1. Upgrade notes


2. New Features

2.1. Acquisitions

2.1.1. Auto-Cancel Line items When All Copies Are Canceled

When a copy (line item detail) is canceled through the Acquisitions interface, the parent line item is also canceled if all copies for that line item are also canceled. The cancel reason given will come from:

  1. The cancel reason for the just-canceled copy if it’s a Keep Debits true cancel reason.

  2. The cancel reason from any other copy on the lineitem that has a Keep Debits true cancel reason.

  3. The cancel reason for the just-canceled copy if no copies have a Keep Debits true cancel reason.

2.1.2. Invoice Closed Date and Closed By Fields

Acquisitions invoices have 2 new fields:

  • Close Date — This is set to the time when the ACQ user clicks the "Close" button in the invoice interface.

    • This field replaces the existing complete field. An invoice is considered complete if a close date value is set.

  • Closed By — This is set to the logged in staff user who performs the "Close" action.

As with the now-defunct complete field, but new fields are cleared in the event an invoice is reopened.

These new fields are visible in the invoice interface under the Show Details action for closed invoices.

Upgrading Invoice Reports

Existing report templates that reference the invoice complete field should be modified to check whether the new close_date field is NOT NULL instead.

Other Upgrade Considerations

At deploy time, all invoices with a complete value of TRUE will have their close_date field set to NOW. A value is required, since this field is now the source of whether an invoice is open or closed.

However, no values will be applied to the closed_by field for already closed invoices.

2.1.3. Patron Acquisitions Requests

The existing interface for staff-mediated patron acquisition requests has been replaced in the web staff client with a re-implementation written in AngularJS, with some minor bug fixes (including access from the Patron interface) and other improvements.

2.2. Architecture

2.2.1. Angular6 Base Application

With Evergreen 3.2, we introduce the initial infrastructure for migrating to a new version of Angular. The structure of the new code is quite different from the AngularJS code and it runs as a separate application which communicates with the AngularJS app via shared storage and in-page URLs that link back and forth between the two.

For this release, users will only be directed to the new Angular site when navigating to Administration ⇒ Acquisitions Administration. Once on this page, some of the admin interfaces will presented as Angular6 interfaces, while others will direct users back to the AngularJS application. The Angular6 interfaces are the simpler, grid-based interfaces.

Acquisitions Admin Angular6 Interfaces
  • Cancel Reasons

  • Claim Event Types

  • Claim Policies

  • Claim Policy Actions

  • Claim Types

  • Currency Types

  • EDI Accounts

  • EDI Messages

  • Exchange Rates

  • Fund Tags

  • Invoice Item Types

  • Invoice Payment Method

  • Line Item Alerts

  • Line Item MARC Attribute Definitions

System Admin Upgrade Notes

Like the AngularJS application, Evergreen releases will come with all web browser staff client code pre-compiled. Admins only need to add an Apache configuration change.

Add the following stanza to /etc/apache2/eg_vhost.conf.

RewriteCond %{REQUEST_URI}  ^/eg2/
RewriteCond %{REQUEST_URI}  !^/eg2/([a-z]{2}-[A-Z]{2})/
RewriteRule ^/eg2/(.*) https://%{HTTP_HOST}/eg2/en-US/$1 [R=307,L]

<Directory "/openils/var/web/eg2/en-US">
    FallbackResource /eg2/en-US/index.html

For multi-locale sites, see the bottom section of Open-ILS/examples/apache[_24]/eg_vhost.conf.in for a sample fr-CA configuration. The section starts with "/eg2/ client setup and locale configuration"

Developer Upgrade Notes

Developers building Angular code on existing installations need to update their version of NodeJS by re-running the -developer prereqs installer.

sudo make -f Open-ILS/src/extras/Makefile.install <osname>-developer

2.3. Cataloging

2.3.1. Add UPC to z39.50 search for OCLC and LOC

Add UPC as a search attribute for both OCLC and LOC targets in z39.50 for cataloging.

