Gandalf,

Unfortunately, increasing the work buffer to 10 days does NOT help! I have mine set there and still run out of work after 2 days, so I have to babysit Collatz and the other Boinc projects.

---

Gandalf,

Chakra is a beautiful to look at distro. Does not have a full implementation of BOINC.

RoboLinux would not install.

Linux-Lite, well, it's YAU (Yet Another Ubuntu) so I don't expect anything different. I did install it this morning and learned it is based on XFCE. Does _every_ developer working on XFCE have only one monitor? Why else would you consider it ok to force the left most monitor to be the main?

My 64-bit Mint17 Cinnamon instance makes it about 18 hours before the resource leakage kills the machine.

Just can't bring myself to pull 20Gig of RAM out of this box so it can run 32-bit.

---

Well, it finally happened. After a few days of running great, I auto applied the latest updates as identified by the update manager. Suffered the NVidia Crash 3 times before 10am. Tried installing other version of drivers, even added edgers PPA again.

Get this!

Now no combination of drivers, with or without nvidia-modprobe will let the CUDA be recognized. Not only that, each time I boot I get the nag-o-gram about running without acceleration.

64-bit, the untested frontier, even with an LTS version.

---

On the SSD for this machine I did a fresh install of Mint17 KDE. I manually added the backport, installed nvidia 331 and nvidia-modprobe. For a couple of days BOINC has happily been using the CUDA.

The Cinnamon partition simply can't be fixed. I did file a bug report, but I'm pretty certain everyone involved will assume "the other guy" is going to fix it so, like the vast majority of OpenSource bug reports, it will be ignored until the release it is filed against becomes unsupported and the bug is auto-cancelled.

---

Today's blog post says they have fixed some memory leaks in Cinnamon.

Now that you mentioned SSD, there has been a few posts over at Seti@home suggesting people to upgrade the SSD's firmware when they have had problems with GPU crunching.