- May 29, 2006
-
-
Ruud Overeem authored
Need head of LCS/Transport.
-
Ruud Overeem authored
Need HEAD of LCS/Common for FileLocator classes.
-
Ruud Overeem authored
Merge MAC-CS1-branch with HEAD. See MAC/Journal for changes.
-
- May 15, 2006
-
- May 01, 2006
-
-
wierenga authored
Handle EINTR and EAGAIN errors by retrying 'sendto' or 'recvfrom'
-
- Jan 02, 2006
-
-
Ger van Diepen authored
Blob code has been moved to a separate package
-
- Dec 21, 2005
-
-
Alexander Mueller authored
Solved. See diff.
-
- Dec 20, 2005
-
-
Alexander Mueller authored
Makes the version tags consistent for GCF
-
- Nov 15, 2005
-
-
Alexander Mueller authored
Files are obsolete.
-
- Nov 03, 2005
-
-
Alexander Mueller authored
Correnctions for the "make clean" target.
-
- Nov 02, 2005
-
-
Alexander Mueller authored
Some more cosmetic changes for rpm building.
-
Alexander Mueller authored
Some more cosmetic changes for rpm building and change release nr. from 6.0 to 6.1.
-
- Oct 24, 2005
-
-
Alexander Mueller authored
Some more cosmetic changes for rpm building.
-
Alexander Mueller authored
Some cosmetic changes for rpm building and moving the gcf-pi.conf.in from PI to Protocols.
-
cvs authored
-
- Oct 19, 2005
-
-
Alexander Mueller authored
Cosmetic change in this build file.
-
Alexander Mueller authored
Changes of build files and new files added to be able to install all GCF packages by means of rpms.
-
- Oct 14, 2005
-
-
Alexander Mueller authored
During further tests and the LDR1.7 integration the KeyValueLoggerMaster and Daemon are not working very efficient together. So this is changed too. Fruther at the moment a API manager has requested the KVLDaemon to skip an PVSS update value the value was not remembered for the situation the connection to the KVLDaemon was broken. Now the remembered man ID will be resend if the connection is reestablished.
-
Alexander Mueller authored
Two faults in the source code of the GCFCommon are the reason of the segmentation fault in the KeyValueLoggerMaster. As expected the seg. fault occours at unpacking of the GCFPVDynArr. This is solved now.
-
Alexander Mueller authored
From now the timer values will be internally always (u)int64.
-
- Oct 04, 2005
-
-
Ger van Diepen authored
Use $(top_srcdir)/Makefile.common
-
Ger van Diepen authored
Use lofar_GENERAL
-
- Sep 27, 2005
-
-
Alexander Mueller authored
Made DB connection info parameterset dependent.
-
Alexander Mueller authored
Coupled to first release (HEAD) of OTDB
-
- Sep 21, 2005
-
-
Alexander Mueller authored
rub first calls "make all" and then "make install". In that case in the Makefile.am misses an include path declaration for the lib. This is fixed now.
-
Alexander Mueller authored
Corrects all tags and readme files.
-
Alexander Mueller authored
Invokes the GCFProperty::setTimedValue direct instead of the GCFProperty::setValue.
-
Alexander Mueller authored
Added pointer check and a isConnected check.
-
- Sep 19, 2005
-
-
Alexander Mueller authored
Inserted the getSearchPath method.
-
Alexander Mueller authored
All asserts are changed in ASSERT from LofarLogger.h. If necessary the GCF_Defines.h is include for including the LofarLogger.h indirect.
-
- Sep 05, 2005
-
-
Alexander Mueller authored
Changed LCS-Common-2_2patch1 to LCS-Common-2_3 Changed HEAD to MAC-GCF-5_2rc2 Changed MAC-GCF-5_1 to MAC-GCF-5_2rc2
-
Alexander Mueller authored
Changed a log message.
-
- Sep 02, 2005
-
-
Alexander Mueller authored
All .conf and .log_prop files will be installed in sysconf_DATA (which means 'etc' beside of the 'bin' directory). The GCFTask::init method sets this 'etc' as default path for all *.conf files, which will be adopted to the same (singleton) instance of the ParameterSet.
-
Alexander Mueller authored
The problem described in this bug report, applies also to the genTypeInfo script, which will be opende by the PALlight package. This is therefore solved too.
-
Alexander Mueller authored
Added the setSearchPath and extends the constructor with the default parameter searchPath.
-
Alexander Mueller authored
Inserted some extra checks to filter out invalid data. Furthermore ETM has notified about this problem and they suggest this is actually a bug in PVSS.
-
Alexander Mueller authored
Solved. And added some comments to the prepare script.
-
- Aug 31, 2005
-
-
Alexander Mueller authored
After investigation of this problem it became clear that the seconds converted to usec will not fit in a "unsigned long". So this must be "unsigned long long". The second reason for the problem is the fact the the Timer class was not able to handle a negative timedelta. This can occur if the running NTP service corrects the system time with a time in the past. This is solved too.
-
- Aug 25, 2005
-
-
Alexander Mueller authored
If an assert must be generated the following code will be placed: <code>if (<param>NOE > 0) assert(<param);</code> Furthermore the GCFEvent::packMember method only copies the data if <param>NOE > 0. The use of sizeof in this case is also changed.
-
Alexander Mueller authored
see comment #12, comment #13 and comment #14
-