Jojo, am reading your problem correctly?
If you have (for example) x=2 and y=4, you want the database to return
functions like x+2, 2x and 3x-2? (I know this is probably much simpler
than the actual functions you'll be using).
And you the want to use those functions in your code?
@list: pickle?
---
Richard "Roadie Rich" Lovely
Part of the JNP|UK Famille
www.theJNP.com
(Sent from my iPod - please allow me a few typos: it's a very small
keyboard)
On 14 Nov 2008, at 04:42 PM, "Jojo Mwebaze" <[EMAIL PROTECTED]>
wrote:
Sorry Alan,
What u described below is exactly what i want to do.. if given x, y
as two datatums and such that f(x) --> y, given y can we determine f
or x?
Assuming the x, y and f are stored in the database, then we can be
able to write queries to search/extract for the f's that are
responsible for the 'y' or use links, database catalogs etc
Thanks for the help.
Johnson
On Fri, Nov 14, 2008 at 2:21 AM, ALAN GAULD
<[EMAIL PROTECTED]> wrote:
> Thanks Allan... we have used CVS for the base system.. but for
users provided functions,
> we think for having them persistent in the database..
I'm puzzled. CVS provides much better facilities for handling code,
especially with multiple
versions (visibility of diffs, who changed what and when etc) that I
can't think of a single
good reason to put it in a database. I can see the point of using a
database as an indexing
system for searching, filtering etc but storing code in a database,
extracting it and then trying
to execute it is just so much more difficult than fetching a
version file and importing
or running it directly. Plus you need to write a basic version
control system on top of
the database anyway.
I really think I must be missing something about your requirements?
Alan G
On Thu, Nov 13, 2008 at 7:18 PM, Alan Gauld
<[EMAIL PROTECTED]> wrote:
"Jojo Mwebaze" <[EMAIL PROTECTED]> wrote
Because we have very many such cases, we can not incorporate such
adhoc
changes in the system.. we are thinking of storing such classes in
the
database and have classes run from the database. if anyone else
feels they
need to use someone's algorithm, they can run it/or extract it from
the
database and run on it on their data.
Sorry if I'm missing the point but this sounds like a traditional
version
control system like CVS or SVN would do the job using the normal
python files.
Or is that too simple?
--
Alan Gauld
Author of the Learn to Program web site
http://www.freenetpages.co.uk/hp/alan.gauld
___
Tutor maillist - Tutor@python.org
http://mail.python.org/mailman/listinfo/tutor
___
Tutor maillist - Tutor@python.org
http://mail.python.org/mailman/listinfo/tutor
___
Tutor maillist - Tutor@python.org
http://mail.python.org/mailman/listinfo/tutor
