I've used nightly CVS snapshots for both the boinc client (from boinc_public-cvs-2004-09-08.tar.gz) and the seti@home client (tried three versions -- seti_boinc-client-cvs-2004-09-08.tar.gz, seti_boinc-client-cvs-2004-09-15.tar.gz , and seti_boinc-client-cvs-2004-09-22.tar.gz). Got successful builds in all cases (after a hiccup or two).

Problem is, when I run either setiathome_test or setiathome_4.3_i386-unknown-freebsd4.10 from the test_workunits directory using the supplied test WU, the client stops processing and I get the following message in stderr.txt:

boinc_parse_init_data_file(): can't open init data file
SETI@Home Informational message -9 result_overflow
NOTE: The number of results detected exceeds the storage space allocated.

With the seti@home CVS from 9/8, this happened after about 15 minutes, so I know that there was a problem. However, with the 9/15 and 9/22 snaps, it's taking around what I think should be a normal WU processing time for this machine.

So my question is, can I go ahead and begin using the stuff I've most recently built to connect to the project and begin processing actual WUs, or is there a real error in the client I've built and thus I'd just be wasting the machine's cycles?

Thanks for any help...

-dnorris

I've compiled recently boinc and seti@home client on a FreeBSD.

I tried on two systems:

1. FreeBSD release 5.2
using GNU make (port gmake)
and using command
gmake -k
(this is said on the boinc www, that non critical errors are omitted)
both boinc client and seti client compiles fine and working without problems.

2. FreeBSD stable 4.8
I also used GNU make
and -k option.
Both programs compiled without problems.
Boinc is working fine. But seti client is producing error
and destroying unit :/

I must add, I used cvs from 20 september 2004

Thank you for the replies.

I was planning to upgrade the machine in question to 5.3 when that is released anyway, so if no other solution comes forth then it's still not that long to wait.