2.3.2. Asynchronous Vandelay Imports

Vandelay imports are now monitored from the browser client asynchronously, meaning the client requests updates from the server instead of waiting for the server to respond to the original import request. This changes allows for incremental progress updates in the browser client.

New Database Table

This adds a new database table vandelay.session_tracker for tracking in-progress vandelay upload activity. A new tracker row is added for each of "upload", "enqueue", and "import" actions, linked for a given session by the value stored in the "session_key" field.

The table tracks other potentially useful data, like the staff member and workstation where the action was performed.

Upgrade notes

Users of NGINX as a reverse proxy may need to set a suitable client_max_body_size value in the NGINX configuration so that large MARC record uploads are not truncated. Note that this would have always been necessary, but since this feature allows larger files to be more reliably queued and imported, the need to set client_max_body_size became more apparent.

2.3.3. Support for Last Inventory Date

Evergreen now provides an option to add an inventory date to items to facilitate the process of performing inventory in libraries. Staff can add an inventory date to an item in one of the following ways: * From the check in screen, there is now an Update Inventory check in modifier. When selected, scanned barcodes will have the current date/time added as the inventory date while the item is checked in. * From the Item Status screen, an action is available to add the current date/time as the inventory date to selected items.

This new feature will also store the workstation that was used when the inventory date was updated.

2.3.4. Parallel Ingest with pingest.pl

A program named pingest.pl is now installed to allow faster bibliographic record ingest. It performs ingest in parallel so that multiple batches can be done simultaneously. It operates by splitting the records to be ingested up into batches and running all of the ingest methods on each batch. You may pass in options to control how many batches are run at the same time, how many records there are per batch, and which ingest operations to skip.

The browse ingest is presently done in a single process over all of the input records as it cannot run in parallel with itself. It does, however, run in parallel with the other ingests.
Command Line Options

pingest.pl accepts the following command line options:


The server where PostgreSQL runs (either host name or IP address). The default is read from the PGHOST environment variable or "localhost."


The port that PostgreSQL listens to on host. The default is read from the PGPORT environment variable or 5432.


The database to connect to on the host. The default is read from the PGDATABASE environment variable or "evergreen."


The username for database connections. The default is read from the PGUSER environment variable or "evergreen."


The password for database connections. The default is read from the PGPASSWORD environment variable or "evergreen."


Number of records to process per batch. The default is 10,000.


Max number of worker processes (i.e. the number of batches to process simultaneously). The default is 8.


Skip the selected reingest component.


Start processing at this record ID.


Stop processing when this record ID is reached.


Read record IDs to reingest from standard input. This option conflicts with --start-id and/or --end-id.


Stop processing after this many total seconds have passed. The default is to run until all records have been processed.


Show the help text.

2.3.5. View Authority Record by Database ID

A new interface allows catalogers to retrieve a specific authority record using its database ID. Catalogers can find those IDs in subfield $0 of matching fields in bibliographic records.

To use the new authority record viewer:

  1. Click Cataloging → Retrieve Authority Record by ID.

  2. Type in the ID number of the authority record you are interested in. Don’t include any prefixes, just the ID number.

  3. Click Submit.

  4. View or edit the authority record as needed.

2.4. Circulation

2.4.1. Autorenewal of Loans

Circulation policies in Evergreen can now be configured to automatically renew certain items checked out on patron accounts. Circulations will be renewed automatically up to a custom limit (the max_auto_renewal field) and patrons will not need to log in to their OPAC accounts or ask library staff to manually renew materials.

Two new action triggers have been added to Evergreen that permit the Auto-Renew feature. They can be found, configured, and enabled in Administration>Local Administration>Notifications/Action Triggers. They are named Autorenew and AutorenewNotify.

The Autorenew A/T definition uses the checkout.due hook to automatically validate and renew (in the reactor) circulations on the day they are due, grouped by user. The output events of this definition is is the input used by the related AutorenewNotify A/T that simply uses a new hook called autorenewal to notify patrons via email of their currently due or auto-renewed items.

In the webstaff’s Patron Items Out page, the new column AutoRenewalsRemaining indicates how many autorenewals are available for a particular circulation.

