This reverts commit 436bbb0a.

The many2many widget is usefull, as it allows to select an existing picking instead of creating a new one when clicking add an item from the list

As it sets the standard_price field, which is defined on product templates only: setting the standard price of a product variant will lead to set the standard price to all the variants of the product template.

There is therefore no meaning to compute and set the standard price of product variants according to the boms

To allow customization, such as do not send the notification to the partner that subscribed himself to the thread (he probably already read the thread, no need to send him an email)

It wasn't possible to delete a user if he had a im chat presence linked to him.

Same problem as before: filters do not compose properly with the implicit
'&', and they are filtering on the wrong model (should filter on probability
in the crm.case.stage model and not on the crm.opportunity.report)

* use static imports in java examples to make them terser
* inline ``domain`` in java and php example to make examples more
  self-contained
* try to extend/improve Model.write's docstring
* add convenience kwarg to fields_get, mostly for user-driven
  introspection

Closes #3689

commit f76d4525 was not actually working: extra keys from
config files are not yet into the config options dict at
import time. The fix is to move the logic inside the method,
like in `find_pg_tool` just below.

Also fix the use of `find_in_path` in report.py: the subprocess
may also raise AttributeError exception, so instead of listing
all the possible ones just re-raise the IOError shallowed by
`find_in_path` when the result is None.

Fixes #3809 #3811

The openerp-server.conf now generates the bin_path record, in order
to resolve calls to external binaries served in the thirdparty dir.

Adpated report.py to use find_in_path and not directly which.

When moving fields name -> provider on payment.acquire, the condition in payment_transfer was not updated.
This lead to no post_msg value in the Wired Transfert acquire.
Fixes #2423, opw 613934

[FIX] google_calendar: update tuto to create credential for google Calendar API. Google changes suddenly the process. Before that was optional, now without it, customers have an error 400 redirect_uri_mismatch

The line being deleted in this revision looks to have been useful when it was introduced in this commit:
36fc910e

As the sale order was updated right away through the update_pricelist method
But since this rev. 22f4c315, the sale order is updated later, and reset the sale_order_code_pricelist_id value in the session right after setting it prevent to apply the pricelist of the promotional code...

[FIX] web_calendar: improve the commit 787147d5. This commit was fixing bug in other module that calendar which was open an event in popup view. Because browse/read by default was unable to read string id. The fix was too restrictif, because calendar need to read/browse string; we cannot cast this id in Int else we loose the virtual part. parseInt('12-20140512') -> 12.

[FIX] calendar: allow to sort in tree view the starting date (start, start_date, start_datetime) (hack for V8)

Problem was that when the user types quickly in the search bar and press
enter, the keydown event of the enter key happens before the keypress
event of the last key entered.  This means that the autocompletion has
a wrong string.  The fix is to move the enter selection detection from
keydown to keyup.

The calendar module generates string values with
a date/time formatted according to the user
language. Those formats may contain non-ascii
characters and are read as unicode strings,
but fed to str{p,t}time, which only accepts
byte strings (in Python 2).

This would cause an exception when loading calendar
notifications for a user using e.g. Chinese with
some CJK unicode chars in the date/time format.

There was two problems:
* it was filtering on the probability of the opportunity, not on the
probability of the stage
* the filters were not prefixed with '&', which means that they gave
weird results when combined in the searchview (in a filter group, filters
are combined with '|', but it breaks when the filters are defined by
an implicit '&')

closes #3761

Fixes #3730

This rev. 73072272 ensured to not (re-)set the state 'confirmed' to exploded moves with a more advanced state (for instance, 'assigned')
Nevertheless, the location chaining is performed on the move confirmation, through the action_confirm method of the stock.move model. Besides, the resulting moves of the _action_explode method had the state 'confirmed' on creation, the 'confirmed' state wasn't set by the method 'action_confirm', meaning that the moves were confirmed without having the location chaining done. Allowing moves to go through the action_confirm method even if the state was 'confirmed' or further triggered the location chaining.

Preventing already confirmed moves to go through the action_confirm method prevented the location chaining, thus.

We now create the resulting moves with the 'draft' state, and then confirm them through the procurement workflow signal 'button_confirm'. Thus, the resulting moves are confirmed by going through the action_confirm method, writing the confirmed state and triggering the location chaining at the same time. We then write the 'assigned' state if necessary.

opw-617235