Paul McNett wrote:
> no PKs: no, it won't work, but don't we assume there will be PK's all 
> over the place already?

BTW, keying on PK was my second attempt. The first time around, I had 
the dicts keyed on the id() of the record object. However, I never 
investigated why, but the record objects were getting redefined at some 
point and thus getting different id()'s.

I think it would be much better to avoid having to key on PK, but I 
decided to get the proof of concept done as fast as possible. I'll look 
into what is redefining the records as this is probably unnecessary and 
also bad for performance, at which point I can go back and key on 
id(rec) instead of PK.

I also thought of keying on row number, but discarded that idea early on 
because a simple sort operation can change the row numbers.

-- 
pkm ~ http://paulmcnett.com


_______________________________________________
Post Messages to: [email protected]
Subscription Maintenance: http://leafe.com/mailman/listinfo/dabo-dev

Reply via email to