1 Reply by Syrphus

  • Syrphus
  • Veteran
  • Offline
  • From: Strathpeffer, Highland
  • Registered: 16-12-2008
  • Posts: 535

Re: XL, R2002, and R6 dates problem

While converting from R2002 to R6 I imported the R2002 database with 89 taxon determination errors relating to 2 taxa.  I dealt with these problems.

While trying subsequently to achieve full functionality in the new R6 installation, I revalidated the R6 database.  That threw up 36 errors relating to determination date (TAXON_DETERMINATION.VAGUE_DATE_START and TAXON_DETERMINATION.VAGUE_DATE_END) being earlier than sample date (SAMPLE.VAGUE_DATE_START and SAMPLE.VAGUE_DATE_END).

Inspection of these records in R2002 showed that two of the records had been edited manually, and I had edited the sample date but not the determination date.

The others - all of which had been entered via the Excel import wizard on 5 separate occasions - were year dates (e.g. ‘1918’) where the sample date had been correctly stored as 01/01/1918 - 31/12/1918, but the determination date had been stored as the date value for the year (01/04/1905 - 01/04/1905).

This was in the context of 1962 samples in the R2002 database of ‘Y’ date type which had been correctly imported and stored. All these imports were carried out with the Excel import wizard, apparently without errors (always only the ‘1 item rejected’ message) and in no case were separate determination and observation dates supplied.  All were done with the same tested routine, where dates  in the XL import file were in text format - precisely to avoid vague date (M or Y) problems.  In these same imports, other ‘Y’ records were correctly stored.

Two points emerge:
1 - the Excel import wizard has apparently been inconsistent its processing of the data import to R2002 (I have also recently identified inconsistency in the processing of determiners, where records are ‘successfully’ imported with no determiner stored).
2 - the R2002 --> R6 transfer did not throw up the date mismatch - that was only revealed by the revalidation of the R6 database.  Revalidation is not advised in the documentation on transfer from R2002 --> R6 (indeed I can find no reference to the revalidation of the database in the ‘Help’ or ‘Getting started’ facilities).

Season's greetings.

Murdo

Syrphus's Website

2 Reply by sallyrankin

  • sallyrankin
  • Administrator
  • Offline
  • Registered: 07-04-2006
  • Posts: 371

Re: XL, R2002, and R6 dates problem

The Recorder 2002 to Recorder 6 data transfer doesn't validate the data as is done when data is imported into either system. The errors detected in this process are mainly referential integrity errors, i.e. it can't transfer an item because an entry is required in another table before it can be added to the database. The ability to revalidate the database or sections of it was a new facility in v6.13. This wasn't in Recorder 2002 or earlier versions of Recorder 6.

There are links to the information on revalidation in the Help under the heading Significant changes in Recorder 6 version 6.13 on the opening page of the Help. Unfortunately the indexing in the Help is not what it should be as resources for recent improvements have not allowed time for sorting this out.

Sally Rankin, JNCC Recorder Approved Expert
E-mail: s.rankin@btinternet.com
Telephone: 01491 578633
Mobile: 07941 207687

3 Reply by Lynn Heeley

  • Lynn Heeley
  • Administrator
  • Offline
  • Registered: 03-07-2008
  • Posts: 345

Re: XL, R2002, and R6 dates problem

Sally,

Now that 6.13 has the revalidation tool do you think it would be useful to add this point to the transferring from R2002 to R6 document (ie suggest revalidating after transfer to detect errors with referential integrity)?

Lynn