rif6894 - Monday, December 5, 2011 11:10 PM:
I wish to instantiate a legacy revision sequence that could include:
1 2 3 4 A A.1 A.2 ...
and
01/10/2009 02/02/2010
But it seems that:
1. The Revisions tables is limited to 1024 characters,
2. You can only specify a revision that is in Revisions tables, and
I am able to create, for example, the legacy data has revisions 1 2 4 6. I can purge revisions 3 & 5. But as there is the 1024 limit on the revisions, it is not possible to include all the possible (230) combinations of the legacy revision sequences. I am using java/SOAP as the legacy data is hosted on a Solaris platform
Suggestions?
AdamStrachan - Tuesday, December 6, 2011 2:07 AM:
Possibly you can edit the Sequence itemType, and change the length of the "sequence" property to 2048 or something.
I haven't tried this, but it might solve your problem.
Just a note, I think you have to be logged in as the super user to edit core itemTypes, so be careful not to break anything!
rif6894 - Wednesday, December 14, 2011 5:09 PM:
Adam
Thanks for your response.
1. The ITemType->Revision->revision length can be increased to 4000 characters. This covers all the variations in the legacy dataset.
2. Further investigation found that the major_rev attribute can be changed (in the Preliminary state) to any value that is in the Revision setting,
3. Promote to Released and Revise, creates the next generation, with the Revision being the next character(s) in the Revision sequence and
4. Deleting intemittant generations is possible, so that there are no extranious object created during the legacy data import.
So legacy data importing is practical within ARAS.
