On Sun, Jun 20, 1999 at 06:40:51AM +0900, Izumi Tsutsui wrote:
> In <19990619224659.F357@jocelyn.rhein.de>
> is@jocelyn.rhein.de wrote:
> 
> > > Yes, I'm spearheading the building of clients for the NetBSD architectures,
> > > and will have then distributed back to them as soon as they finish getting
> > > the proxy bug out of the 1.1 client.  The clients take a while to validate
> > > as they need a complete data set to run through, this takes a pretty long
> > > time on our slower archs.
> > 
> > was the 68k client optimized for 030, 040, 060?
> 
> setiathome-1.2.m68k-hp-netbsd1.3.3 is working on my sun3/80 (030)
> so it does not contain 040/060 instructions.
> But please leave 020/030 version anyway. m68k rc5-64 client
> does not run on my 3/80. It dumps core with "illegal instruction".

I'm not in the power of replacing it, anyway. But I suggest to create a
040-optimized and a 060-optimized one.

> BTW, the m68k seti@home client seems to have some problem.

works fine for me. I'm at 13% or something.
Only could  be faster.
> 
> ---
> [...snip...]
> Doppler shifting data by 0.001849 Hz/s
> Starting work for Doppler shift rate 0.277278, FFT length 131072
> 2.447172% done
> Doppler shifting data by 0.001849 Hz/s
> Starting work for Doppler shift rate 0.279126, FFT length 131072
> Doppler shifting data by 0.001849 Hz/s
> Starting work for Doppler shift rate 0.280975, FFT length 131072
> Starting work for Doppler shift rate 0.280975, FFT length 65536
> Doppler shifting data by 0.001849 Hz/s
> Starting work for Doppler shift rate 0.282823, FFT length 131072
> Doppler shifting data by 0.001849 Hz/s
> Starting work for Doppler shift rate 0.284672, FFT length 131072
> 2.498474% done
> Doppler shifting data by 0.001849 Hz/s
> Starting work for Doppler shift rate 0.286521, FFT length 131072
> 100% done
> Done with work unit.
> Sending result - connecting to server.
> All data sent.
> ---
> 
> It took about *only* 30 hours, but what happened between ~2.5% and 100% ?
> Is this normal behavior? I guess it should take ~1500 hours...

Hm. No idea.
	-is