This rev. is similar to db9b91ff, but treats the x2many case.
Calls to the 'commitChange' function of renderers must not be done
inside a mutex, mainly since 241704536, as they could end up in a
deadlock. For x2manys, it was actually the case in practice in
Inventory > Operations > Inventory Adjustments > Create > select
'One product only' > select a product > try to edit the real
quantity in the editable list > focusout > deadlock.
The reason why the deadlock occurred in this view and not in, e.g.
sale.order, is due to the presence of the field_float_scannable
widget. Indeed, the barcode engine intercepts input events and
manually re-triggers them. As a consequence, the 'change' event is
not triggered on the field, and it does not notify the environment
of its new value by itself, until the field x2many asks it to
commit its value. Note that this might be an issue as well, as
onchanges are thus not triggered.
Task 1878254
Rev. f241441 recently added the possibiliy to define custom
controls (instead of the single 'Add a line') in editable list
views. However, it didn't take into account keyboard navigation.
Suppose that the focus is on the field preceeding the list, and
the user clicks on TAB, the code assumed that there was only one
control, and focused it. However, when there were several control,
the last one was focused (which is typically the less common one),
and there was no way to navigate to the others.
This rev. focuses the first control by default, and allows to
navigate through controls with RIGHT and LEFT. Pressing TAB still
moves out of the list, to the next field.
Note that i'm not happy with this fix, as the logic doesn't seem
to be in the right place. Indeed, the notion of controls is the
concern of the editable list, not the x2many field, so the list
should handle this logic by itself. Unfortenately, this is not that
easy to change, as for now the list doesn't even know that we are
navigating to/from its control(s), until we click on it/them.
Changing that will be for later, in master.
We also did a linting pass on the concerned code.
Task 1878254
We're not shipping the sourcemap files, and while the assets minifier
strips out the mappings they're getting hit in debug=assets which is
bot useless and problematic when running odoo-bin without a proxy for
static folders: if sourcemaps are enabled (which is apparently the
default in all browsers if devtools are opened at this point) the
browser tries to fetch the sourcemap, which does through the
SharedDataMiddleware which doesn't find them and passes the query on
to the regular Application which goes through the entire dispatch &
NotFound process.
If website is installed, that process ends up rendering website.404,
which can be pretty costly until everything is properly cached:
# initial request
"GET /web/static/lib/bootstrap/js/index.js.map HTTP/1.1" 404 - 319 0.198 0.954
# a few requests later
"GET /web/static/lib/bootstrap/js/index.js.map HTTP/1.1" 404 - 27 0.022 0.060
And the way sourcemaps are fetched (might be headers, might be the
stampede as browsers will try to fetch a dozen sourcemaps as fast as
possible) seems to make this problem much, much worse: instead of 300+
the requests take 800+ queries each, multiple seconds, and requests
get worse as time goes on (didn't investigate the exact reason for
that) *and* they apparently don't ever get cached (at least they don't
after half a dozen reloads of the client).
This reverts commit c539311. The default taxes will now be found in the invoice's onchange() which makes more sense and will work on all income/expense accounts without having to set the taxes on all these accounts.
This will be used to add the 'Upload' button on the tree view of vendor bills, by extending the regular view.
Was part of PR #26740
Was part of task #1880089
This is done by adding a new field on mail.thread objects and setting it when posting a message with attachments on a model inheriting from mail.thread.
Was part of PR #26740
Was part of task #1880089
This commit fixes the handling of the service parameter
in a mocked environment:
- updated docstring: now an object (previously an array)
- no longer used by mocked server (unused parameter)
- enforces default value for this parameter (empty object)
The test was failing because it was using an old approach to mock
services in JS tests. Also, it provides ajax service in the list
of services, which was not supported in a mocked environment.
With this commit, when an ajax service is provided in the test,
it will use this service instead of using mocked RPCs from the
mocked server.
Changes made at #23879 changed the dependency of l10n_cn to include a large
list of cities.
Move to a different module to make the l10n_cn more lightweight.
Closes#26592
When filling the results of a read_group using
_read_group_fill_temporal we were using the name of the field.
