this is done by upgrader in production, so it's consistent to do it
before test

slapos_cloud:testSlapOSCloud fails with PYTHONHASHSEED=385

This is no stable with hash randomization, with PYTHONHASHSEED=593 this
test failed.

With the previous approach, the test depend on the order of attributes,
but this order is not stable when hash randomization is enabled and we
don't care about order of attributes in this test.

Since ERP5 commit 6546e47f51 (erp5_software_pdm: a Software Product is
a Service, 2022-09-07), the message changed

same as nexedi/erp5@2e366054

It's been a while the server is ready.

See erp5@c45c2295

   Replace it with an generic action directly pushed to erp5 (erp5_forge)

See merge request nexedi/slapos.core!606

   This prevent raise in case some non-ascii char is passed into the string.

  We only expand by the end of the period, so we dont generate an invoice 15 days in advance anymore

- make url script return no url if there are not connection parameters
- update tests accordingly

   ...to failures on password generation

See merge request nexedi/slapos.core!604

See merge request nexedi/slapos.core!603

   The password has to work accordinally the policy, if fail retry a few times.

   There is no more patch

  backport from ERP5

   In past, we used to rely on ERP5 Login (passwordless) for authentication via Certificates, but we switched to use excusively a dedicated portal type (Certificate Login).

   The ERP5 Login now follows the default implementation, so no need custom scripts.

   This helps to figure out what it is really testing.

  Some scripts were removed and on other cases started to be tested.

   After 6 years, it is expected that all instances are migrated. The migration is outdated, since it we use Certificate Login rather them ERP5 Login on Compute Nodes and Software Instances.

    There is no need to pre-create it, since it would be required you go throw External service (Facebook or Google) to create the default connectors.