The time stamps should be accurate to within about +-1ms now.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@120 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
2. Initialization of PTT to OFF is fixed.
3. WSJT_Source_Code.tst updated.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@117 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
2. Fixed grid2deg to be case-insensitive for grid locator.
3. Bumped Revision number to 5.9.3.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@116 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
Updated README_592.TXT in preparation for Windows release.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@115 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
all files previously using conf.h
- pass defs found into the Makefile instead
- add parallel port enable to configure --enable-parallel vs. --enable-serial
- use code (partially contributed by yours truly anyway) from cwdaemon to add
parallel port support for *nix.
I've not tested parallel port operation, it should match what cwdaemon
does, which is convenient.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@113 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
it was causing my grief with tx testing of OSS
- fixed configure to fill in just start_oss.c for AUDIO
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@110 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
- cleaned up compiler warnings (added a few casts)
- transmit appears to work
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@108 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
transmit not tested, I don't even know how to get the portaudio
version to transmit. ;-)
Works on receive just fine though.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@107 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
2. Removed the two-buffer offset in fivehz.f90, it did not work right.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@104 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
source files needed for the audio system chosen. Note in particular
a2d.f90 no longer needs to be wrapped with a #ifdef PORTAUDIO since
it is now only compiled and linked when needed.
Tested on portaudio here, but untested on alsa. Please test that
I have not screwed it up on alsa or windows. ;-)
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@102 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
- split start_threads.c into start_alsa.c and start_portaudio.c
- make configure/Makefile add start_alsa or start_portaudio as needed
instead of start_threads
Untested on ALSA, works fine with PORTAUDIO
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@100 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
FSK441 and JT6M, as well as JT65. (Still no generated pings, however.)
2. Updated README_592.TXT in preparation for open release of version 5.9.2.
Updated READ
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@97 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
- left in the types import as well for possible debugging.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@93 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
array (from fortran), it should be converted to string before using it as one.
This fixes the odd zeros() problem.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@92 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
fixed some dependancies to force recompile if a file is changed.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@86 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
2. Tweaked code that measures sampling rates (in fivehz and fivehztx) so
that it works well with both PortAudio and bare ALSA.
3. Added a "Using PortAudio" message to a2d.f90.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@84 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
Notice that I have an explicit #ifdef USE_PORTAUDIO for later use.
The alsa code was swapping around the original iret1 iret2 and thread1
thread2 also making it confusing. This should 100% agree with what
Bob had now for USE_PORTAUDIO.
This code should be re-factored if we are keeping alsa, start_threads.c
should have no audio code in it. Something like jtaudio_portaudio.c
jtaudio_alsa.c jtaudio_oss.c would be what I would do, with the
file to compile chosen at configure time.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@83 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
Removed the divide-by-2 of input samples in portaudio callback.
Re-activated #ifdef USE_PORTAUDIO in a2d.f90.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@80 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
- Try to guess at a default between portaudio and alsa for now
- changed "samplerate.h" to <samplerate.h> in resample.c
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@76 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
- always define -pthread
- make sure right libs are passed along
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@74 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
--enable options.
This will have to be cleaned up a lot more to minimise #ifdef's
a2d_() is needed for portaudio which is needed by both windows and *nix
but only a small piece of start_threads.c is needed for portaudio
the ALSA part is #ifdef'd out when USE_PORTAUDIO is enabled.
audio_init.f90 and a2d.f90 now have #ifdef's on USE_PORTAUDIO
instead of Win32.
Lots to do still.
git-svn-id: svn+ssh://svn.code.sf.net/p/wsjt/wsjt/trunk@73 ab8295b8-cf94-4d9e-aec4-7959e3be5d79
