There seem to be quite a few test runners that haven't run any tests this week (or at least I'm not seeing any of the new Boost.Math tests showing up), frankly it's hard to see how any of these can be official release platforms if they're run rather infrequently: which is not to say they're not useful non-the-less! I'm just flagging this up, in case folks were thinking that any of these *were* being run frequently, those with a "*", currently have no recent tests for that compiler/platform: *AIX, IBM xlc, last run Fri 5th Oct. *Darwin, gcc 4.0.1, OSL, last run Fri 5th Oct. *Darwin, gcc 4.0.1, Sandia, last run Thurs 4th Oct. *Darwin, gcc 4.0.1, Sandia, last run Fri 5th Oct. *Irix64, SGI mipspro, Sandia, last run Fri 5th Oct. Linux, gcc-4.1.2, Faux, last run Sat 29th Sept. Linux, gcc/pathscale/pgi, Sandia, last run 5th Oct or earlier. Linux, intel c++, Sandia, last run 5th Oct. Linux, gcc, OSL, 14th Sept. Linux, gcc, BGD-Ubuntu, last run 30 Sept. Sun OS, Sun-5.8, OSL, last run 5th Oct. Win32, Borland C++, Siliconman: these have been run recently, but against an old version of the trunk. *Win32, gcc-cygwin/minw32, speedsnail, Last run 1st Oct. Most of the other tester have run once or twice last week, with only the following showing reasonably rapid (sub 48hrs) turnaround times (this is my just my subjective impression based on monitoring the results this week): HP-UX IA64 gcc. RSI Droid Linux gcc-4.1.2 Caleb Epstein SunOS-5.10 gcc. Huang-WinXP-x86_32 VC8, Intel-10 RudbekAssociates-V2 VC7.1, VC8. Huang-Vista-x64 VC8 Intel-10 BCB Boost, Borland 5.6.4, 5.8.2, 5.9.2 Please note: I'm not trying to malign anyone here, we *REALLY* appreciate all the time, effort and resources that all the tester put in: especially if you've made it my "frequent" list :-) I also note that quite a few of the tests were last run on the 5th Oct, which may indicate a common cause (tarball problems?). Anyway I thought I'd better flag this up, so we can get that 1.35 release moving along :-) Regards, John Maddock.