[wsjt-devel] Ad-hoc Testing - 1.4-rc1 on Linux FC20 x86-64

2014-09-26 Thread Josh Rovero
Built via cmake, no problems with full day of operation -- P.J. "Josh" Rovero Ham Radio: KK1D http://www.roveroresearch.org http://www.roveroresearch.info -- Meet PCI DSS 3.0 Compliance Requirements wi

Re: [wsjt-devel] Credits mismatch

2014-09-26 Thread KI7MT
Hi Joe, I don't know why I didn't think about this before, but, I ran some tests using my Laptop, Dual Core Centrino 2.0Ghz w/4.GB ram i386 Ubuntu install. I opened System monitor, and watched the resource usage (HD, Swap, CPU, etc). The base memory starts out at about 450MB used. It peaks at abo

Re: [wsjt-devel] Credits mismatch

2014-09-26 Thread ki7mt
Hi Joe, Ok, I'll have a look at this. Can probably just pull both -j and -- -kj flags out. All my main boxes have 4+ cores so I rarely see slow downs on *Nix, Windows is another story, it's always slow. As you stated, the -j is for Hamlib3, and -- -kj is for the Cmake invocations, you can axe

Re: [wsjt-devel] Credits mismatch

2014-09-26 Thread Joe Taylor
Bill -- > AFAIK the JTSDK-QT uses 'cmake --build . --target -- > -kj' which passes everything after the '--' to the underlying > make tool, which is 'make' in Linux. I don't know what is > but I thought it was either a sensible value for the hardware or just '-kj'. > > when you say you are runni

Re: [wsjt-devel] Credits mismatch

2014-09-26 Thread Bill Somerville
On 26/09/2014 22:30, Joe Taylor wrote: > Bill -- Hi Joe, > >> Sometimes make goes overboard with the number of parallel processes, you >> can limit the number using '-jN' and with two hyperthreaded cores like >> your machine '-j4' is probably optimal. > As I mentioned, I removed the "-j" flag entir

Re: [wsjt-devel] Credits mismatch

2014-09-26 Thread Joe Taylor
Bill -- > Sometimes make goes overboard with the number of parallel processes, you > can limit the number using '-jN' and with two hyperthreaded cores like > your machine '-j4' is probably optimal. As I mentioned, I removed the "-j" flag entirely on the one explicit call to make -- and the scrip

Re: [wsjt-devel] Credits mismatch

2014-09-26 Thread Bill Somerville
On 26/09/2014 22:00, Joe Taylor wrote: > Hi Greg, Hi Joe, > > For what it's worth: on my rather elderly Linux box running (X)ubuntu > 14.04, hamlib is built without problems. But the machine is brought to > a crawl (presumably with very many page faults?) when trying to build > wsjtx. I notice ma

Re: [wsjt-devel] Credits mismatch

2014-09-26 Thread Joe Taylor
Hi Greg, For what it's worth: on my rather elderly Linux box running (X)ubuntu 14.04, hamlib is built without problems. But the machine is brought to a crawl (presumably with very many page faults?) when trying to build wsjtx. I notice many copies of cc1plus are running in parallel. I remov

[wsjt-devel] hamlib 3

2014-09-26 Thread David
Hi Billtrying to use the Script to build your latest Hamlib 3 followed the script to the letter and i get this error checking for gcc...gcc checkingwhether the C compiler worksNO configure: error: in /home/djmunn/hamlib-prefix/build C compiler cannot create executables -- 73 David VK4B

Re: [wsjt-devel] hamlib 3

2014-09-26 Thread Bill Somerville
On 26/09/2014 21:45, David wrote: Hi David, > Hi Billtrying to use the Script to build your latest Hamlib 3 > > followed the script to the letter and i get this error > > checking for gcc...gcc > checkingwhether the C compiler worksNO > configure: error: in /home/djmunn/hamlib-prefix/build

Re: [wsjt-devel] Credits mismatch

2014-09-26 Thread ki7mt
Hi David, All, I use a script on new VM's (clones from unmodified VM installs) to perform quick build tests with. I added a modified version to JTSDK-NIX that incorporates all of Bills latest updates for the rc1 build as well as, the HEAD development branch ../wsjtx This script is only for WSJ