What is the reason to use ObjectData? I only see drawbacks: you depend on
Java simply to retrieve a value (do not talk about searching), BLOBs are
very unnice to deal with and compatibility is also an issue.
My belief is that it makes much more sense to use a structure like this for
the attrs:
VISITOR_ATTR_TYPES
-------------------
ATTR_TYPE_ID TYPE_NAME
1 NAME
2 EMAIL
3 SHOE SIZE
4 BAD HABITS
and store attr values in
VISITOR_ATTRS
-----------------------
VISITOR_ID ATTR_TYPE_ID VALUE
1 1 Rafal
1 2 [EMAIL PROTECTED]
1 3 42
2 1 Fedor
.....
This way you have enough flexibility to add/remove various attributes
without any java coding and schema changes by simply manipulating entries in
VISITOR_ATTR_TYPES. But this way you do not depend on Java to retrieve
values and you can run queries against any attr. There is a question whether
it makes sense to move system-used attrs (uid, passwd, email) into this
structure or leave in VISITOR table or, maybe, duplicate in both.
Makes sense?
On Fri, 22 Sep 2000 15:16:30 +0200, Turbine wrote:
> John McNally wrote:
> > Assuming this is not done while the application is live, it should be
> > possible to add a column to the database to accept data from an
"attribute
> > (the Perm hashtable)". The admin adds the column which has the name of
the
> > key. The app is started and TurbineUserPeer is then aware of the
column.
> > It is a problem as it is currently written that the new empty column
would
> > overwrite the value stored in OBJECTDATA the last time the user was
saved.
> > I think this should be changed, so that when including the attributes
stored
> > in named columns if a key is already present in Perm, (and maybe that
the
> > column data is null or "", etc.) the data in Perm stays, so that the
next
> > time the user is saved the column will get populated. Anyone see a
problem
> > with this?
>
> Generally, the decisons which attributes should be kept in separate
> columns should be made before the applications starts populating
> the database. But what you wrote makes perfect sense. We should read from
> named columns, and then from OBJECTDATA. (and document why it's done
> this way :))
>
> Rafal
>
> --
> Rafal Krzewski
> Senior Internet Developer
> mailto:[EMAIL PROTECTED]
> +48 22 8534830 http://e-point.pl
>
>
> ------------------------------------------------------------
> To subscribe: [EMAIL PROTECTED]
> To unsubscribe: [EMAIL PROTECTED]
> Search: <http://www.mail-archive.com/turbine%40list.working-dogs.com/>
> Problems?: [EMAIL PROTECTED]
>
fedor.
_______________________________________________________
Say Bye to Slow Internet!
http://www.home.com/xinbox/signup.html
------------------------------------------------------------
To subscribe: [EMAIL PROTECTED]
To unsubscribe: [EMAIL PROTECTED]
Search: <http://www.mail-archive.com/turbine%40list.working-dogs.com/>
Problems?: [EMAIL PROTECTED]