2.4.2. Emergency Closing Handler

Staff are provided with interfaces and mechanisms to create library closings that, in addition to affecting future circulation and booking due dates, and hold shelf expirations, will automatically move existing circulation and booking due dates and hold shelf expiration times. This new functionality is conceptually described as Emergency Closings and business logic implementing it as the Emergency Closing Handler. It contains additions and adjustments to the user interface, business logic, and database layers. Access to this functionality is available through the Closed Dates Editor interface in the staff client which has been ported to AngularJS.


This development has created new business logic code to inspect, in real time, existing circulation, booking, and hold records, and modify such date and time stamps so that the circulation, booking, or hold will end in the same state it would have if the closing had existed at the time the circulation or booking occurred, or the hold was placed and captured. Of specific note, hourly loans will have their due date adjusted to be the end of the day following the closing.

When the Emergency Closing is saved, any fines accrued during the closing may be voided, as settings dictate, with the exception of circulations that have been marked as LOST or LONG OVERDUE. That is, even for LOST and LONG OVERDUE circulations with due dates that fall within the Emergency Closing, no fine adjustment will be applied. Emergency Closing processing is permanent, and cannot be rolled back.

This functionality is explicitly initiated by staff action. If staff do not request an Emergency Closing, existing circulations, bookings, and holds will not be processed and adjusted. However, if staff request any Closing that starts nearer in time than the length of the longest circulation duration configured for use in the Evergreen instance they will be prompted with the option to create the closing as an Emergency Closing.

Action/Trigger hooks have been created for circulations and bookings that are adjusted by the Emergency Closing Handler. These will facilitate the creation of notifications to patrons that the due date has changed and to alert them to potential changes in accrued fines.

Booking start dates are explicitly ignored in this implementation. Because an Emergency Closing is, by its nature, an unexpected event, it will be up to staff to address any bookings which intersect with a new Emergency Closings. Reports can be used to identify booking start dates that overlap with a closing and that may require staff intervention.

Staff requesting and Emergency Closing must have the new EMERGENCY_CLOSING permission. Some text describing the feature.

2.4.3. Patron Preferred Name and Name Search Keywords

Preferred Name

Adds a new set of patron preferred name fields for prefix, first, middle, last, and suffix allowing patrons to provide preferred name information. Preferred names are optional and each acts as an overlay to the analogous primary name field, making it possible to provide preferred name values for individual fields.

For example, a patron named William Erickson may have a preferred first name (pref_first_given_name) of Bill, in which case the preferred name would be Bill Erickson. Note a preferred last name is not required in this case as the code uses primary name values as defaults when not replaced with a preferred version.

  • Patrons will see primary names displayed in the catalog when set.

  • Staff will see both primary name and preferred name in the patron summary side bar.

  • Patron searches for any given name field will search both the primary and preferred name data.

  • Preferred name fields are available in Action/Trigger templates and are present in various patron-focused print templates.

Name Keywords

Adds a new field to store miscellaneous patron name search terms. These values are only for searching and do not appear in any interfaces, apart from the patron summary side bar and the patron edit UI.

Included is a new search field in the patron search UI which searches keyword values and all other name fields. It’s essentially a global patron name keyword search.

2.5. Client

2.5.1. Disabling of legacy XUL staff client

The legacy XUL staff client is no longer supported in Evergreen 3.2.x and the server-side installation no longer supports a direct connection by a version XUL client by default. All users of Evergreen 3.2.x are strongly urged to complete their switch to the web staff client as part of upgrading to 3.2.x.

Evergreen administrators who for some reason continue to wish to deploy the XUL staff client can do so at their risk by supplying STAFF_CLIENT_STAMP_ID during the make install step and using make_release to create installers for the staff client. However, no community support will be provided for the XUL client.

2.5.2. Permission Group Display Entries

In some cases, it is useful to have the ability to reorder permission, or to make only specific groups available in the permission group selector for specific Org Units. An interface has been made available to allow this.

Group Tree Display Entry Interface

Permission Group Display Entries can be reordered, added, or removed via Administration → Local Admin → Permission Tree Display Entries. Select the Org Unit you wish to edit the entries in.

