the conjugated wave will focus onto a point on the other side of the (negative refraction) lens.
For antenna research this will probably mean really a lot. Most probably we will also be able to locate sources of (microwave) radiation with great precision.
I have been runing Xfree from cvs for a couple of days now and it actually works better with 4.0.3 than 4.0.2, at least for me (on a GeForce256 DDR, Dual PIII). In 4.0.2 the Xserver died with strange error message. For the first time AGP works as well. I'd say. Download!/jarek
Cellular techniques provide essentially what we want in terms om BW/unit area. This is a cost issue, just how close are we prepared to place base stations to each other. Actually, this can be made cost effective by using multiple beam antenas which then essentially create smaller cells. Add to that a some computing power and you can add interference reducing techniques (now problably mostly used in millitary systems), although such techniques can be used with simple antenas as well. This enables more efficient frequency reuse patterns in cellular networks as well. My quess that smart antenna (beamforming down link and interferance rejection up-link) will be the future. Systems like that are being employed today./jarek
On x86 hardware the linux kernel enters 100 times per second. We have direct access to the hardware and can set up DMA for data and commands to the card. The time spend in the kernel is negligible. Context switching is of the order of micro second(s) for PIII's.
My guess would be that, at best, BE somehow can get out of the way for smart optimizations. They mention SIMD use in PIII but I don't understand why that wouldn't be used in linux or windows to the same extent.
Now this is getting strange. I this story is true, then Nvidia is at the moment distributing GPL'd code without source. If they at least removed the code from their site when they realized they were in violation of GPL, then this could be delt with as an honest mistake (at least if the developer in question agrees with it). But now, Nvidia is chalenging the GPL license. That has to stop.
Funny, I managed to get the binary release, which turned out to be for 2.2.12-20 (ancient) kernel.
Then I went for the src.rpm and realized that the story by now was probably on slashdot, which turned out to be correct. Oh well, I'll get it tomorrow:)
As far as I remember, getting 2.4 to run with say RH 6.1 is no problem what so ever. Just make dep... and install with your favorite method. A few apps will be hit with quite substantial modifications of the layout in/proc but usually that will not prevent you from getting the system up and running. Changing to 2.4 from 2.2 is much easier compared to the transition from 2.0 to 2.2. If you want to try out devfs, read the docs and install devfsd.
In what way will devfs remove the need for the major/minor scheme? As far as I know, devfs only provides a major cleanup of the namespace bloat in/dev. Otherwise, drivers are still referenced by major and minor numbers through the vfs layer.
Well, maybe because clock rates is not the whole truth. Maybe because if we focus on clock rates companies won't go for the other architectures (compare risc and cisc) being afraid they loose the marketing war over clock frequencies. Focusing on the wrong parameter may not get you where you want.
Yes, Pro/E would be nice indeed. Assuming that a port would bring along Mech/structure, thermal and the other add-ons. Imagine that on a DS-20. A real killer.
Don't agree. The glx-driver found on their ftp site performs very nicely with q3test (and my PII 450:) and has not caused a single problem during the 2 months I've used it./jarek
I believe you are correct. Making things smaller teaches us more about how to controll the quantum mechanical effects on a technological level. Put another way, More's law will make a quantum leap as we, using computers, will be able to simulate physical phenomena in linear time (or linear complexity).
but then again, you have to admitt that transmeta is in a place where investors are looking for projects while E2K (if it's for real, I don't know) is on the other side of the earth however you look at it. On a general note, having met a number of rusian scientists I dare to say that rusian science deserves A LOT of respect. They have great tradition in science and education. Don't put them down because the political system during the post war era has destroyed the country. There is nothing wrong with the people.
We have a number of people that need to access Pro/E to a varying degree. A couple of network licenses would solve that just fine. Imagine a Pro/E "Server" on an alpha and a nubmer of Pro/E GLX clients (preferably Intel). This could become a reality in the near future.
If I remember correctly, 2.1 is compatible with 2.0 unless the program depends on specific symbols in the library (something I red on the gnu libc web page although I can't find it now).
It should be on http://www.gnu.org/software/libc/libc.html but...
That, ACL's and volume management (at least I believe that is a kernel issue) is what I need. Esp. ACL's. Unfortunately there seems to be litle interest in ACL's on the kernel mailing list. Anybody knows if there is a ACL project going on. The page most people were pointing to doesn't seem to exist any more.
Slashdot Mirror
According to Steve Lord, one of the XFS developers, current xfs from cvs has problems. (link to Steves post on linux-xfs) /jarek
It's a fair question. There is obviously some competion from MySQL and PostgreSQ since the people using them are not paying CAL's (Client Access Licenses) to Oracle. What Oracle can charge for is where their code gives value over MySQL and PostgreSQL. Now, the databases do not live in an isolated world. Integration with other software is important and we see a lot of that in OSS. PHP works like out of the box with MySQL and PostgresSQL. With Oracle and Informix, there some work to do before you get it to run. Zope, python and mysql, I haven't tried but I guess it's a nobrainer.
If this integration in OSS continues, products like Oracle and Informix will find it harder to provide added value since integration with other products will eat in to that and the OSS competition will seem even stronger.
So, the difference between OSS:ing and simply adapting pricing to competion is to get into the OSS integration framework. If developers start to think equaly about PHP+MySQL as PHP+Oracle, more people will knock on Oracle door to ask how to solve various database related problems. This is allready a large revenue stream for Oracle (I guess) and the question companies like Oracle have to answer is wheather to climb onto the chimney of what _may_ be a sinking boat or redefine their business into service and consulting and try to make a living out of that.
