2020-09-24: release new 1.9.5 build 2192

Today we release which fixes a bug in causing dataset ranges in multi-tenancy applications not being honored in the Find data for the field window. Even though the issue seemed specific to multi-tenancy implementations, we recommend updating to build 2192 if you installed build 2147 earlier.

2020-09-21: release Axiell Collections 1.9.5

Today we release the fifth update of Axiell Collections 1.9 (build, offering the bug fixes described below. For some new functionality to become available you may need to clear the cache of your browser after Collections has been updated on the server. The quickest way do do this is often to use the Ctrl+F5 key combination (in Windows) or Command+R (on Apple machines) on the Collections login page.

Bug report no.

Short problem description


Name-based geography fields were no longer showing in the Map view.


When creating child records using the bulk create button on the hierarchy browser, inherited data from the parent was not overwritten with explicitly provided new data.


An standard search with hierarchical operator resulted in a Column name or number of supplied values does not match table definition error.


A Language is required error occurred when trying to save a Loans record in 4.5 model applications.


Updating a date on the part_of_reference.date.start (ef) field on a Series record, would also update the date value in the other Series records that linked to the same parent Activity record, if the parent/child hierarchy relationships link to the same metadata record.


Advanced searches with the random operator would generate an error.


A search-and-replace in a multilingual field added a non-existent language attribute without value to the field.


The default sort order on the Workflow Activity Details no longer worked.


Performing an advanced search during a derive action returned a bad request error.


The refresh buttons in the Record details view table grids had disappeared.


The Geographical map widget showed plotted polygons in non-transparent black for all but the first record.


Adding a new source record to an article added the new record automatically to the Resources dataset.


Pressing Enter after filling in your login details in the login dialog of an application in which password resetting was enabled, opened the Forgot your password screen.


Manual sorting did not work after you re-ran a saved search.


The search results, when trying to link one or more marked records to a record in another database, were shown in a list box only one row high, with a very small scroll option.


Pressing Ctrl+Enter in a repeated field (to create a new occurrence) also deleted the previously entered term in the current occurrence of that field.


The Change password option in the Account drop-down in the main menu on the left had gone missing.


There were missing or wrong French translations regarding the related records functionality and the warning about mandatory fields not having been filled in.


Performing an advanced search on a non-indexed field always returned No records found.


The Occurrence menu in the bulk create window was corrupted.


The search results of a derive were shown in a list box only one row high, with a very small scroll option.


Standard search form hierarchical operators were missing for a field linked to the geothesaurus.


After clicking the Keep marked records icon in the Result set, it would display No results.


For multilingual inheritable fields, only the values in the first language were inherited, and without a language attribute (which could cause data corruption).


Deleting a value in a data language other than the invariant language would reset the invariant language to null.


Cutting a record from Desiderata didn't work.


A manually filled-in write-back field (that was also a merged-in field) was emptied before storage, even if a before-storage adapl filled the associated linked field with a new value in order to force a new linked record with that write-back field.


Trying to derive records from an external source resulted in a Bad request error.


A number of dialogs had an OK and Cancel button with the same functionality. One of the buttons was redundant.


When editing records in the Result set view, clicking on a field that has an associated zoom screen would always open the zoom screen instead of allowing you to edit the record in the Result set.


Enumerative fields did not show a particular value in the drop-down list or in the field, if that value had no English translation and no translation in the current UI language (if that language was different from English).
The fix means that in this case the so-called neutral value of the drop-down item will be shown.


No message You are already using this record, do you want to continue editing? when you tried to edit a record that you had locked yourself.


Hierarchical search operators didn't work on the new hierarchical link structure.


After derival of a record, an empty record was shown.


Upon deriving of a record, all non-exchangeable fields were copied along, regardless of whether the derival means a copy or a move.
The fix means that if derival implies the copying of a record, non-exchangeable fields won't be copied, while if derival implies the moving of a record, non-exchangeable fields will be retained in the moved record.


Inconsistent behaviour in (still) highlighted, filled mandatory fields.


A mandatory field in a task screen was not marked in light red after clearing its default value.


Mandatory fields in a task screen could be left empty.


Merged-in invariant language values were not displayed (or printed).


After adding a new field to the displayed fields (columns) displayed in the Result set, the new column did appear but the field values were not displayed.


Field info was not available for date fields and numeric fields.


It was impossible to add date fields or numeric fields to the clipboard.

2020-09-21: Create term and Create and edit term buttons renamed

The underlying functionality has not changed but the Term creation section with the Create term and Create and edit term buttons in the right bottom corner of the Find data for the field window have been renamed to the Link creation section with the Accept link and Create and edit link buttons, to more clearly reflect their function, as it's not always terms you are accepting or creating:

with the Accept link button you just accept a term, name, location, exhibition or some other identifier of a new linked record for now while it will only really be saved as a new linked record once you save the main record you are currently working in
with the Create and edit link button you create, edit and save a new linked record now, even before saving the main record you are working in.


