Subscription request workflow uses simulation_state rather them validation one.

   What happens when user don't pay reservation.

     Since we use Caddy rather them apache by default, change the field to input a value on tests.

slapos_subscription_request: As dates are simetrical on Invoice and Payment, the value is not negative anymore

   This with price not zero is covered by another tests.

   Failures:

      portal_skins/slapos_subscription_request/SubscriptionRequest_checkRelatedAccounting:5:2:Dangerous default value {} as argument (dangerous-default-value)
      portal_skins/slapos_subscription_request/SubscriptionRequest_checkRelatedAccounting:15:2:Unused variable 'refund_service' (unused-variable)

   The input_movement_list weren't properly filtering causing grouping references been set with it's late occurency.

   It makes requests as non-subscription and for subscription cases, and assert if the invoice is correctly linked via simulation

See merge request !277

  This is probably required multiple renames or upgrade project code base

  This fixes name convention tests:

     /portal_skins/slapos_crm_monitoring/SupportRequestModule_viewExportMonitorOPMLList
     'message': 'erp5-Guideline.Form.Report.Dialog.Is.Postfixed.With.Dialog: Dialog form name must follow ${portal_type}_view.*Dialog naming'

     /portal_skins/slapos_crm_monitoring/SupportRequestModule_viewInstanceMessageList
     'message': 'erp5-Guideline.Form.Report.Dialog.Is.Postfixed.With.Dialog: Dialog form name must follow ${portal_type}_view.*Dialog naming'

     portal_skins/slapos_crm_monitoring/SupportRequestModule_viewSupportRequestMonitorLinkList
       'message': 'erp5-Guideline.Report.Name.Uses.Portal.Type.Followed.By.View.Report.Name.And.Report: Report form name must follow ${portal_type}_view.*Report naming'

Because some programs did not have proper configuration for log files, they
were using automatically generated file names for log files. Correctly
define stdout_logfile configuration for all programs, used with
redirect_stderr=true, so that both stderr and stdout are in the same log
file.

  This is a safe measure in case the alarm to delete the instance is temporarily disabled.