Ah yes, on Einstein it was my fault. The OS X version that I entered was derived from a person's computer running ABP2. It doesn't say in the Applications page which version is needed, nor in the thread on the forums. I read Darwin 10.3.0 and misdeduced. :-)
Collatz show it like that in their app page: http://boinc.thesonntags.com/collatz/apps.php "Mac OS 10.4 or later running on Intel 2.02 (cuda)" On Fri, Jun 4, 2010 at 12:44 PM, Charlie Fenton <[email protected]> wrote: > At 11:39 AM +0200 6/4/10, Jorden van der Elst wrote: >> >> I've adapted >> http://boinc.berkeley.edu/wiki/GPU_computing#Attach_to_projects_with_GPU_applications >> to show OSes. > > Thanks Jord, but I'm afraid that change is not quite correct, for a couple > of reasons. I've edited the wiki some more to try to help ease the > confusion. > > First, the NVIDIA web site says that for the current version 3.0 of the Mac > drivers require Mac OS X version 10.5.6 or later (which is Darwin 9.6.0 or > later): > <http://developer.download.nvidia.com/compute/cuda/3_0/docs/GettingStartedMacOS.pdf> > > The older NVIDIA driver version 2.3 supports Mac version OS X 10.5.2 (Darwin > 9.2.0) or later, but for clarity I didn't include that information on the > Wiki page. Anyone running OS 10.5 should have upgraded to OS X 10.5.8, > which is a free update. And people installing the CUDA drivers for the > first time on Windows or Linux should get the newest drivers. > > Second, those changes have confused the Darwin version with the OS X > version. BOINC doesn't support OS X 10.3.0, and Intel Macs don't run on OS > X 10.3.x at all. > > As I tried to explain before: > OS X 10.6.3 = Darwin 10.3.0 > OS X 10.5.8 = Darwin 9.8.0 > OS X 10.4.11 = Darwin 8.11.0 > OS X 10.3.9 = Darwin 7.9.0 > > In general: > - Darwin version 10.x.y corresponds to OS 10.6.x > - Darwin version 9.x.y corresponds to OS 10.5.x > - Darwin version 8.x.y corresponds to OS 10.4.x > - Darwin version 7.x.y corresponds to OS 10.3.x > > The third part of the Darwin version number should be ignored; as far as I > have seen, it is always zero. > > -- -- Jord. _______________________________________________ boinc_dev mailing list [email protected] http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev To unsubscribe, visit the above URL and (near bottom of page) enter your email address.