Entries may be added using the Add functionality, creating entries based on permission groups that have not been added to the tree for the Org Unit you wish to add them to.

Group Tree Display Entry Admin UI
Moving an Entry

Moving an entry will shift its position up or down in the patron profile selector for a given Org Unit.

  • Select an entry

  • Press either the Move Up or Move Down button. The entry will be moved up or down, accordingly.

  • Click Save to save your edits.

You may only move up or down entries that have sibling entries.
Removing an Entry

If you want a particular Org Unit to not have access to specific entries, you may remove an entry. Removing an entry will remove it from view. The entry will be removed from the database.

  • Select an entry and press the Remove button.

Adding an Entry

You may add entries from permission groups that are not currently reflected in the permission group tree. This is useful for moving entries to different parents, or making them root entries.

Add Entry modal
  • If desired, select an entry to be used as the parent entry.

  • Press the Add button.

  • Select a permission group from the dropdown.

  • If you’ve selected a parent entry, you may check the Add Root Entry box to override that parent and add the entry on the root level.

  • If you did not select a parent entry, the entry will be added on the root level of the tree.

2.5.3. Browser Client Settings & Preferences Stored on the Server

Browser client settings and preferences that should persist over time are now stored as settings on the server. This allows settings to follow users and workstations and reduces problems associated with losing settings as a result of clearing browser data.

The browser client honors setting values stored as user settings, workstation settings, and org unit settings, depending on which setting types are locally configured.

Setting Types
  • No setting can be both a user and workstation setting. They are mutually exclusive.

  • Any setting can be an org unit setting in addition to being a user or workstation setting.

Read-Only Settings

Read-only settings are useful for defining values that staff can use but not modify. For example, admins may wish to prevent users from locally modifying the grid configuration for a given interface so it remains consistent for all users.

A setting is read-only when an org unit setting type exists (regardless of whether a value is applied) and no user or workstation setting type exists.

Server-Stored Workstation Settings Workstation Admin View

There’s a new "Server Workstation Prefs" tab to the stored preferences workstation admin interface. From here, users can view which preferences are stored as server-stored workstation preferences and delete select values.

Upgrade Notes

A new permission APPLY_WORKSTATION_SETTING has been added to control who may apply values to workstation settings. Use something like the following to apply the permission to all staff accounts (mileage may vary):

INSERT INTO permission.grp_perm_map (grp, perm, depth)
    (SELECT id FROM permission.grp_tree WHERE name = 'Staff'), -- name may vary
    (SELECT id FROM permission.perm_list WHERE code = 'APPLY_WORKSTATION_SETTING'),
    0 -- or 1, 2, etc.

Workstation setting types matching values previously stored in the browser (via localStorage or Hatch) are created as part of this feature. During upgrade, admins should consider whether any of these new setting types should be transferred to user and/or org unit settings instead. Setting type changes can be made at any time, but when a setting type is deleted all of its data is deleted, so a change in type means re-applying the settings in the browser client.

Values stored in the browser will automatically migrate to server settings as each setting is accessed in the browser client. Once migrated, the in-browser copies are deleted.

If a setting type does not exist where the browser expects one, the value is stored in-browser instead and a warning is issued in the console.

2.6. OPAC

2.6.1. Batch Actions In the Public Catalog

The public catalog now displays checkboxes on the bibliographic and metarecord constituents results pages. Selecting one or more titles by using the checkboxes will dynamically add those title to the temporary list, which is now renamed the cart.

Above the results lists there is now a bar with a select-all checkbox, a link to the cart management page that also indicates the number of of titles in the cart, and a link to remove from the cart titles that are selected on the currently displayed results page.

The search bar now includes an icon of a cart and displays the number of titles currently in the cart. Next to that icon is a menu of cart actions.

The cart actions available are Place Hold, Print Title Details, Email Title Details, Add Cart to Saved List, and Clear Cart. In the web staff client, the cart actions also include Add Cart to Bucket. When an action is selected from this menu, the user is given an opportunity to confirm the action and to optionally empty the cart when the action is complete. The action is applied to all titles in the cart.