Just a thought
/jarek
With negative index of refraction you are able to recreate the source field at some other "image" position. That means no resolution limits set by the wavelength of the light and and focusing power (usually called numerical aperture) set by the lens. Note that the negative refraction lens conjugates the phase of the wave. For a plane wave, it just looks like the wave reverses direction at the other side of the (flat) lens. For a spherical wave originating at some point, the conjugated wave will focus onto a point on the other side of the (negative refraction) lens.
For antenna research this will probably mean really a lot. Most probably we will also be able to locate sources of (microwave) radiation with great precision.
/jarek
I have been runing Xfree from cvs for a couple of days now and it actually works better with 4.0.3 than 4.0.2, at least for me (on a GeForce256 DDR, Dual PIII). In 4.0.2 the Xserver died with strange error message. For the first time AGP works as well. I'd say. Download! /jarek
Well, I read kernel traffic and kernel mailing list but somehow this escaped me. Thanks slashdot. /jarek
great stuff btw
because he thinks so too, that reiserfs might go in when the 2.4.x is stable.
/jarek
Cellular techniques provide essentially what we want in terms om BW/unit area. This is a cost issue, just how close are we prepared to place base stations to each other. Actually, this can be made cost effective by using multiple beam antenas which then essentially create smaller cells. Add to that a some computing power and you can add interference reducing techniques (now problably mostly used in millitary systems), although such techniques can be used with simple antenas as well. This enables more efficient frequency reuse patterns in cellular networks as well. My quess that smart antenna (beamforming down link and interferance rejection up-link) will be the future. Systems like that are being employed today. /jarek
On x86 hardware the linux kernel enters 100 times per second. We have direct access to the hardware and can set up DMA for data and commands to the card. The time spend in the kernel is negligible. Context switching is of the order of micro second(s) for PIII's.
My guess would be that, at best, BE somehow can get out of the way for smart optimizations. They mention SIMD use in PIII but I don't understand why that wouldn't be used in linux or windows to the same extent.
Now this is getting strange. I this story is true, then Nvidia is at the moment distributing GPL'd code without source. If they at least removed the code from their site when they realized they were in violation of GPL, then this could be delt with as an honest mistake (at least if the developer in question agrees with it). But now, Nvidia is chalenging the GPL license. That has to stop.
/jarek
Funny, I managed to get the binary release, which turned out to be for 2.2.12-20 (ancient) kernel. :)
Then I went for the src.rpm and realized that the story by now was probably on slashdot, which turned out to be correct. Oh well, I'll get it tomorrow
As far as I remember, getting 2.4 to run with say RH 6.1 is no problem what so ever. Just make dep ... and install with your favorite method. A few apps will be hit with quite substantial modifications of the layout in /proc but usually that will not prevent you from getting the system up and running.
Changing to 2.4 from 2.2 is much easier compared to the transition from 2.0 to 2.2.
If you want to try out devfs, read the docs and install devfsd.
In what way will devfs remove the need for the major/minor scheme? As far as I know, devfs only provides a major cleanup of the namespace bloat in /dev. Otherwise, drivers are still referenced by major and minor numbers through the vfs layer.
For the last 10 years, I've probably seen a unix reboot less than I've seen NT 4.0 reboot the last week. /jarek
Alan certainly has relevant opinions about that part of the kernel but I believe Dave Miller is the one who wrote most of the new NET-4 code.
/jarek
Well, maybe because clock rates is not the whole truth. Maybe because if we focus on clock rates companies won't go for the other architectures (compare risc and cisc) being afraid they loose the marketing war over clock frequencies. Focusing on the wrong parameter may not get you where you want.
/jarek
.. linux is really becoming __main stream__
:)
get it
/jarek
this Daemia girl is really getting to me, and Stripe doesn't help either :)
Maybe I finally can hurt them plenty..
Happy fragging
/jarek
Yes, Pro/E would be nice indeed. Assuming that a port would bring along Mech/structure, thermal and the other add-ons. Imagine that on a DS-20. A real killer.
/jarek
Don't agree. The glx-driver found on their ftp site performs very nicely with q3test (and my PII 450 :) and has not caused a single problem during the 2 months I've used it. /jarek
I believe you are correct. Making things smaller teaches us more about how to controll the quantum mechanical effects on a technological level.
Put another way, More's law will make a quantum leap as we, using computers, will be able to simulate physical phenomena in linear time (or linear complexity).
/jarek
but then again, you have to admitt that transmeta is in a place where investors are looking for projects while E2K (if it's for real, I don't know) is on the other side of the earth however you look at it.
On a general note, having met a number of rusian scientists I dare to say that rusian science deserves A LOT of respect. They have great tradition in science and education. Don't put them down because the political system during the post war era has destroyed the country. There is nothing wrong with the people.
Cheers,
/jarek
damage you image on the net, I can't think of any better way to do it. Sue a free developen and parish the rest of you (companys) life.
/jarek
We have a number of people that need to access Pro/E to a varying degree. A couple of network licenses would solve that just fine.
Imagine a Pro/E "Server" on an alpha and a nubmer of Pro/E GLX clients (preferably Intel). This could become a reality in the near future.
If I remember correctly, 2.1 is compatible with 2.0 unless the program depends on specific symbols in the library (something I red on the gnu libc web page although I can't find it now).
...
It should be on http://www.gnu.org/software/libc/libc.html but
/jarek
That, ACL's and volume management (at least I believe that is a kernel issue) is what I need. Esp. ACL's. Unfortunately there seems to be litle interest in ACL's on the kernel mailing list. Anybody knows if there is a ACL project going on. The page most people were pointing to doesn't seem to exist any more.
/jarek