Whenever we add or delete a channel in the bus/longpolling system, we
want the long poll to restart immediately, to be able to receive the
correct messages.
Before this commit, we had to wait for the longpoll timeout to finish
before receiving the correct messages.
We have a need to be able to abort a pending xmlhttprequest. This is
necessary for the longpoll system so it can restart properly.
With this commit, we simply propagate the abort method from the jquery
deferred to the deferred returned by the session rpc method.
Note that this is not documented on purpose. The reason is that we do
not want to commit on maintaining this feature.
It is kind of awkward to have a deferred which can be aborted, but once
it is chained, the abort feature is lost. So, for now, this will stay a
lower level feature used exclusively for our needs.
The commit [3079415fb17232117b0d5059a23ae89232eaf5e8] allows the read_group groupby parameter not to be contained in the fields parameter. This makes possible the simplification made here. Similar changes could certainly be applied elsewhere.
For historical reasons, a field in the read_group grouby parameter was required to be in the list of fields parameter without any necessity. This is not the case anymore. A byproduct of this commit is that it fixes a bug in pivot view introduced earlier in [47395c6].
Before installing pads, create a second company
On the first company, create a task and initialize its pad
Switch companies and try to edit the task
Before this commit:
There is a traceback because since the pad credentials of the second company are not set,
the url cannot be computed
After this commit, we prevent the Null field to concat with a string
hence there is no traceback
OPW 1837059
closes#24263
On a DB with thousands of events including recurring ones, fetching the
complete list of events might lead to a memory overflow due to the
prefetching of all virtual events.
We reduce the memory footprint by handling no more than 50,000 records
at a time and clearing the cache at each step. This limit is purely
arbitrary, but seems to be a good compromise between performance and
memory usage.
opw-1832482
Before this commit, whenever we do an operation on any of these custom views,
such as folding it, the custom view was saved by creating a new one.
This behaviour comes from a lost feature to undo operations on custom views.
Since we do not have this feature anymore, it makes no sense to create new
custom views on save, instead of applying the changes in place.
With this commit, custom views are edited in place.
Fixes#23712
Before this commit, whenever we do an operation on any of these custom views,
such as folding it, the custom view was saved by creating a new one.
This behaviour comes from a lost feature to undo operations on custom views.
Since we do not have this feature anymore, it makes no sense to create new
custom views on save, instead of applying the changes in place.
With this commit, custom views are edited in place.
Fixes#23712
When we make a payment, we need the write access on account.journal
because check_manual_sequencing on account.payment is a related field
of account.journal and was used to decide on numbering the check or not.
With this change, the write access is not required.
opw-1834343
opw-1835888
closes#24281
before this fix, hitting the ALT key did not show the shortcuts on the
buttons
after this fix, the events of the web_client are extending those of
abstract_web_clients where the keyboard navigation is hooked up, so the
shortcuts are visible
Before this fix, the focus was not set when opening a kanban view
After this fix, the focus is set on the search view (unless there is
a quick create, in which case the focus will be on the quick create)
In the reference widget, the focus was always on the input (many2one) even
though the first select (with the model) was not set.
Thanks to @kaerdsar for the original PR.
Partially closes#24070
The website sale wishlist tour checks if you are correctly redirected to the
cart.
To do so, it checks if the cart title 'Shopping Cart' appears on the screen.
But since there is a new checkout in saas-11.3 with 006ce9c88 that removed this
title, we need to trigger on something else.
Before this rev., in the ir.model.fields form view, it crashed
when the user specified an invalid model name as a relation for
fields of type many2many (the onchange crashed). This rev. displays
a warning instead.
opw 1837401
Instead of:
1. testing the existance
2. creating the directory if not exists
(which would not work if a concurrent creation is done between step 1 and 2)
do:
1. try to create the directory
2. skip if it fails
Ask forgiveness, not permission
Closes#24210
On a pivot view expand a field in column
another one in row
Then put a domain in the search view
Save the whole thing (ir.filter) as favorite
Before this commit, the ir.filter was not saved the right way when a domain was present
(Let alone its restoration)
This was because of
- the internal default groupbys of the pivot model was not updated, and what we got in the params was overwritten
by what existed before
- a slight typo in the code made the row updated with the col values
OPW 1835865
OPW 1829925
closes#24215
This is confusing for end users because on the fleet form/kanban view the
model is displayed. The information displayed in the name is the default
depreciated cost on the model. But in reality the depreciated cost of the
car could be different.
When an unlogged user creates a livechat, typing the /help command in the
backend would result in the help command sending a traceback.
The function processing the command would assume the existence of a channel
partner name, which is not the case when the user starting the chat is unlogged.
opw:1835484
When producing a byproduct in a BoM, the quantity for the byproduct was
increased by twice its produced quantity.
This is essentially a revert of commit d7fdb636b8
which was erroneously forward-ported to 11.0,
where byproducts were already correctly counted.
opw:1834226
Before this fix, when a user don't have write access on a document, he
can't send an email with attachment because, the attachments is added on
the document.
The change improving security was introduced by the commit:
https://github.com/odoo/odoo/commit/6494f511718893eec3573c60c0a62e04d386359d
With this fix, for mail, the attachments is added on the
'mail.compose.message' and re-render the fields attachments displayed in
the current view. This change respect the access rules like all others
attachment of 'mail.compose.message' (can send an email when the user have
read access on the document).
The media dialog of the html widget display all documents attachments and
the added attachments Allowing users to use these in the body of their
message.
Here is a scenario that could happen before this commit:
1. web client is started
2. it binds a callback on hashchange
3. it performs some async work (read initial action for user)
4. the user (or more likely, a tour) changes the url
5. the hashchange handler is called, and it starts a new action
6. the async work for the web client is over, it then executes the
default action and/or clicks on the first menu
7. we are now in the default action state, instead of the desired action
state (the one from the new url)
With this fix, we simply do nothing when there was an hashchange before
the initial action was done.
Here is a scenario that could happen before this commit:
1. web client is started
2. it binds a callback on hashchange
3. it performs some async work (read initial action for user)
4. the user (or more likely, a tour) changes the url
5. the hashchange handler is called, and it starts a new action
6. the async work for the web client is over, it then executes the
default action and/or clicks on the first menu
7. we are now in the default action state, instead of the desired action
state (the one from the new url)
With this fix, we simply do nothing when there was an hashchange before
the initial action was done.
When sending the test email of a mass mailing campaign, it might be
observed that the result is different from the final email sent to
customers.
This is because the `body` field of the final mail message is an `Html`
field with options `sanitize_style=True` and `strip_classes=True`. On
the other hand, the `body_html` field the test mail is a simple `Text`
field which is not sanitized.
We sanitize the `body_html` field with similar options.
opw-1820064
Here is a scenario that could happen before this commit:
1. web client is started
2. it binds a callback on hashchange
3. it performs some async work (read initial action for user)
4. the user (or more likely, a tour) changes the url
5. the hashchange handler is called, and it starts a new action
6. the async work for the web client is over, it then executes the
default action and/or clicks on the first menu
7. we are now in the default action state, instead of the desired action
state (the one from the new url)
With this fix, we simply do nothing when there was an hashchange before
the initial action was done.
As for commit 3f975f9769, the partner_type field is readonly and cannot be modified when not in draft.
It makes the form invalid, so form buttons cannot be used, e.g. to access the journal items, etc.
Christophe Matthieu