openerp-community team mailing list archive

[Ferdinand <office@xxxxxxxxxx>]

On 2014-01-14 18:17, Enapps :: Vadim Chobanu wrote:

Hi
we had and expect a similar performance of server load as for any other 
database fields which are not indexed and thus can be sorted and filtered.
the "only" difference is that for a database field the work is done by 
postgres (after reading all records) and for non db field this is done 
by python AFTER evaluation of the function field.
the time consuming part is the evaluation for the function field.
if this becomes a performance problem - other solutions (aggregates) 
must be implemented.

but
* many of the most common tasks do not involve 1m records - although the 
underlying resources may be of high volume (Example: show all products 
with qty !=0 for inventory per stock location - or print accounts with 
balance !=0)
IMO the argument is "wrong" - if the data is not filtered / sorted in a 
human understandable way, we may preserve CPU resources, but spoil much 
more expensive and scarce human resources and this creates bad user 
experience.
* a group by is sometimes helpful for a handful of items although the 
grouping is done by id and not ordered by name
* the (only) work around  is to make a csv export and sort/filter the 
data in a spread sheet which involves the computation of the function 
field anyhow. I doubt that this is what users expect and that this is 
less time consuming.

hope that helps
Ferdinand
>   Hi,
> Quite interesting.
> Would you happen to have any ideas on how this will work for 'not so
> small' volumes of data? Lets say a company with 5000 products and 1mil
> stock moves.
> If users would go to products list and click on order by quantity, what
> kind of wait time are we talking about
> Thsanks,
>
> Vadim
>
>
>
> -----Original Message-----
> From: Openerp-community
> [mailto:openerp-community-bounces+vadim=enapps.co.uk@xxxxxxxxxxxxxxxxxxx]
> On Behalf Of Office ChriCar
> Sent: 14 January 2014 10:52
> To: Alexandre Fayolle; openerp-community@xxxxxxxxxxxxxxxxxxx;
> qdp@xxxxxxxxxxx
> Subject: Re: [Openerp-community] sortable / searchable stock levels (Re:
> New WMS. Call for testers.)
>
> On 01/14/2014 11:30 AM, Alexandre Fayolle wrote:
> > Hi,
> >
> > One feature I was hoping to finally get in the WMS rewrite is
> > searchable / sortable stock levels.
> >
> > The typical use case our customers have is to be able to filter out
> > products with Quantity on Hand == 0 when displaying stock by location.
> > Sorting by quantity on hand is also a quick way to see which products
> > have a negative stock level (and therefore an inventory problem). This
> > is not possible in oerp <= 7 without custom addons. Is it planned for
> v8?
> searchable and sortable function fields (non database fields)
>
> I have funded and Panos has  provided a server 6.0 side patch (attached)
> for sorting and searching of non db fields.
> If OpenERP does not accept this it could now finally go into OCB.
> IMO it is impossible to explain to users why some fields are
> searchable/sortable and some not and client side (browser) sorting does
> not make sens as long the client deals only with subsets (80 records or
> so)
>
>
> as well as one2many_sorted (using translated terms - what users expect)
> http://bazaar.launchpad.net/~camptocamp/c2c-rd-addons/7.0/files/head:/one2
> many_sorted/
> <http://bazaar.launchpad.net/%7Ecamptocamp/c2c-rd-addons/7.0/files/head:/o
> ne2many_sorted/>
> this allows different sorts in different o2m situations (stock_move,
> account_move_lines,....)
>
>
>
> --
> Beste Grüße
> Ferdiand Gassauer
> ChriCar Bet. und Ber. GmbH
>
> _______________________________________________
> Mailing list: https://launchpad.net/~openerp-community
> Post to     : openerp-community@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~openerp-community
> More help   : https://help.launchpad.net/ListHelp


--
Ferdinand