On Fedora Rawhide I do the following:
svn co https://svn.code.sf.net/p/wsjt/wsjt/branches/wspr
cd wspr
sh au*sh
./configure
---- change all instances of f2py in Makefile to f2py3
make
python3 -O wspr.py
This also worked on Fedora 20 the last time I tried it.
Fedora uses python2 for some functions that have noot been ported to
python3,
hence the need to have both versions installed and available.
On 06/17/2014 04:45 PM, ki7mt wrote:
Hi Richard,
Sorry, was away for a few days and just now caching up. See below.
On 06/10/2014 10:52 AM, Richard Shaw wrote:
Ok, I got it fixed but I'm not sure what the exact problem is.
WSPR doesn't like the default LDFLAGS on Fedora (and probably the
same for Debian?)
LDFLAGS="-Wl,-z,relro
This is the same for Debian / Ubuntu. F2PY does not like / will not
accept -Wl, .. .. even with --as-required, so on the build servers, I
set the LDFLAGS to null == "" initially, then perform the build.
If anyone is interested I have a couple of patches that I needed to
fix the build as f2py didn't like LDFLAGS as well as adding DESTDIR
support to the make install target.
I'd be interested in the F2PY patches if you could send them over, as
I've not found a way around it, other than setting to NULL. Did you
patch Numpy source or the f2py.py file
One thing I noticed that is probably overkill, currently the shell
script to start wspr is installed into /usr/share/wspr and then
symlinked to /usr/bin. There's nothing wrong with installing the
shell script directly into /usr/bin and just dropping the .sh
extension. As long as it is executable and has the correct shebang on
the first line it will work fine.
I updated this in the Debian rules file to: mv -T $(BINDIR)/wspr.sh
$(BINDIR)/wspr but forgot to change that in Makefile.in. Will catch
that one on the next update.
Thanks,
Richard
------------------------------------------------------------------------------
HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
Find What Matters Most in Your Big Data with HPCC Systems
Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
Leverages Graph Analysis for Fast Processing & Easy Data Exploration
http://p.sf.net/sfu/hpccsystems
_______________________________________________
wsjt-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
------------------------------------------------------------------------------
HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
Find What Matters Most in Your Big Data with HPCC Systems
Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
Leverages Graph Analysis for Fast Processing & Easy Data Exploration
http://p.sf.net/sfu/hpccsystems
_______________________________________________
wsjt-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
--
Chuck Forsberg WA7KGX [email protected] www.omen.com
Developer of Industrial ZMODEM(Tm) for Embedded Applications
Omen Technology Inc "The High Reliability Software"
10255 NW Old Cornelius Pass Portland OR 97231 503-614-0430
------------------------------------------------------------------------------
HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
Find What Matters Most in Your Big Data with HPCC Systems
Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
Leverages Graph Analysis for Fast Processing & Easy Data Exploration
http://p.sf.net/sfu/hpccsystems
_______________________________________________
wsjt-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