Clicking on the cart icon brings the user to a page listing the titles in the cart. From there, the user can select specific records to request, print, email, add to a list, or remove from the cart.

The list of actions on the record details page now provides separate links for adding the title to a cart or to a permanent list.

The permanent list management page in the public catalog now also includes batch print and email actions.

Additional information
  • The checkboxes do not display on the metarecord results page, as metarecords currently cannot be put into carts or lists.

  • The checkboxes are displayed only if JavaScript is enabled. However, users can still add items to the cart and perform batch actions on the cart and on lists.

  • A template config.tt2 setting, ctx.max_cart_size, can be used to set a soft limit on the number of titles that can be added to the cart. If this limit is reached, checkboxes to add more records to the cart are disabled unless existing titles in the cart are removed first. The default value for this setting is 500.

Developer notes

This patch adds the the public catalog two routes that return JSON rather than HTML:

  • GET /eg/opac/api/mylist/add?record=45

  • GET /eg/opac/api/mylist/delete?record=45

The JSON response is a hash containing a mylist key pointing to the list of bib IDs of contents of the cart.

The record parameter can be repeated to allow adding or removing records as an atomic operation. Note that this change also now available to /eg/opac/mylist/{add,delete}

More generally, this adds a way for EGWeb context loaders to specify that a response should be emitted as JSON rather than rendering an HTML page using Template::Toolkit.

Specifically, if the context as munged by the context loader contains a json_response key, the contents of that key will to provide a JSON response. The json_response_cookie key, if present, can be used to set a cookie as part of the response.

Template Toolkit processing is bypassed entirely when emitting a JSON response, so the context loader would be entirely responsible for localization of strings in the response meant for direct human consumption.

2.6.2. New class for searchbar when on the homepage

This adds the .searchbar-home class to the div that contains the searchbar when on the homepage. This allows sites to customize the searchbar differently on the homepage than in other places the search bar appears (for example, offering a large, Google-style search bar on the homepage only).

2.6.3. Username Login Hint

To make customization easier, the username hint on the OPAC login page ("Please include leading zeros…") has been moved to a separate TT2 template. If you have customized the hint text, you will need to add your modifications to username_hint.tt2.

3. Acknowledgments

The Evergreen project would like to acknowledge the following organizations that commissioned developments in this release of Evergreen:

  • British Columbia Library Cooperative


  • Georgia Public Library Service

  • Indiana State Library

  • Lake Agassiz Regrional Library

  • MassLNC

  • North Texas Library Consortium

  • Northwest Regional Library

  • Consortium of Ohio Libraries

  • Pennsylvania Integrated Library System

  • South Carolina State Library

We would also like to thank the following individuals who contributed code, translations, documentations patches and tests to this release of Evergreen:

  • Felicia Beaudry

  • Jason Boyer

  • British Columbia Libraries Cooperative

  • Andrea Buntz Neiman

  • Eva Cerninakova

  • Galen Charlton

  • Garry Collum

  • Jeff Davis

  • Bill Erickson

  • Jason Etheridge

  • Lynn Floyd

  • Jeff Godin

  • Blake Graham-Henderson

  • Francisco J Guel-Mendoza

  • Kyle Huckins

  • Mary Jinglewski

  • Angela Kilsdonk

  • Kathy Lussier

  • Katie G. Martin

  • Jennifer Pringle

  • Morkor Quarshie

  • Mike Rylander

  • Jane Sandberg

  • Chris Sharp

  • Ben Shum

  • Remington Steed

  • Jason Stephenson

  • Cesar Velez

  • Dan Wells

  • Stephan Woidowski

We also thank the following organizations whose employees contributed patches:

  • British Columbia Libraries Cooperative

  • Calvin College

  • Catalyte

  • Equinox Open Library Initiative

  • Kenton County Public Library

  • King County Library System

  • Linn-Benton Community College

  • MassLNC

  • Sigio

We regret any omissions. If a contributor has been inadvertently missed, please open a bug at http://bugs.launchpad.net/evergreen/ with a correction.