In anglosaxon accounting's stock valuation, the case where invoice and company currencies were different was not managed properly, because of a syntactic error in the condition checking this case.
Was PR #17259
The inner ’.s_parallax_bg’ element has a negative z-index value to be
rendered behind any other inner content.
This may cause the element disappear, because z-index is evaluated on
the entire page.
This commit will restrict the ’s_parallax_bg’ behaviour at snippet
level, assigning a z-order to the container too.
The `Background` template report (added in 3d04e4488) has been reported
in some instance to have CSS which could cause an empty page at the end
of the report.
This is caused by the following CSS rule which allow to have the
background at the bottom of the content:
.o_report_layout_background {
min-height: 1000px;
}
this is probably caused by a difference in header/footer height, font or
version of wkhtmltopdf (this has been reproduced on wkhtml 0.12.1).
A fix is hard to do for this and keeping the background at the bottom of
the report content because of how wkhtml behave. So to mitigate (and
hopefully it doesn't come back) the issue, this commit just change the
CSS min-height.
opw-726546
opw-743315
closes#17072
The filter on activity_date_deadline has been added during a fix, see
28c158cd13
It's incorrect for the computation of the 'Won in opportunities'
because it excludes opportunities not having activity_ids.
Furthermore, the way to compute the 'won' amount is not equals to the
filters when clicking on the button. Then, we need to keep only
opportunities standing in a kanban column having a probability of 100%.
opw-742489
Before this commit, the tour tips were overlapping the chat windows.
This was due to the fact they always had a z-index of 1051 wherever
they were. If a tip was placed in a location where none of its
ancestors created a new stacking context, the closed tip was then
positioned above everything else, including the chat windows.
A simple solution would have been the one proposed by this commit:
https://github.com/odoo/odoo/commit/856e78b35a47a8732e5bd4da1cce247df70e5414
However, as it explained, this would not be entirely correct. In the
website, elements are using z-indexes and would thus hide closed
tooltips when it is not wanted if those have no z-index at all.
The solution this commit is making is to force the creation of a
stacking context for tooltips' parent. The z-index will still exist
on all tips (closed and not closed) to force them above every other
elements, but now only elements of the same parent. This is indeed
more correct (but maybe still not entirely): website tooltip are
often placed in the body (as it is the element which scrolls) and
are thus still placed above elements like the navbar as their
z-index is left high. However, when placed in the appswitcher, they
will now be placed above appswitcher elements only (and thus not
above outside elements which are themself above the appswitcher
(e.g. the chat windows)).
opw-743330
in res.company write(), create a sub method _validate_fiscalyear_lock()
so that it is easier to inherit to modify or extend its
functionality.
Was PR #16615
Introduced by 8f39ad2be6 which was
intended for 9.0. It does not work well in 10.0 because here users
have the possibility to specify their own tracking numbers directly in
the POS interface. This causes issues when eg. the user specifies the
same lot number as is already reserved by the system.
opw-742657
Creating a mail template which is complex and compatible with all mail
clients is a very difficult task as each mail client supports a
different subset of basic CSS rules.
This commit fixes 3 problems which were occuring in outlook and/or
thunderbird but not in gmail:
- Main layout horizontal overflow
-> change 24px horizontal padding of main layout to 3% and set the
width to 94% (so that the sum is 100%)
- Fixed 600px content was full width instead
-> Change the .o_mail_wrapper to be a <table> element instead of a
<div> element. The table is full-width and contains one row of 3
cells: an empty left one, a fixed 600px-width center one and an
empty right one. This allows to have the same layout as before but
without using the unsupported max-width css rule.
- Texts were not using the right color and font-size
-> <p> elements were not properly inheriting the color and font-size
properties, so force the style on <p> elements.
opw-743156
If there are open rescue sessions there will be multiple opened
pos.session records with the same user.
A better solution would be to make the POS client send back the
pos.session id it belongs to and use that instead, but this would be
unsuitable for a stable release. It will instead be done in master.
opw-743318
In 4e0e376b0 an iframe for the readonly preview was added when
displaying html fields.
In e5112ee00 this iframe was only used if style-inline was set in the
widget options.
This brought an incompatibility if the not iframed preview contained an
iframe (eg. a youtube video which is then resize to a 30px height) which
is solved by this commit.
opw-725161
closes#16955
When calculating the cost price or the traceability between the consumed
and produced quants in a production order, we need to take into account
the extra moves that are added in the action_done (move_validate)
opw-715541
The slideshow image gallery code was quiet old and ugly. This commit
refactores it so that it works better, especially its responsiveness.
Before this commit, the number of thumbnail images was limited to 10.
If there were more images than 10, the user was able to navigate
between groups of 10 images thanks to arrow buttons. The problem is
obvious: on small devices, 10 images is too much. Now this "10" limit
is not fixed, it is determined according to the width of the browser
at the time the user loads the snippet.
Some animation and style was also wrong in some cases.
This commit was made as a 10.0 fix, the snippet should keep improving
in advanced versions.
Putting fixed width on button is obviously always a bad idea as
traduction may change the inner text size. Besides, removing the
fixed width does not change the design anyway.
