Setting a PIN containing < or > silently cleared it: the length check ran before escapeSql(), whose strip_tags() then emptied the value, and the UPDATE stored the blank while flashing success. Length is now checked after sanitising.
Password and PIN changes require the current password. SSO contacts are exempt - no local password to check, and the IdP already did it.
New index on logs(log_user_id, log_client_id) for the portal profile and activity pages, which were scanning the whole table twice per profile view. admin/audit_logs.php's date filter rewritten as a half-open range so KEY log_created_at is usable - DATE(log_created_at) BETWEEN made it non-sargable.
Portal statement page and PDF now render in the client's currency, matching the guest view and the emailed statement.
Quick Send asks for confirmation; confirm-link extended to submit buttons.
Portal audit entries logged an empty name - client/post.php used , which only exists agent-side.
Send Email opens a contact picker instead of hardcoding the primary contact; Mark Sent asks how it went out and takes a note. Both log recipients / method / note to the document history. Handlers moved GET -> POST.
Fixes two pre-existing bugs in the email path: archived contacts were still receiving billing copies, and every copy was addressed to the primary contact by name.
None of the *_client_id columns carried an index, so agent/includes/inc_all_client.php was doing a full table scan per badge count before the page rendered a byte - 27 of them. The main agent context runs 8 counts and the admin context none, which is why the client side nav was the only one that felt slow.
24 indexes, shaped (<entity>_client_id, <entity>_archived_at) where the query pairs the two and the client column alone otherwise. The archived column is always trailing and never an index of its own - on a healthy instance nearly every row has it NULL, so it cannot lead. recurring_tickets, services, invoices, recurring_invoices and calendar_events get a single column because their queries carry no archived filter; invoices is in that group because the money total there spans archived rows on purpose. payments.payment_invoice_id and transfers.transfer_revenue_id are join keys rather than client scoping but sit on the same page's critical path.
Migration checks information_schema.STATISTICS before each ALTER, since MySQL has no ADD INDEX IF NOT EXISTS and re-adding is an error - safe to re-run and safe where an index was added by hand. db.sql carries the same keys so fresh installs skip the update step.
Wider than the side nav: these are the columns clientScopeSql() filters on, so every client-scoped list page was paying for their absence. Expect the ALTERs to be the slow part of the upgrade on a mature instance.
Backups are now AES-256 encrypted zips in three types (full, database
only, master key), catalogued in a new backups table, built by cron rather
than the web request, and kept under uploads/backups with retention in the
nightly job. The encryption key is one value per install held in config.php,
never in the database and never in the file name.
Restore is shared by the setup wizard and the new scripts/restore_cli.php,
which is the only path without an upload size limit. It verifies the key and
unpacks the archive before dropping anything, and dumps the current database
first so a failed import is rolled back.
A backup dumps, zips and encrypts for minutes without issuing a query, so on
a server with a short wait_timeout the connection is closed underneath it and
the UPDATE marking the backup complete is what fails - long after the archive
was written correctly. The connection is now held open for the job and
re-established before any write that follows long file work, including the
database phase of a restore. Retention recovers rows a dropped connection
left behind: still Running after six hours becomes Complete if the archive is
on disk, Failed if it is not.
cron.php's own failure path is hardened to match. It recorded job failures
through the same connection the failing job had just killed, so an uncaught
exception ended the dispatch and no trace of the original error survived.
Failures now also echo to stdout, so cron mails something useful when the
database is unreachable.
Security: the setup wizard's restore step is now closed on any install that
has users, whatever config.php says. $config_enable_setup defaulted to
enabled when the flag was absent, and the flag is only written at the end of
a successful install, so an install abandoned partway left an unauthenticated
endpoint that would drop every table, import an attacker-supplied archive,
and overwrite uploads/ including the .htaccess that stops PHP running there.
Affects 26.07 and earlier. Restoring over a live install is now CLI only.
Statuses can be flagged to pause the resolution clock; sla_history records the intervals a ticket's clock actually ran and the deadline is re-based on remaining budget when it resumes. Adds SLA Summary and SLA by Client reports, an SLA state filter on the ticket list, SLA colouring on kanban cards, and an Urgent column on the Tickets by Client report. DB update 2.5.1. Also fixes resolution SLA verdicts being skipped when resolving via kanban or the client portal.
Response/resolution targets stamped at creation from per-client/priority assignments, business-hours due date math, warn/breach alert stages via cron/ticket_sla.php, ticket list coloring, per-ticket SLA override, admin page for plans/assignments/business hours. DB update 2.5.0. No behavior change unless SLAs are assigned. Bulk reply now only counts Public replies as first response.
API keys now run as a user and inherit that user's module, operation, and client permissions. Drops per-key client scoping and removes existing keys (must be recreated). Adds an edit modal to change a key's user.