2020-08-19: release Axiell Collections 1.9.4

Today we release the fourth update of Axiell Collections 1.9 (build, offering the bug fixes described below. For some new functionality to become available you may need to clear the cache of your browser after Collections has been updated on the server. The quickest way do do this is often to use the Ctrl+F5 key combination (in Windows) or Command+R (on Apple machines) on the Collections login page.

A note about running Collections 1.9.3 and higher in Microsoft Edge

New Collections functionality introduced with version 1.9.3, namely the ability to run multiple instances of the same application in a single browser (using the new ?tabId parameter in the URL to Collections) causes older, non-Chromium versions of the Microsoft Edge browser to keep refreshing the page. If you are using this browser, the issue is easily corrected though, by upgrading to the new, Chromium version of the Edge browser (released Februari 2020).  Although Microsoft is gradually rolling out the update of its browser to all users automatically, like it does with Windows, you may still have to upgrade Edge manually now to get past this issue. See the following web page to learn how to find out which version of Edge you are using and how to update to the new version if necessary. Note that the Chromium version of Edge can also be recognized by its new icon in the Windows task bar:
The mentioned issue does not appear in other browsers.

Bug fixes in 1.9.4

Bug report no.

Short problem description


The Search settings dialog showed an incorrect French translation when the UI language was French.


For the add/remove current record to saved search functions, French translations were missing.


For any occurrence groups that are configured to be a table grid, you could no longer add new occurrences. The options to add rows above and below were available, but they wouldn't generate a new row regardless of which option you chose. Also, the refresh button that is normally at the bottom of the grid, was missing.


An object reference exception appeared for unknown fields in an internal link setup when saving a record.


A <<DUMMY+FN>> reference in a Word template printed the image file name instead of the image itself.


The Write set dialog showed an incorrect French translation when the UI language was French.


When adding a new linked Persons and institutions record from the Books dataset, the field input.source in people.inf was filled with people>people instead of document>book.


Several column sorting options in the Result set did not have a French translation.


The Create button in the search/create dialog opened via Links did not have a French translation.


An adapl running with a Word template did not have the processed record to its disposal in memory.


When exporting data with multiple occurrences to CSV in Hebrew, the data in the resulting CSV was mangled: a lot of data was put into the next row.


When creating a new record, removing the configured default value for a field manually produced an error on saving the record.
The fix means that an error no longer occurs after you've deleted the default value in a new record, but on saving the record this value will reappear. You can only remove a default value from an existing record by editing the record again and deleting the value again.


Pasting a copied value in a multilingual field created an additional empty occurrence value with an empty language attribute.

2020-08-07: release Axiell Collections 1.9.3

Today we release the third update of Axiell Collections 1.9 (build, offering the bug fixes and new functionality described below.

Bug report no.

Short problem description


When you searched an Archives catalogue for any record and opened the Hierarchy browser view, then an Invalid column name 'priref' error was generated because of an invalid SQL query.


Running the advanced search like description.type = * and not description.type = includes or task -> job.number = 89006 (a combination of indirection and and not) would throw a SQL error.


A search that retrieved certain specific records in the Collections cloud, could cause Collections to stop responding. And if the page was refreshed, the session was logged out.


Searching with the related operator did not find any records.


The fixed query expand(TI = %worddata%) and Df = ITEM (searching on an inherited title in combination with specifying a particular record level) generated an invalid object error.


An access point for input date in the Standard search did not work.


In settings.xml, <Setting Key="DomainUsersOnly" Value="false" /> did not work anymore if application authentication was set to None.


A %field%-based variable domain was preventing the Find data for the field window from displaying terms when opened from the Simple search.
By default, when there is a variable domain defined for a linked field in order to filter it based on a value in another field in a record (e.g. %dataset%_FORM), and the user wants to pick a term for that field while in Simple search, the Find data for the field list comes back empty while the auto-complete list (the drop-down list for the field) does come back with terms, but with all terms for the field, not terms filtered by the domain.
The solution requires an adjustment of the relevant access point(s) - the fields you can search in the Simple search have been set up as so-called access points (aka methods). You can now associate an access point with a specific domain, by setting the fixed domain in the Domain name property on the Method properties tab of the desired access point. So then the access point will only search terms within that fixed domain. If you need to be able to search the relevant field within different domains, then you'll have to create a separate access point for each of them.


When creating a new record from the Record details view while a record is selected in the results view, the information in the table grids from the previously displayed record was copied to the new record if the tabs containing the table grids remained open.


