[open-economics] Yourtopia questions

Mon Feb 14 09:23:52 UTC 2011

Dear all,

If it is difficult to load the new questions without removing the
previous data, should we then just go forward deleting that data? I
think the price of giving up the data collected so far is worth paying
if the questions can be exchanged soonish then. 

What do you think?

Best wishes,

Dirk

On Mi, 2011-02-09 at 23:02 +0100, Stefan Wehrmeyer wrote:
> Hi guys,
> 
> > Unfortunately I do not understand a few things. The new csv contains
> > modified descriptions for the new description-boxes. Also it contains
> > short labels for the piechart on the results page. So there are two new
> > categories that we did not use to have. Hence the csv IS in a different
> > format.
> > 
> > Also, since you did not like when I deleted columns in the
> > Google-Spreadsheet I did not delete them either in the csv but added new
> > ones (new_label, new_description, new_piechart)
> 
> I actually renamed the columns in indicator.csv from new_label to label and from label to old_label etc. so to not break the importer. I added the capability to write the piechart column to the importer (but piechart is not used yet in the templates).
> 
> The import works, but as outlined in issue 25 (not issue 21), using the load feature of the importer without deleting indicators first puts too many indicators in the db and deleting everything first with delete_all deletes also the answers which would also be bad. So integrating the new indicator.csv needs a custom importer script as far as I can see.
> 
> https://bitbucket.org/okfn/openhdi/issue/25/load-new-questions-from-indicatorcsv
> 
> Apart from that, I integrated the new description in the templates in a hopefully decent way.
> 
> Cheers,
> Stefan
> 
> 
> 