so that unindexObject SQL query are invoked after another MariaDB transaction for SELECT is committed to avoid deadlock.

because now Alarm_zGetAlarmDate and Alarm_zUpdateAlarmDate use different SQL connection, thus different MariaDB transaction.

now catalog update and catalog search use different SQL connection, thus different MariaDB transactions.

otherwise testSyncML fails.

In nexedi/erp5@39369169 the new option
'nofile' was added to set the soft limit of open file descriptors to its hard limit.
In nexedi/slapos!1451 (comment 193296) it was criticized that
the name of the option - 'nofile' - is too vague and in nexedi/slapos!1465 a new
name for the option was searched. Finally we agreed on the more explicit 'with-max-rlimit-nofile'.

See merge request nexedi/erp5!1843

  This gadget aims to edit json based on json schema.

See merge request nexedi/erp5!1853

since for column like 'delivery.start_date'
the field would be like 'listbox_delivery_start_date'

This was missing from f3435570 (ui_test: update
verifyImageMatchSnapshot screenshots for firefox 115, 2023-10-22)

See merge request nexedi/erp5!1840

See merge request nexedi/erp5!1835

See merge request nexedi/erp5!1841

See merge request nexedi/erp5!1831

See !1850

Pause to make sure the server has time to process the request.
This is of course not a good fix, but the future of this code is really
uncertain, so this should at least fix the failing test.

When account number method is set to account reference and the
transaction line has a node acquired from the invoice source or
destination organisation or person and when this node has a reference,
then this was displayed as an account reference.

We set a reference on Client 1 organisation to reproduce the problem.

Make it part of the caching method id so that changing the preference
immediately show new accounts.

This also moves the "user_name" argument that was an artificial parameter
not used only here to make the cache vary to the cache id, for
consistency and clarity.

This fixes dict formatting which was formatted like this with too long
values:

  {
    "a":
    "b"
  }

after we updated cloudooo.erp5.net with nexedi/cloudooo!32

See merge request nexedi/erp5!1785

This method raises the same error as jwt's claim validator, as it was intended to be
called from the same exception handling block. So move it from "else" blocks into
the corresponding "try" block.

  We must ensure that some Web Section (under a website with must-revalidate caching policy) can have a session that isn't stored.

  Different from no-cache (that is similar to must-revalidate + max-age=0), no-store tells the client (and frontends) that it must not cache the content ever.

  Since we issue an no-store and no 304s, the Last-Modified could confuse the frontend if If-Modified-Since header is present on the request.

  If-Modified-Since is sent, the ERP5 wont return with 304, so we don't expect that the frontend does it (but it does), so since we don't expect that user stores the last request it makes no sense to return with Last-Modified.

A Telephone type object is defined by its number, and it's country (ie: country code). See Telephone portal type and property sheet.
So here, having only the number is incomplete.

using pdfjs-3.11.174-legacy-dist.zip

Note: viewer.js is modified to do nothing in webViewerLoad (one line
with return "ERP5 patch: disable here" is added at the top of this
function) because we do the initialization in the gadget.
I think something similar was also done with the previous version.

The tests were extended to cover:

 - opening an empty or broken PDF: this should not crash
 - clicking save: this should not crash and the updated document should
   render properly. The save action replaces the ERP5 document content
   with the content from the editor, so if user for example added
   annotations or draw on the PDF, the modified version is saved.
   We don't use this feature and it's not so easy to test, so for now
   the test only covers the "it should not crash and it should not
   break the document".
 - entering wrong password for encrypted PDFs should not crash

This was working fine (although not python3 ready), but was not tested