Hallöchen!
llanitedave writes:
> [...]
>
> I guess it's mostly a normalization question.
>
> And while I was typing out a long explanatory discussion to
> enlarge on the problem, I stumbled across the answer.
>
> I'll need to use a separate table for each sample type to store
> its unique set of attributes.
>
> For example, a fluid sample type might have the fields: volume,
> concentration, pH, and expiration date. A soil sample might have
> the fields: Clay content, color, grain size, moisture content.
>
> All sample types will have the common fields sample ID, sample
> type, date collected, collector, source, and storage location.
>
> So I'll need a generic "Sample" table and then related tables for
> each sample type.
In our nanotech institute, we have already created a samples
database with Django.
We used an approach very similar to yours: There is the "Sample"
model with almost no fields -- more or less, just the sample name.
Then, there is a "Process" model with a timestamp, and an M2M
relationship with "Sample".
And finally, there is a plethora of models derived from "Process"
through multi-table inheritace. Etching processes, deposition
processes, measurements etc. Even properties inherent to a sample
like sample size or sample substate are organised in special
processes.
Now, the sample's datasheet simply is the list of all related
processes, in timestamp order.
> The difficulty is that many sample types are still undefined, and
> will have to be defined by the individual users. That means
> they'll have to have an interface that allows them to create
> tables and fields in the database. I suppose I can put together a
> form for that which will ensure it gets done in a limited,
> consistent and standard way.
We haven't thought about this yet, and we won't do this probably.
What we *may* implement someday is that the administrator can create
new processes without programming. But this still includes
restarting the server.
On the other hand, we've defined a "general purpose" process with a
spreadsheet-like 2D array, and the user can set the column names
arbitrarily.
Tschö,
Torsten.
--
Torsten Bronger, aquisgrana, europa vetus
Jabber ID: torsten.bron...@jabber.rwth-aachen.de
or http://bronger-jmp.appspot.com
--
You received this message because you are subscribed to the Google Groups
"Django users" group.
To post to this group, send email to django-us...@googlegroups.com.
To unsubscribe from this group, send email to
django-users+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/django-users?hl=en.
