Initial import (was under CVS).
vonkorff did the following:
Fixing splitter
Fixed graphics for windows version
Working on progress reporting bug and others
Today I mostly worked on the "hardcoded breakpoint bug"
Changed timer on ap_fold.cpp from 60 seconds to 300 seconds.
Fixed debug_msg output to go to stderr and stdout
increased rsc_memory_bound to a more reasonable value.
Removed extra factor of 1024 from rsc_memory_bound.
4.22
Fixed "best" pulses recording so it remembers time series
Upped rsc_fpops_est to match actual run times.
Fixed overflow error in pulsegraph code
Changed some encodings to _x_hex
Fixed segfault in splitter. Matched last block done with dataseq.
Adding in.dat and dll files for use on snake
After Eric got Astropulse to compile and run on snake
Code for printing the state file.
Changes to assimilator are intended to prevent it from
Several routines now live in setilib and so do not needd to be in local source.
Fixed boinc_fraction_done error
Switched byte order in sbtf -- there is no endian issue
Stop assimilating best_pulses
Fixed segfault from unallocated array
Testing
Fixed ap_fold thresholds
Visual Studio complained about sqrt's argument not being
Fixed ffa thresholds.
Added ffa_scale
Oops, really added ffa_scale this time
added ffa_scale
* ap_fold.cpp:
Auto updated. Added ffa_scale
Temp - printouts w/ delete [] threshold_table
ap_fold.cpp
Deactivated debug_msg from automatic set to true
analysis_configs.xml: dm chunk to 15000 instead of 1151
Changed code to reflect new BOINC API
A few comments
Automake changes. Nongraphical portions of the unix client build. The
Astropulse unix graphics now build, but segfault.
Astropulse graphics on UNIX now work... This was accomplished by moving the
Modified project for v6 graphics build on windows.
Returned state_signal to astropulse.h
Added num_std_devs member to ap_signal and state_signal
Added support for -block option for seeking to a given dataseq
Added a placement new to initialize ap_shmem.
ap_fileio.cpp:modified condition for similar signals and signal insertion
Fixed problems with start_block
First commit.
Two changes:
Small oops, changed 2: to 1:
stderr output is now shorter
rsc_fpops divided by 5
Changed delay_bound from 14 days to 30 days.
Removed boinc_time_to_checkpoint
changed fopen to boinc_fopen
Fixed ISCTYPE.C error by changing wuname to type unsigned char.
Moved checkpoints so they are always at start of for loops
No significant changes
Output to stderr will show the state structure when calculate_fold_level
Updated CPU time graphics
Debug printfs to check for blank statefile
After reading from statefile, make sure we didn't read all 0's. If we did, try again up to 100 times, every 5 seconds.
Changed credit calculation to FLOP counting method.
Attempt to fix 0'd statefile error by exiting and restarting.
Trying to fix 0'd statefile error by exiting and restarting.
Updated flops value for claimed credit.
Removed many error messages
Updated FPOPS for credit calculation.
Now using dir_hier_path to find place to put output file
something about LD_LIBRARY_PATH
First version
Moved trigger file and disk space checks and reduced verbosity of the astropulse
Current configs....
analysis_configs: fixed high_pass to 100000.0
Handles the case where there is no <best_pulses> tag.
Fixed validator bug w/ rel difference vs. abs difference for repeating signals.
Changed max allowed signals from 30 to 60 -- need to count both single and repeating!
Fixed free disk space checking for seti@home splitter.
Fixed infinite loop looking for correct DSI in pulse_split.C
Rom's fix to screensaver bug in graphics_main.cpp
remove radar flag added
changed client_state.C to client_state.cpp in BOINC_DIR lookups
First commit of this file
fix client_state.C to .cpp like ../client/m4
Lines having to do with radar removal
char argc -> int argc