(https://github.com/odoo/odoo/commit/283c842289f1bc06bc30a877864e868127418e80)
But _read_group_format_result needs the name of the groupy which can
differ from the name of the field as groupby supports the following
syntax for dates `field:groupby`
This ultimately lead to a crash when trying to groupby multiple date
fields in a graph view.
Linked to Tasks : #1879604 and #1879168
The xpath on the res config settings view was attached to a div added in the
account module but the record was inheriting from base and not account.
As snailmail account inherits from account, it was somehow working.
* website, website_sale, website_sale_delivery
While we used BS3, we used our own responsive classes for text
alignment (text-right-sm, text-center-not-lg, ...). Now BS4 classes can
be used to achieve the same behavior so there is no point in declaring
our own ones anymore.
Also, as the breakpoint names changed (lg -> xl, md -> lg, ...), the
current code was not working as intended anymore anyway.
* website_sale
Bootstrap chose not to implement this feature, probably relying on their
grid system (see https://github.com/twbs/bootstrap/issues/21943).
We have a case where a responsive size of 0px is needed though: applied
on a table, this forces the table to its contents' natural width. So
this commit implements the behavior and adds the size of 0 to replace
the non-responsive class 'o_table_column_fit'.
Since d5b687a an error is logged when trying to insert duplicates in
ir_model_data. An error appears when installing this module because of
the duplicate.
As this module is not yet tested, this error flew under the radar.
Note: a new test that installs l10n* is developped
get the product_name thanks to a function, so that we can reuse it somewhere else (account_predictive_account)
Part of PR #26876.
Part of task #1882329
Show the source and destination location in the detailed operation view
if the multilocation is active. Also hide locations if only the delivery
package settings.
Issue: 1882695
The group_by was passed from context from the action spawning a new view.
This meant that it could give a group_by a field that was not defined on the
destination view, causing a js traceback.
opw 1884035
Modal generic:
- Fix footer overflowing buttons by pushing them to a new line.
Page manager:
- Add an SEO column to easily see which pages have incomplete SEO.
- Add an edit SEO column to navigate to the SEO modal on said page.
- Hide "edit in backend" column as debug.
- Generally improve the whole view by adding appropriate titles on icons, etc.
SEO modal:
- Improve general style.
- Add description length alert.
- Restructure the keyword table/suggestions.
- Reset page meta to their initial value on Discard.
- Various small usability improvements.
PR: #26897
task-1850579
Rumor say that that 1x/100... the variant is not selected and that the tour crash.
So we are not sure that it will avoid that, but it will be not worse.
I will keep my fingers crossed for that.
Before this commit, it was not possible to open systray menus in mobile.
This is due to popper.js that is wrongly dynamically positioning the
dropdown menu.
This commit fixes the issue by disabling dynamic positioning by popper.js
for the systray menus, so that it uses ou custom styles for dropdown menu.
Closes#26434
Revision on https://github.com/odoo/odoo/commit/d91ec258e6a31153beb9eb16cfe8f916546b7688
The commit above reverts a commit when we were using bootstrap 3.
The equivalent of XS in bootstrap 3 is SM in bootstrap 4.
So the correct revert with bootstrap 4 is to use SM.
Also removing useless style on 'o_mobile_menu_opened'.
Closes#26434
Tests from hr_recruitments work luckily until now.
Because module maintenance imply base.group_user for group_equipment_manager
<record id="group_equipment_manager" model="res.groups">
...
<field name="implied_ids" eval="[(4, ref('base.group_user'))]"/>
</record>
and module hr_maintenante imply group_equipment_manager for hr_user
<record id="hr.group_hr_user" model="res.groups">
<field name="implied_ids" eval="[(4, ref('maintenance.group_equipment_manager'))]"/>
</record>
After this commit launch test on hr_recruitment module without any other module will works.
Design-theme repo show this error because repo tests hr_recruitment without installing maintenance
(Repo install website_* modules -> website_hr_recruitment -> hr_recruitment -> hr)
Backport-of d5b0f99943 to have runbot design-theme green
When choosing a project template on a service (product
config), and then archive it, the generated project
on sale line confirmation will be inactive too.
This commit forces the generated project to be active.
Task #1884328
Aaron Bohy