Because when it does, it can create a cycle in the graph of
implied groups. This happens when the option 'Sort products
in sections with subtotals and page-breaks' is checked in the
Sales settings. In this case, being in group_user implies to be
in group_sale_layout, and being in group_sale_layout implies to
be in group_user.
This is a problem for groups display in the 'Users' form view: the
employee groups are no more linear, and thus no more displayed as
a selection field, but there is an 'Employee' section where each
of the employee group can be checked. Activating the option in the
Sales setting should have no impact of the groups hierarchy of HR.
Moreover, there have been several imilar issues on production
databases, probably due to this cycle, but that we haven't been
able to fully understand/reproduce yet: with this option in the
'Sales' settings activated, after a mysterious operation, the
'Users' form view was broken, because the selection field
'sel_groups_'for HR was still referenced in the view, but didn't
exist anymore (as HR groups weren't displayed as a selection field
anymore).
When the list view of 'ir_cron' is opened, the key 'active_test' is set
to False in the context to see all cron, even the archived ones.
The problem by passing 'active_test' in context is that it's propagated
to all actions launched from this view, for example whan you run
manually a cron, this key is passed to the cron action, which may leads
to call some actions on archived records which is not intended.
To fix this issue, we removed the key 'active_test' from context and
instead create a filter that show all crons set by default.
Commit 9d4efc8 added the sequence field on model
'purchase.order.line' and added this field in the
tree view (with 'handle' widget), to allow to
re-order the purchase order lines in the purchase
order form view.
However, for this to actually work, the default order
on the model should be on sequence field, otherwise
the rows will still be ordered by id even if the user
changes their sequence value by re-ordering them.
If an exception during the merges (such as a file descriptor overrun), we
would otherwise depend on the next garbage collection to close the
files. But the next GC may never come.
For example if we ran out of OS file descriptors during merge, all future
requests will crash for the same reason, and the process will never recover
because the GC will never run.
Much easier to explicitly close the files all the time.
can't be opened that way.
For now, only views that can be opened with a view manager are
supported, i.e. views like form, list, kanban, graph...
So this fix filters out the views of type qweb and search from the
list.
When creating a down payment for a SO, the applied taxes must
be the taxes of the company order. In this way, it's the same behavior
as a product with multiple taxes is added on a SO, only the order company
taxes are applied.
opw:741958
Let's consider the stockable product P with an internal category set
with cost_method = 'real' and 'average'. When creating a PO with P
and a unit_price=0, after delivering the product, the cost transfered to
the stock valuation account was equal to the product cost multplied by
the product quantity instead of the unit_price set on the PO multiplied
by the quantity.
PS: The quant is created with the stock.move in _quant_create_from_move and the unit_price
set on the stock.move is taken with function get_price_unit.
opw:740865
Steps to reproduce:
1. Create a payment of 999 with bank
2. Create a statement line of 1000 on bank journal
3. Go to reconcile interface, choose the 999 line and make an adjustment of 1
4. The payment has been reconciled with the statement line, and one new journal
entry has been created for adjustment of "1"
5. Cancel statement line (you need to enable this functionality on bank journal)
Bug:The payment journal entry of 999 is deleted and payment remains confirmed
After the fix: The payment journal entry of 999 is UNRECONCILED and the payment
entry still exists.
opw:741619
... records on models having a sequence field (e.g. project.task).
When a record is dragged and dropped, the following RPCs are done:
- write (the new group id on the moved record)
- read (the moved record)
- resequence (the column where the record has been moved to)
To perform the resequence RPC, we need to retrieve the list of ids
of the records being in the column. We do that by iterating over
the cards in the column and reading the id in their data attribute.
However, a traceback might occur if the resequence was triggered
exactly when the user was holding a record over the column being
resequenced, because jquery-ui creates a placeholder for the record
being dragged and dropped, and this placeholder doesn't have the
required data attribute.
The field `ssnid` is used on the payslip. However, it is not available
in any view, therefore it is impossible to fill it.
We replace by the `identification_id`, which is the field commonly used
for identification.
opw-742204
Closes#16798
- Activate multi-currency: EUR is main currency, USD is secondary
currency
- USD currency rate is:
1.2 on 2017-01-01
1.5 on 2017-02-01
- Create a journal in USD (e.g. Cash)
- Create a tax with Cash Basis activated
- Create an invoice of 100 USD on 2017-01-15
- Register a payment of 100 USD on this invoice on 2017-02-15
- 4 journal entries are created:
the payment itself
the exchange rate difference
the cash basis linked to the payment
the cash basis linked to the exchange rate difference
The cash basis linked to the exchange rate difference should not be
created, it is not necessary. The amounts are already included in the
cash basis linked to the payment.
opw-728767
Before this rev., there might be a crash when the user quickly
clicked several times on the browser's back button. This crash
occurred in the code saving/restoring the scroll position of the
current action.
When clicking fast on the back button, several actions are
executed but don't wait for each other (those actions are triggered
by the hashchange event), so when it is fast, and mainly when the
appswitcher is involved, it may happen that we try to get the
scroll position of an action that isn't loaded yet.
opw~741814
oco-odoo