Perhaps I'm missing something here, but if you check out this article, you'll note that Bell Labs transmitted 3.28 terabit over 300 kilometres by multiplexing different wavelengths over a single fibre strand.
While I don't completely disagree with the above post there are several things to keep in mind; Solaris still performs better in areas such as SMP/Large memory configurations. I've been running Linux on and off on an e4k with 14 cpus and 8 gigs of RAM (flipping between that and Solaris 8 beta snapshots).
Linux still has a ways to go with the SMP implementation. This is probably because of the fact that Linux was more geared towards the Intel platform which doesn't have anywhere near the number of processors that you can pack on a mid to high end enterprise Sun. I suspect it also has to do with the fact that the process of multithreading the various kernel subsystems is a complicated task and Sun seem to have done this better than most, if not all, the commercial Unix software/hardware vendors.
Please don't think I am trying to dis the Linux development effort, I'm typing this on an Ultra 2 running Linux. I should further point out that I am in no way affiliated to Sun Microsystems or any of its subsidiaries.
Phaedrus
PS: I would be interested to see the benchmarks that you used to determine that "under Linux, most O/S-limited operations ran at about four times their speed under Solaris."
Disclaimer: I don't have anything to do with Artx nor do I know "Hannibal" on any basis other than the fact that I read his article.
I'm a bit concerned about both sides of this debacle. It seems to me that people on the one hand are more than willing to denounce some marketing droid from a graphics startup, but on the other hand are willing to make the leap of faith that the author of the aforementioned diatribe doesn't have an agenda and is actually capable of writing a report that isn't tainted with a high degree of subjectivity.
What he said may very well have been true to one degree or another, but that still doesn't excuse the fact that he posted *private* email. Now, maybe I'm just dating myself here by saying this, but I seem to recall that in the good old days of usenet, doing things like that would get you carbonized, no matter how right your arguments were.
It read like a very vindictive article and given what I've seen and heard from the press in the past I'm certain he's not got the entire story straight either (whether it was intentional or not). People never-the-less took for granted that he was absolutely right and the ensuing rabid lemming effect took hold, everyone is up in arms and all the city folk are storming the castle with pitchforks and torches.
I think people need to think twice before they trust anything, especially when it was published on the web.
Slashdot Mirror
Perhaps I'm missing something here, but if you check out this article, you'll note that Bell Labs transmitted 3.28 terabit over 300 kilometres by multiplexing different wavelengths over a single fibre strand.
While I don't completely disagree with the above post there are several things to keep in mind;
Solaris still performs better in areas such as SMP/Large memory configurations. I've been running Linux on and off on an e4k with 14 cpus and 8 gigs of RAM (flipping between that and Solaris 8 beta snapshots).
Linux still has a ways to go with the SMP implementation. This is probably because of the fact that Linux was more geared towards the Intel platform which doesn't have anywhere near the number of processors that you can pack on a mid to high end enterprise Sun. I suspect it also has to do with the fact that the process of multithreading the various kernel subsystems is a complicated task and Sun seem to have done this better than most, if not all, the commercial Unix software/hardware vendors.
Please don't think I am trying to dis the Linux development effort, I'm typing this on an Ultra 2 running Linux. I should further point out that I am in no way affiliated to Sun Microsystems or any of its subsidiaries.
Phaedrus
PS: I would be interested to see the benchmarks that you used to determine that "under Linux, most O/S-limited operations ran at about four times their speed under Solaris."
Disclaimer: I don't have anything to do with Artx nor do I know "Hannibal" on any basis other than the fact that I read his article.
I'm a bit concerned about both sides of this debacle. It seems to me that people on the one hand are more than willing to denounce some marketing droid from a graphics startup, but on the other hand are willing to make the leap of faith that the author of the aforementioned diatribe doesn't have an agenda and is actually capable of writing a report that isn't tainted with a high degree of subjectivity.
What he said may very well have been true to one degree or another, but that still doesn't excuse the fact that he posted *private* email. Now, maybe I'm just dating myself here by saying this, but I seem to recall that in the good old days of usenet, doing things like that would get you carbonized, no matter how right your arguments were.
It read like a very vindictive article and given what I've seen and heard from the press in the past I'm certain he's not got the entire story straight either (whether it was intentional or not). People never-the-less took for granted that he was absolutely right and the ensuing rabid lemming effect took hold, everyone is up in arms and all the city folk are storming the castle with pitchforks and torches.
I think people need to think twice before they trust anything, especially when it was published on the web.