Evergreen 3.14 Release Notes

Support is added for handling X12-format ASN messages, which are generally used to deliver shipping status information about EDI-processed materials orders. Development targeted the general X12 message format standard for data parsing, and configuration suitable for processing the expected contents of ASN messages in particular. While the initially targeted production vendor test data was from Midwest Tape, the general X12 parser was built using documentation and non-production examples from several vendors, and is expected to handle X12-format ASN message from most library materials vendors.

This functionality is implemented in a way that does not require intervention or configuration by the Evergreen administrator beyond the normal EDI configuration. If an X12-format message is delivered by the vendor, Evergreen will notice and attempt to handle it transparently.

#2003106

The open-ils.search drone had the ability to cache the Z39.50 target definition data per process.

This fix removes per-child-process caching of Z39.50 target definitions to address several issues caused by the caching:

#2044854

Two LogFormat entries have been added to the sample Apache configuration in eg.conf.in that are useful when a proxy runs in front of Evergreen’s Apache. These configurations mimic the default "common" and "combined" log formats, but log the remote client’s actual IP address when mod_remoteip is enabled. The entries are repeated here in case you prefer to manually update your Apache configuration.

To use these formats, simply replace the "common" or "combined" at the end of the CustomLog entries in the eg.conf file with either the "proxy-common" or "proxy-combined" entry, depending upon which you want to use.

From:

to:

for example.

#1908540

A new script for bulk loading and updating patrons from the server now exists in Open-ILS/src/support-scripts called patron_loader.pl. It is installed to <prefix>/bin (/openils/bin/patron_loader.pl for standard installs). It can be run manually or from cron.

Sample invocation:

Required parameters:

Optional parameters:

Database settings loaded by default from opensrf.xml

Required Columns:

Although data for the above columns are optional in some situations the columns still need to exist in the file.

Optional Columns:

Mapping:

Not all data sources can customize the data exported to the CSV so some mapping is allowed.

The config.patron_loader_header_map table allows for mapping incoming header names to ones that are natively expected. For example, imagine that a school wants to use the uid as password and the column header will always read uid then you can enter it like this:

Two value types can currently be mapped as well, home_library and profile in patron_loader_value_map. These map values in their respective columns instead of the headers. For example, imagine a school who exports student profiles of Middle School and High School but both need to load as the Evergreen profile of Student. It would be represented with two entries:

You can also map home libraries like this:

As a convention the Evergreen database column names are mostly used for the actor.usr columns but it was found in testing that home_ou was very confusing so the label of library is used instead and internally adjusted to use home_ou.

The column ident_type is treated specially. It is required by actor.usr and does not have a default but usually doesn’t correspond to a exported value from others systems so it defaults to 3 or Other but you can define it through an optional parameter.

Overview:

The script is very conservative checking for an existing cardnumber and usrname. If either is found on an account that differs from the one using the match point then it will skip adding or updating that user. The match point specified is considered authoritative and it will update the matching account unless debug is on.

Currently only two set of address columns are supported add1_foo and add2_foo. The script assumes the addresses being added are authoritative mailing addresses, removes any existing mailing addresses, adds these and sets the user’s mailing_address field to the one from the addr1_street1 field or addr2_street1 if there is no addr1_street1. If only a partial address is given the entire address will be written so long as there is a street1. Empty strings will be used for the other values. If there is no address given then addresses will not be touched. Part of the aggressiveness of removing non-specified addresses is to ensure identifying information for patrons is removed when updating, especially for the use case of schools bulk updating juveniles.

Database and Logging:

The database holds a actor.patron_loader_log table that logs sessions and failed rows.

#1786524

The format.date and format.time org unit settings are no longer used by the staff client and have been marked Deprecated. If you are not using these OU settings in a local customization they can be removed from the database at your convenience.

Evergreen 3.14 no longer supports new installations on PostgreSQL versions less than 13. Evergreen 3.14 is still compatible with PostgreSQL versions of 10 or higher. Existing installations may upgrade to Evergreen 3.14 without requiring a PostreSQL upgrade.

Compatibility with obsolete PostgreSQL database versions is not guaranteed in the next major Evergreen release. Users are encouraged to upgrade to a more recent PostgreSQL version if they are running PostgreSQL versions between 10 and 12.

The PostgreSQL community recommends a dump and restore from the older PostgreSQL database to the newer one when upgrading over major versions. For more information, see the PostgreSQL upgrade documentation.

This work moves the Record Buckets feature into a new and reimagined Angular interface, and adds several new feature improvements.