If during a search-and-replace in a linked field a non-unique name was to be inserted, then by default the first indexed record for this name was used, instead of prompting the user for the correct linked record to use.


When you opened a view in its own window, then the new window was constrained to the browser window it originated from. It could not be dragged outside the main browser window.


In a resized Edit Map Location window, the mouse pointer and the blue drawing dot no longer coincided.


When using the geocoding Search functionality in the map menu of the Geopgrahical map view, the latitude and longitude metadata shown in a pop-up on the map was partially displayed outside the map.

2020-08-07: running multiple instances of the same application in a single browser

Collections 1.9.3 allows running multiple instances of the same application in a single browser. This allows you to work on different things in Collections at the same time. So you can open the same application in multiple tabs or in multiple windows of the same browser and you only need to log in or out once. So logging out of Collections will mean that the other tabs on which you had opened Collections will revert to the login screen as well: all instances of Collections in this browser will be logged out (also if the browser itself has been opened in multiple windows). You may close instances of Collections without logging out of the other instances.
For Collections to keep track of opened instances of Collections and clean up other instances when logging out, a new parameter is visible in the URL: tabId (e.g. http://localhost/Collections/?tabId=226179-27).
If you do not see this tabId parameter in the URL after you have updated Collections on the server and logged in, your browser is likely caching old code, so you'll have to clear the cache. The quickest way do do this is often Ctrl+F5 (in Windows) or Command+R (on Apple machines).

2020-08-07: a note about the Bulk create one or more records functionality

If you find that after the installation of 1.9.3, the Start and Number of records to create settings are no longer present at the top the Bulk create records screen, then a Bulk insert method still has to be added to the relevant data source in the current application definition via Axiell Designer. Only when this method is present, are users now allowed to use this functionality to add more than one new record at a time (requiring these settings).


2020-07-20: release Axiell Collections 1.9.2

Today we release the second update of Axiell Collections 1.9 (build, offering the bug fixes and new functionality described below. (Releases 1.9 and 1.9.1 were retracted.)

Bug report no.

Short problem description


Some task adapls didn't work properly (1.9.1 only).


Unable to log in: LDAP error (1.9.1 only).


"startIndex cannot be larger than length of string" error when URL to image was to be displayed in the Media Viewer.


Sometimes drop-down lists for search fields in the Standard search would not be filled with values.


Deriving records from external sources didn't work anymore.


After opening the related records view, clicking on New, Search or any other option in the main menu produced an error.


When an error message appeared on screen while editing in the result set, it wouldn't go away after a few seconds.


A certain incorrect combined advanced search broke sorting.


It was not possible to link all records from a saved search to the current linked field if the record number of those records had to be displayed in a table grid for the linked field group.


After using the Ctrl+S shortcut to save a conservation record, any following use of this shortcut would be non-operational. Only re-logging into Collections would fix the issue.


After entering a partial value in a linked field and having opened the Find data for the field window by clicking the icon next to the field, the opened window could not be closed anymore by using the Cancel button.


In some cases, sorting records decreased the number of records listed.


The increase of a result set due to sorting did not take place if the result set consisted of an opened advanced search set.


Logout could take up to ten seconds.


The record history for a multilingual field was only updated for the last addressed language in the Edit multilingual texts dialog.


The record history for a multilingual field showed an added value in a second data language as having replaced the value in the original language.


An Object reference not set to an instance of an object error occurred when entering a new linked relation value in the Thesaurus.


Export to Excel of record without language value for current data language returned a Object reference not set to an instance of an object error.


An Active Directory role with write access to the application and full access to a task, still did not see the task in the user interface.


A hierarchical search gave an Invalid object name 'CTE_3' error.


After doing a specific search, then in an unsorted result set some records showed as duplicates, with the same prirefs.


It was not possible to delete certain links from object records.


The adapl associated with a raw output was no longer executed.


Creating a new record in a particular database returned an Unexpected character encountered while parsing value: B. Path '', line 0, position 0 error.


A new record could be created in a data source without the New record method.


A specific after field adapl compiled in debug mode returned a Collection was modified; enumeration operation may not execute error.


The Invariant checkbox in the field properties could be marked while it was and is not meant to be editable. The edited checkbox was never saved. Now the checkbox cannot be marked any more.


When importing numerical values with a comma as their decimal separator into a database that uses a dot as the decimal separator in numerical fields, the comma was deleted from the imported value and not replaced by a dot.
The fix does not mean any decimal separators in the imported exchange file will now always be automatically converted to the decimal separator set globally for the database: you should first set it in the new Decimal separator option in the CSV settings dialog (which can be opened from within the Import window by clicking the Settings button).
See the next paragraph for more information about these new settings.


