hi all,
setup : loggernet admin 4.1 and CR1000.
Connection is established via IP.
I have setup the scheduled collection via the setup-screen (TOA5/ Append / Data-logged-since + Collect-all-on-first)
My problem is that everytime a collection is triggered (if scheduled or by "collect now" doesn't matter), a new backup-file is created, and i can't see any reason, why this should happen.
The TableDefs are shown as up-to-date within ConnectScreen and StatusMonitor.
I suspect it will be something minor that i am not able to spot at the moment, but as am quite clueless and i hope to find some help here.
Can anybody tell me,if there is a way to get the exact reason, why Loggernet "thinks" a backup-file has to be created?
Thanks in advance!
Before it appends to a TOA5, TOB1, or TOACI1 data file, LoggerNet will first scan the header of that file and compare the information that it contains with the information reported in the datalogger table definitions. For TOA5, the following conditions can result in errors that will cause LoggerNet to refuse to append to the file:
- table names do not match
- The number of columns do not match
- One or more column names do not match
- The number of units strings does not match the number of column names
- The number of process strings does not match the number of column names
There have been cases where the datalogger program specified a double quote character (") in the units string and this character unbalanced the quote marks in the TOA5 format which in turn threw off the TOA5 parser. This same thing may also be possible to specify in the FieldNames instruction.
Without looking at the datalogger program and/or the resulting data file, it is impossible to judge more closely the cause of the error.
hi jtrauntvein,
I have indeed an unintended unbalanced double-quote within an unit-string.
Hard to spot ;)
Will fix it today and am quite sure this will resolve the issue.
thx a lot for the fast response and detailed listing.
hi,
one question left.
Its clearly a error introduced by me, but shouldn't this be "catched" by the pre-compiler?
Are these kind of programming errors planned to be caught in future versions of the compiler?
I remember i have hit the same kind of error few weeks before with the stationname instruction, which resulted in the tablefile instruction to fail, because of invalid filename(looks like i have some problems with proper quoting XD ).
StationName=FOO" -> OK
StationName="FOO -> FAIL
StationName(FOO")-> OK
StationName("FOO)-> FAIL
the compiler also eats this (have not tested the behaviour):
if foo
bar=1
endif