#2063146

Feature improvements include:

In addition to feature improvements, this work addresses functional bugs related to record buckets, including:

This work also introduces some changes to the grid component:

This also introduces a change to the title area in Angular screens. Instead of using the blue alert style, page titles are now left-aligned and have a template area afterward. This is intended to be used for single buttons that represent a primary action on the screen, e.g. Add New [Thing]. This change makes room for a second template area that can be placed to the right of the title, used here for a search box.

This work adds the following new permissions. Note that the CALL_NUMBER, COPY, and USER permissions are placeholders for potential future work:

Public copy notes and tags are now exported by marc_export in the 852 subfield z when the --items option is used.

Copy tags come before copy notes, and tags with a URL come before those without. The URL of a copy tag (if any) is added to 852 subfield u. The copy tags are ordered such that the first subfield z should correspond to the first subfield u and so on if there is more than one URL.

#2045440

The Require Monographic Part when Present Library Setting, circ.holds.ui_require_monographic_part_when_present, has been expanded to also require that all copies on the parent bibliographic record have a part when the setting is active.

This setting requires a patron to select a monograph part when placing a hold if any parts exist on the bibliographic record. In a logical extension of this behavior, the setting now also requires catalogers to use parts on all copies when any copy at the library have parts.

The change in behavior only affects the library/organization unit where the setting is set or its sub-units.

#2018014

SuperCat and unAPI can now export public copy tags when copies are exported.

To request the export of copy tags via unAPI, add acpt to the URL "includes" section.

#2047442

New database tables

Hold reset reasons allow staff to see when and why a hold request has been reset. Reset reasons are generated any time a hold has been reset, whether that’s a manual reset from a staff member or automatically because a hold has reached the hold retarget interval. This can be very useful for debugging the hold targeter or identifying bad actors in the system.

Types of Reset Reasons

There are ten different types of reset reasons that can be identified.

Viewing Reset Reasons

Staff can view reset reasons for a hold via a patron’s holds tab.

New srfsh script

This feature includes a script that you can setup as a cron job to maintain the action.hold_request_reset_entry table:

purge_hold_reset_reason_entries.srfsh

This script defaults to "1 year" but can be overridden on a library by library basis via YAOUS.

New library settings

#2012669

The patron self-checkout interface is now available as an Angular port of the interface.

https://myhost.mydomain/eg2/staff/selfcheck

New Print Templates

New self-check print templates are available under Administration ⇒ Server Administration ⇒ Print Templates

#1840773

The staff client now supports both light and dark modes. By default, the staff client will use the color mode setting from your operating system. That is to say, if you have turned on Dark Mode for your operating system, or if your operating system defaults to Dark Mode, the Evergreen staff client will now also display in Dark Mode.

If you’d like to use a different color mode in the staff client than your operating system setting, you can do so using the color mode selector in the navigation bar at the top of the staff client. If you want to resume using your operating system’s color mode, you can set the color mode selector to "Auto".

#1740529

Adding support to the Staff Client for Multi-Factor Authentication. Supported factors:

#2071636

There is a new setting to hide carousels from the public catalog. This can be useful in cases where carousels are created for use in other sites (e.g. on library websites), rather than for use in the public catalog.

You can turn off carousels in config.tt2 by setting ctx.show_carousels to 'false'.

#2009903

OPAC overhaul of accessibility in several categories:

#2048666

Improves styling and accessiblity and adds the following library settings:

#2065448

The new Report Security functionality is primarily configured through the addition of XML attributes to elements in the Fieldmapper XML file, fm_IDL.xml. These new attributes fall into three categories:

All restriction definitions can make use of the full set of fields on the restricted source (LEFT side for core source and link-element projected sources, RIGHT side for class-level projection-restricted sources), the staff user that scheduled the report run, and any aribtrary liternal value, though typically the last will be a set of one or more permissions to be tested.

Many of the existing, permission-related database functions can act as redaction and join/projection restriction functions. Additional functions are supplied as part of this development in order to faciliated restrictions based on Patron Opt-In values in effect at the time a report is run.

See the Reports Security documentation for more details.

In addition to report content security restrictions, report output visibility is now, optionally, restricted based on whether the accessing user either owns the report output, or they have at least the VIEW_REPORT_OUTPUT permission at a location to which the folder in which the output lives has been shared.

Administrators can add additional required permissions via the OILSProxyPermissions Apache configuration value in the report output <Location> section of the eg_vhost configuration file.

#2043142

If your PC management system or selfcheck looks for a screen message value of OK to evaulate user standing this can be enabled by adding the want_patron_ok value to a sip login and setting the value to true. Using a login setting like this allows you to only send this screeen message to those clients that expect it.

#1613335

The Angular Staff Catalog now has a Serials Quick Receive feature, similar to the one from the traditional catalog.

#1906462