Hi, Le Mercredi 3 Mars 2010 19:00:14, Les a écrit : > Continuing to work on Piklab and Ktechlab, I have gone back to piklab > for now and discovered the following: > 1. When you configure the toolchains, piklab looks in the exec path > for gpasm and gplib and apparently ignores the environment, because when > gputils installs the library manager and assembler are loaded in > the /usr/bin directory and found via path from the command line. > 2. I added a symbolic link to the /usr/share/libexec/sdcc directory A symbolic link need a source and a target. How you created your symlink ? > (the search path for the executables used by piklab) to both gpasm and > gplib, they show up as: > "gpsim" not recognized > "gplib" not recognized > > But pressing the find button for gpasm shows a textbox with the > following in it: > > Command for executable detection: > /usr/libexec/sdcc/gpasm -v > Version string: > > gpasm-0.13.7 beta > > > Pressing the button for gplib shows: > Command for executable detection: > /usr/libexec/sdcc/gplib -v > Version string: > > gplib-0.13.7 beta I get the same behaviour without any symbolic link. Piklab say « "gpasm" not recognized » but clicking the button pops up: Command for executable detection: gpasm -v Version string: gpasm-0.13.7 beta So, piklab is able to find gpasm but doesn't find it. This is contracdictory... Are you able to assemble a file? A project? I have any asm file to test, sorry :-( Breakpoint: an answer is needed here! You can have a try with the 0.15.7 version? http://koji.fedoraproject.org/koji/getfile?taskID=2029116&name=piklab-0.15.7-1.fc12.i686.rpm Parent task is: http://koji.fedoraproject.org/koji/taskinfo?taskID=2029112 It seems it fix this bug. Unfortunately, this version introduce a new bug: http://sourceforge.net/tracker/?func=detail&aid=2945494&group_id=138852&atid=743140 Because it, I can't accept to update in Fedora. Unfortunately, upstream is unresponsive... I found where is the problem, but I don't have a solution :-( In toolchain_config_center.cpp, at line 53, there is a test: if ( it.key()==sgroup.name() ) But if "sgroup" don't exist, there is a bug I try to patch with: if ( sgroup != NULL ) But, this don't work ;-) I don't understand the code, the C object, the Qt C object. If somebody is able to provide me a patch tou fix this this bug, I'll update the piklab package. > Should I bugzilla this? Please, no! :-D Regards, Alain -- Les pages de manuel Linux en français http://manpagesfr.free.fr
Attachment:
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ electronic-lab mailing list electronic-lab@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/electronic-lab
