aqbanking3 branch

Christian Stimming stimming at
Sun Jul 6 15:40:18 EDT 2008

Am Sonntag, 6. Juli 2008 21:00 schrieb Micha Lenk:
> Yes, Gnucash does store the AqBanking uid in the gnucash file itself.
> Gnucash files are usually in gzipped XML format, hence you can simply
> open it for instance with zless and search (hit '/') for the hbci slots
> (<slot:key>hbci</slot:key>) and the enclosed other elements holding the
> actual data.
> I have no strong opinion about it, but probably Gnucash should store the
> account number and the bank code only to avoid such 'problems'.
> @Christian, do you remember why the AqBanking uid is stored by earlier
> Gnucash versions (those you were involved in)?

Originally, gnucash stored only the bank code and account number. Then at some 
point during aqbanking (1, probably), the "unique id" for each of aqbanking's 
account were introduced. I implemented this as a new mapping (with higher 
priority), because it looked as if this would be "even more unique" than bank 
code + account number. However, over time it turned out the original mapping 
was probably sufficient. Eventually I think gnucash should probably stop 
using the "aqbanking uid" and instead only use the conventional bank code + 
account number.


More information about the gnucash-devel mailing list