Ubuntu 10.10 Benchmarks

[phoronix]

Phoronix: Ubuntu 10.10 Benchmarks

With Ubuntu 10.10 having been released yesterday on the 10th of October, many Ubuntu users will be upgrading to this latest release in the coming days. However, for those that are concerned about the performance of this latest release that is codenamed the Maverick Meerkat, here are some benchmarks comparing its performance to Ubuntu 10.04.1 LTS as well as last October's Ubuntu 9.10 release.

http://www.phoronix.com/vr.php?view=15335

[Azmo]

As I have said earlier, a simple table for the OS component information would be much easier to read.
Let me make it abundantly clear what i mean.

<table border="1">
<tr>
<td></td>
<td>Ubuntu 9.10</td>
<td>Ubuntu 10.04.1 LTS</td>
<td>Ubuntu 10.10 RC</td>
</tr>
<tr>
<td>Linux kernel</td>
<td>2.6.31</td>
<td>2.6.32</td>
<td>2.6.35</td>
</tr>
<tr>
<td>GNOME</td>
<td>2.28.1</td>
<td>2.30.2</td>
<td>2.32.0</td>
</tr>
<tr>
<td>X.Org Server</td>
<td>1.6.4</td>
<td>1.7.6</td>
<td>1.9.0</td>
</tr>
<tr>
<td>GCC</td>
<td>4.4.1</td>
<td>4.4.3</td>
<td>4.4.5</td>
</tr>
</table> 

[smitty3268]

As I have said earlier, a simple table for the OS component information would be much easier to read.

No kidding. It's currently easier to google "Ubuntu 10.10 kernel" to figure out what version is included than it is to try and pick through the jumble of numbers in there.

At one point, i think Michael said he was going to fix this, but it hasn't happened yet. Just like the forum edit timeouts...

[Shining Arcanine]

As I have said earlier, a simple table for the OS component information would be much easier to read.

Michael only responds to feedback from premium members. Feedback from freeloaders falls on a death ear.

The same goes for requests for benchmarks that would corroborate the theory that the drop in the Apache benchmarks is because of the barriers (i.e. benchmarks in new kernels with that off).

Higher quality review sites for Windows systems do these sorts of things, but since Phoronix covers Linux, it has little competition because very few people are interested in doing benchmarks on a regular basis, much less in doing benchmarks of Linux on a regular basis. Because of that, Michael has no reason to improve his standards for journalism on this matter.

[MartjeB]

Premium user here.

As I have said earlier, a simple table for the OS component information would be much easier to read. Let me make it abundantly clear what i mean.

<table border="1">
<tr>
<td></td>
<td>Ubuntu 9.10</td>
<td>Ubuntu 10.04.1 LTS</td>
<td>Ubuntu 10.10 RC</td>
</tr>
<tr>
<td>Linux kernel</td>
<td>2.6.31</td>
<td>2.6.32</td>
<td>2.6.35</td>
</tr>
<tr>
<td>GNOME</td>
<td>2.28.1</td>
<td>2.30.2</td>
<td>2.32.0</td>
</tr>
<tr>
<td>X.Org Server</td>
<td>1.6.4</td>
<td>1.7.6</td>
<td>1.9.0</td>
</tr>
<tr>
<td>GCC</td>
<td>4.4.1</td>
<td>4.4.3</td>
<td>4.4.5</td>
</tr>
</table>

[TemplarGR]

Michael only responds to feedback from premium members. Feedback from freeloaders falls on a death ear.

The same goes for requests for benchmarks that would corroborate the theory that the drop in the Apache benchmarks is because of the barriers (i.e. benchmarks in new kernels with that off).

Higher quality review sites for Windows systems do these sorts of things, but since Phoronix covers Linux, it has little competition because very few people are interested in doing benchmarks on a regular basis, much less in doing benchmarks of Linux on a regular basis. Because of that, Michael has no reason to improve his standards for journalism on this matter.

1) Those freeloaders are viewing ads though...

2) It is true that Phoronix's benchmarks are flawed. For example the article comparing versions of Mesa with R600c using only 2.6.32 kernel is useless... I visit Phoronix mostly for the news.

Now i don't know if Michael will take my freeloader advice seriously, but i believe that if you spend much of your time doing something, then at least do it right, make it worthwhile...

For example, i believe Michael spend a considerable amount of his time for the article on R600. But this effort was in vain. If he could spend a little more time, changing at least one or two kernel versions and benchmarking again, he would have a much more usefull article, and the amount of additional effort would be minimal in comparison.

[Michael]

2) It is true that Phoronix's benchmarks are flawed. For example the article comparing versions of Mesa with R600c using only 2.6.32 kernel is useless... I visit Phoronix mostly for the news.

Now i don't know if Michael will take my freeloader advice seriously, but i believe that if you spend much of your time doing something, then at least do it right, make it worthwhile...

For example, i believe Michael spend a considerable amount of his time for the article on R600. But this effort was in vain. If he could spend a little more time, changing at least one or two kernel versions and benchmarking again, he would have a much more usefull article, and the amount of additional effort would be minimal in comparison.

The older kernel version had to be used as the vsync mode switch with the very latest DRM was broken at the time of testing, as I had told Alex Deucher back at XDS, so the FPS would be 30/60.

[Michael]

Michael only responds to feedback from premium members. Feedback from freeloaders falls on a death ear.

That's incorrect, though premium feedback does carry more weight.

Regarding the tabled system/software info data, as I said originally, that will be coming with the new PTS3 / Phoronix Global where that information will be easily embeddable on other pages and be much more rich than even you guys are expressing with your little HTML table code shown in this thread. For those that use PTS, you know how it's formatted right now and once Iveland is complete you will see the new implementation for how it's going to be displayed on Phoronix.com.

[MartjeB]

as I said originally, that will be coming with the new PTS3 / Phoronix Global where that information will be easily embeddable on other pages and be much more rich than even you guys are expressing with your little HTML table code shown in this thread.

Excuse me? Your tone is really arrogant and condescending, Micheal. It's not the first time I note this. For once, please, listen to your users and treat them like people, not like dumb, ignorant fucks.

*Discontinues Phoronix Premium subscribtion

[smitty3268]

Excuse me? Your tone is really arrogant and condescending, Micheal. It's not the first time I note this. For once, please, listen to your users and treat them like people, not like dumb, ignorant fucks.

*Discontinues Phoronix Premium subscribtion

This seems unnecessarily over the line, but yes, Michael is missing the point.

Of course, the html table is simple. That's the point, to create something that could be added to the article in about 10 seconds, so that it will hopefully actually be used rather than waiting for some distant better fix to come along. I mean, personally, I would rather have the existing information out of the article completely rather than have it in it's current form, because it's just so hard to read and it breaks my flow going through the text. So I would say, option #1 - remove it entirely until you get the better fix, option #2 - go ahead and add the simple table instead until you get the better fix. Option #3 - ignore it until you get the better fix seems stupid to me.