When switching between records, the displayed thread was still the one of the
previous records until the messages of the current record were loaded, which
could be confusing.
Moreover, when quickly browsing records using the pager, it could happen that
the resulting displayed thread wasn't the one of the displayed record (if
we visited a record with a long thread history, the RPC to load that history
might return after the one of the current record).
This commit aims to always synchronize the displayed thread with the current
record, by
- directly hide the previous thread when browsing records
- displaying a spinner after 1s to indicate that the thread is loading and
(possibly) not empty
- ensure to only render the thread of the last visited record
The goal of this commit is to have better search results when fuzzy
searching. Before this commit, the position of a match was irrelevant
to the search results. After, a match in a first position has more
weight than a match in last position. This is better, because we often
want to find results using the first letters.
This is controversial, but I could not find a better way to do this
without rewriting the whole library myself (which might not be a bad
idea, it is only 50 lines long)
In a company, there might be several contract types, e.g.:
- 38 Hours/Week, 5 Days/Week
- 45 Hours/Week, 5 Days/Week
- 24 Hours/Week, 2 Days/Week (work on week-end only)
Therefore, it is necessary to be able to define the average number of
hours per day on a resource calendar.
available in channels and chats, but not livechats as the public user isn't
really "in" the channel (the command would reply that you are alone in the
channel, which isn't necessarily the case)
Use the mention mechanism of the composer to autocomplete on canned responses
(keyword :) with fuzzy search.
Directly perform the string replacement in the composer, instead of
server-side as before.
Add a menu item to access an editable list view to create and edit canned
responses.
Add two canned responses as demo data.
to prevent crashes if the user types, e.g. a \, in the partner's autocompletes
in Discuss ('Add a DM' in sidebar and 'Invite people to channel' dialog).
A 200ms delay is set on lookups for mention suggestions to prevent triggering
RPCs when the user is still typing. However, there shouldn't be a delay when
the suggestions are looked up in the prefetched suggestions.
Also fixes a bug that might occur when the user starts typing a mention which
triggers an RPC, but erase it before the RPC returns.
The BasicComposer widget is now used in chat windows, which enables the mention
feature. This is only for the backend, as for the livechat windows in the
frontend, the composer is still a simple input.
Classname 'o_chat_input' has been renamed into 'o_chat_composer' (the root
classname of the basic composer) for uniformity in the JS and LESS code of
chat windows.
* removed hr.timesheet.report as it's mostly redundant with regular
views on analytic lines
* moved "detailed activities" to reporting, renamed it, added a pivot
view
* redirected timesheet link from HR dashboard to the detailed activities
action
- added department_id to analytic lines as that link is filtered by
department
- fixed bug in kanban view: button context wasn't evaluated in the
correct context
It's about 'Receipt Reference code' which gets generated when Order is
created from POS .
After this, click on 'RETURN PRODUCTS' button, the return Order doesn't
fetch the "Receipt Ref" of the origin Order, so there is no way to find
the origin order from where the return of the products done.
- Y axis : according to 12 hour or 24 hour format
- Event title : following time format in preferenced language
NB: The conversion from 'A' to 'TT' to match fullcalendar v1 should be
removed when fullcalendar is updated to v2, as the syntax will change.
See : http://fullcalendar.io/wiki/Upgrading-to-v2/
When migrating stock, some dummies methods have been removed. Indeed those
calls make more problems than they solve by adding unnecessary complexity
around simple methods. However it seems some calls have been left in the
process :
- _get_act_window_dict which is basically an env.ref or act_window.for_xml_id
- _get_products which is basically having all product_ids from template by
iterating on product_variant_ids.
Wizard models are migrated to new API. No change is performed to code
structure or functional structure. Some code cleaning may be performed when
migrating to simplify the code. Other files may be updated due to the move
towards the new API.
Incoterms model is migrated to new API. No change is performed to code
structure or functional structure. Some code cleaning may be performed when
migrating to simplify the code. Other files may be updated due to the move
towards the new API.
Incoterms models and views are moved in their own file. Impacted:
- stock.incoterms
This commit only moves files. No change is done.
This last commit is done by renaming the remaining code in stock.py file. We
hope this helps keeping a bit of history by being able to retrieve the original
stock.py file.
Procurement model is migrated to new API. No change is performed to code
structure or functional structure. Some code cleaning may be performed when
migrating to simplify the code. Other files may be updated due to the move
towards the new API.
Product models are migrated to new API. No change is performed to code
structure or functional structure. Some code cleaning may be performed when
migrating to simplify the code. Other files may be updated due to the move
towards the new API.
Impacted models
- product.product
- product.template
- product.removal
- product.putaway
- stock.fixed.putaway.strat
Production lot model is migrated to new API. No change is performed to code
structure or functional structure. Some code cleaning may be performed when
migrating to simplify the code. Other files may be updated due to the move
towards the new API.
pvy