The Huge Disaster Within The Linux 2.6.35 Kernel

[Michael]

Most ads are for page views. There's text and graphical ads. Simple text ads don't really produce a lot of revenue.

[sabriah]

When I looked at http://www.phoromatic.com/kernel-tracker.php a few minutes ago there was no improvement yet, and the latest date was 2010-05-30, with 5% threshold. Phoromatic chugs along, relentlessly.

When I looked at http://www.phoromatic.com/kernel-tracker.php a few minutes ago there was no improvement yet, and the latest date was 2010-06-01, with 5% threshold. Phoromatic chugs along, relentlessly.

[mtippett]

When I looked at http://www.phoromatic.com/kernel-tracker.php a few minutes ago there was no improvement yet, and the latest date was 2010-06-01, with 5% threshold. Phoromatic chugs along, relentlessly.

The only caveat there is that the Ubuntu kernel daily builds have stopped.

http://kernel.ubuntu.com/~kernel-ppa/mainline/daily/

A new daily build arrived today, so we'll see if the regression has been fixed.

The way Phoromatic trackers currently work is just pickup the most recent daily build. So if there is no updated build it will re-use the most recent.

Regards,

Matthew

[not.sure]

The only caveat there is that the Ubuntu kernel daily builds have stopped.

Ubuntu kernels?????? phormaix uses Ubuntu kernels? I think that makes the whole thing even less useful.
Are you sure you don't want to use plain mainline?

[Elyotna]

Doesn't it build kernel from git every day ? I thought so :/ .

[sabriah]

Doesn't it build kernel from git every day ? I thought so :/ .

Hmmm, that is what I presumed too.

It wouldn't surprise me if Ubuntu applies "in-house" patches which would make it even more difficult to trace regressions for the uninitiated.

While it may be good for Ubuntu-folks, it does limit its general use elsewhere. And, the dates of their builds probably don't match the builds of the plain vanilla builds.

While I think the concept is excellent I think a plain vanilla kernel from git is the optimal way to go. My 2c.

[energyman]

whenever I reported bugs in the past I always made sure I would use an unpatched kernel.org kernel. That way you don't introduce another layer of potential bugs.
If phoronix uses ubuntu kernels for this the article is even less worth than before.

[cynyr]

whenever I reported bugs in the past I always made sure I would use an unpatched kernel.org kernel. That way you don't introduce another layer of potential bugs.
If phoronix uses ubuntu kernels for this the article is even less worth than before.

^^ Truth.
I was lead to believe that this was "vanilla" git kernels that were being tested, not some version of the kernel with ubuntu backporting/foreward porting a ton of stuff to. I will blame ubuntu until the same problem is shown with vanilla packages all around(udev, kernel, gcc, ffmpeg, etc). If it's not a problem with current stable/testing/experimental vanilla packages, the blame is on the distro of choice, 99% of the time. Yes i have a problem with ubuntu, and to a lesser extent redhat/fedora and suse for "patching" packages in incompatible ways.

Reminds me of the old Suse7/8/9 redhat7/7.3/8 days, where a vanilla kernel wouldn't boot, because the init scripts depended on some feature that didn't exist in "vanilla".

Most ads are for page views. There's text and graphical ads. Simple text ads don't really produce a lot of revenue.

Some ad views are better than none no? I used to use adblock and add my own rules, but that just got to be too much, even after i switched to "*.domain.org/*" in it if your ad moved or was flash based. Then easylist came about and i never went back.

[Michael]

It's the mainline PPA packages being tested, not with Ubuntu's backporting mess and other cruft.

[Michael]

Some ad views are better than none no? I used to use adblock and add my own rules, but that just got to be too much, even after i switched to "*.domain.org/*" in it if your ad moved or was flash based. Then easylist came about and i never went back.

Not if it still leads to a loss in ad revenue. Even if the impression count was tripled over current figures, but all ads were only text-based, it still would lead to a loss. Such ads simply don't pay well.