When printing to a certain output format, a "Session expired due to inactivity" was caused and no printout was generated.


It was not possible to open a screen tab if a non-repeatable field (erroneously) had multiple occurrences anyway.


After editing an inherited field and saving the record, the Hierarchy browser was not automatically refreshed to display the new value.


LDAP/AD Authentication on login took too long in the case where a user was a member of many groups.


Collections always opened the first application folder from a list of available application folders.


The OK button in the Bulk create window stayed inactive for no visible reason.


The Edit map location window menu had a Search option even when there was no geocoding API key present in the settings.xml.


If a field's enumerated list was dynamically copied over from a linked record, then the list was not filled with values.

New Decimal separator and Date format settings for import and export, or how to import (and export) numerical values and dates properly

Importing data from an exchange file into new or existing records in your database should not be taken too lightly: if used correctly it's a fast and efficient way to get lots of external data into your database, but it can also be a quick way to get wrong or corrupt data in your database. Let's discuss two particular data types to be careful with: numerical fields and date fields.

- The principles behind numerical and date fields

Numerical fields either have a comma or a dot as decimal separator. This setting has been made once for all your databases. Independent of this setting is how numerical values are presented in the Collections user interface: this depends on the current interface language in Collections, so even though numerical values may be stored with a dot as the decimal separator (which is the typical setting), they may be presented with a comma. For import though, the database setting is what matters.
The format (data type) in which dates in date fields are stored, on the other hand, has been set per data dictionary field and each field also has its own presentation format which may differ from its data type. Now usually all date fields in your databases will have the same date data type and if a presentation format has been applied to these fields, then that will usually be one and the same too, otherwise entering dates might get confusing. The point is though, that prior to building an exchange file you'll have to check per date field to import which format it uses: you can do that by right-clicking a date field in Collections and selecting Properties in the pop-up menu. The Field type property will indicate the data type of the field:

European date

Can only contain dates like 31/01/2002.

Date (general): this date field will accept a date in five possible formats. The following notations are allowed:

(EUR) dd/mm/yy

(e.g. 31/12/94)

(EUR) dd/mm/yyyy

(e.g. 31/12/1994)

(ISO) yy-mm-dd

(e.g. 94-12-31)

(ISO) yyyy-mm-dd

(e.g. 1994-12-31)

(Julian) yyyy-ddd

(e.g. 1994-365)

The European type date format dd-mm-yyyy is accepted as well, but is now deprecated.

DataISO aka ISO date

Can only contain dates like 2020-01-31 or partial dates like 2020-12 or 2020.

DateUSA aka American date (mm/dd/yyyy)

Can only contain dates like 01/31/2020.

- Including numerical and date values in your exchange file and setting up the import accordingly

Random decimal separators in the imported exchange file will not be converted automatically to the decimal separator set globally for the database: you should first set it in the new Decimal separator option in the CSV settings dialog (which can be opened from within the Import window by clicking the Settings button). By setting the Decimal separator option you are actually telling the import procedure that it is to expect all imported numerical values to have this decimal separator and if it is different from the globally set decimal separator for the target database, then those separators will automatically be converted to the right one. So it is important that all numerical values in the import file use the same decimal separator and that you identify this separator in the CSV settings.


If there's a mismatch between the decimal separator of an imported value and the CSV setting, then the relevant separator will actually be removed from the value, importing a value at least ten times too big! Collections won't report an error because there's no foolproof way to determine if a comma or dot is used as decimal separator or as a thousand separator (it's allowed but not mandatory to use thousand separators in imported numerical values).
Further note that in the .csv file, any numerical values using commas should be enclosed by double quotes if the Field separator is a comma as well.


The Date format option works similarly: use it to specify in which format all dates in the exchange file have been formatted. They will be converted from that format to the format in the target field in the database. This option is less strict than the Decimal separator option in the sense that it also correctly imports dates if they are already formatted in the target format whilst the Date format option has been set to some other format. For example, if the target format of the fields in the database is ISO date (e.g. 2020-07-13) and your exchange file contains dates in both ISO date and European date format (e.g. 13/7/2020 or 13/07/2020) then setting the Date format option to dd/mm/yyyy will import all dates correctly. A deviating date format in the exchange file will generate an error message and the relevant records won't be imported.
If you do not set a specific Date format for import, Collections expects each imported date to be formatted like its target field, otherwise the date will be imported incorrectly.
Further note that imported years must always consist of four digits!

- Exporting numerical values and dates

The Decimal separator and Date format settings have also been added to the export CSV Settings dialog too. In there, they do the opposite: they determine to which decimal separator or date format the exported numerical and date fields should be exported. (The database records themselves won't be changed of course.)