2020-06-10: release Axiell Collections 1.8.3

Today we release the third update of Axiell Collections 1.8 (build, offering the bug fixes described below.

Bug report no.

Short problem description


Attempting to print a workflow after completion made Collections crash and shut down all user sessions.


Hierarchical searches (e.g. with generic, narrower) didn't work on alpha-numerical indexes.


Uploaded jpeg files with .323. in the filename were not accepted. The .323 part was incorrectly interpreted as the file extension.


When before deletion of a mission item record an adapl tried to prevent that deletion, an Unable to cast object of type 'Adlib.Eval.AdaplErrorEventArgs' to type 'Adlib.Eval.StringEventArgs' error was generated.

2020-06-03: release Axiell Collections 1.8.2

Today we release the second update of Axiell Collections 1.8 (build, offering the bug fixes described below.

Bug report no.

Short problem description


A Bad Request error message occurred when creating a Workflow record.


Printing to a certain output format gave a "Specified cast is not valid" error.


It was not possible to 'manually' link a mission item record to a mission using the record editor. Selecting a mission item from the link screen returned a 'bad request'.


The record number (which is used to identify mission items) did not show in the link screen for mission items.


If a linked field had a variable domain, attempting to create or edit a record with that field gave an error.


People records were not updated when they were added to new domains from within a linked field in another database.


A Result set view filter on an enumerative field didn't work.


Certain tasks couldn't be started, because the OK button in the task screen remained greyed out, although all mandatory fields in the task screen had been filled.


A combined search with a fixed query yielded an error about a field that could not be found.


Drop-down lists could be very slow to return cursor control in Workflow screen.


Certain forward relations for reversely linked fields were not visible in the Related records view.

2020-05-13: release Axiell Collections 1.8.1

Today we release the first update of Axiell Collections 1.8 (build, offering the bug fixes described below.

Bug report no.

Short problem description


The URI data type did not check for a valid protocol/schema.


Certain searches returned Session Expired in the browser.


The Unsorted option for a field column was missing in the Result set view and in the Saved searches dialog.


A record appeared empty and caused an internal server error 500 when marked, when mono-lingual data was present in the same multi-lingual field occurrence as actual multi-lingual data.
The fix means that any mono-lingual data in a multi-lingual field (which shouldn't occur in the first place) will now be ignored if it is located in the same occurrence as multi-lingual data.


An error on tag rO occurred when trying to create a new collect record in a specific application.


When the default value for a field is a non-preferred term, the default value was not replaced by the preferred term when the record was stored.
The fix means that the preferred term replacement is now done immediately when the new record is created.


A new object number in an import file, while importing with the Match and/or Update only option set, led to an "object reference not set to an instance of an object" error.


A default value was added to an already saved record if the field was still empty, and then the field could not be cleared.
The fix means that default values are never added to existing records, but if they appear in existing records they can now be cleared. Default values are only added to new records.


Collections didn't allow the editing of records with a circular reference (to correct the circular reference).
Normally, circular references (in hierarchical relationships for example) do not appear in data, but if they do they can now be corrected by just editing the record and deleting or changing the problematic reference. Collections now actively checks if a reference is not circular when you add a reference (and only then) and you'll have to correct it immediately, otherwise you can't save the record.
A circular reference occurs when you enter the same reference in both the Part of and the Parts field, for example, or when you enter a reference in the Parts field that is actually already a parent higher up in the hierarchy to which the current record belongs.


A multilingual field, filled by ADAPL, could be made to contain an instance with an empty language attribute.


Printing to a prompted output format didn't show the prompt screen and gave an error, when there was also an adapl associated with the output format.

2020-04-28: release Axiell Collections 1.8

Today we release Axiell Collections 1.8 (build, offering the new functionality and bug fixes described in the paragraphs below.

Bug report no.

Short problem description


When selecting a term from outside the fixed domain for the current linked field, in the Find data for the field window, either nothing happened or a Bad request error occurred.


Sometimes, workflows gave a "session expired" error.


Collections did not support the created attribute in records ingested with the WebAPI.


The auto-populating of URI fields didn't work anymore.


The Reset function in the Result set settings did not work.


The default invariant language setting was not working for filled-in multilingual fields after saving those changes for the first time.


Deployment of a new Collections version often fails because mscvcr120.dll is locked by another process. The issue itself cannot be solved, but when it happens, try recycling the application pool and possibly stopping the website temporarily and deploy the same package again: the error should not occur this time. (A stopped website should be restarted again.)
Also, the deployment package has now been renamed from Axiell.Alm.MVC.KendoSPA.zip to AxiellCollections.zip, a more logical name.


The Field History for linked fields and merged-in fields would display empty.
The fix does not mean that now the previous linked value or previous merged-in values are shown in the Field History (which can be opened by right-clicking a field and selecting History in the pop-up menu), but that for each of those fields just the current and previous linkref value of the associated link reference field is shown. This is still the only available historical value for linked fields, their link reference fields and any merged-in fields, because even without selecting a different value for the linked field, the linked record might have changed: such changes cannot be kept in the current record, only in the linked record itself. For example: current_location.barcode (2N) is a merged-in field with the current_location.name (2A) field with link reference tag 2a. The field history for the barcode field will now actually show the field history for tag 2a. The Field Name header row in the Field History indicates current_location.barcode (2N) [link => 2a], and the values in the Old value and New value columns mention [linkref] behind the value to make this clear.
Note that of linked fields, their link reference fields and any merged-in fields, the Record History (which can be opened from the Result set context toolbar) only lists the linked field, yet again with the historical value of its link reference field. For example:


The layout of the map menu was broken.


By default, integer fields were displayed as formatted with thousands separators, but that is not always the desired behaviour: in the library Search year field for example, an integer field of length 4, one would like the year to be displayed without a dot or comma.
The fix means that by default, integer fields will now be presented without dots or comma's. If the dot/comma presentation of integers is still required for a field, it must now be set via the new Presentation format option for integer data dictionary fields in Designer, by your application manager.


Moving locations in the hierarchy browser, by dragging and dropping them to a different location, if sorting was set for tag nt (the narrower field), had no effect.
The fix means that if sorting has been set for a field in the hierarchy, then trying to move a term within that hierarchy will only allow this to be a change in the hierarchy, not in the sorting order (as the latter would be overruled anyway).


An error was thrown upon opening a template containing an image in a table cell with no exact row height, when an image cache was set in the settings.xml.
The lack of a fixed image height in a table cell causes a crash on resizing the image with a height value of 0. The fix means that the system will now make the height value equal to the width value per default if no exact row height has been set.


In Virtual scroll mode, the Word wrap setting for the Result set view caused the record list to be displayed incorrectly.
The fix consists of the Word wrap setting now being disabled for Virtual scroll mode, because the two settings are really incompatible.


Sorting the result of an expand search reduced the number of found records: all items with inherited data in the search field had then gone missing from the list.


When editing in the Result set, a (variable) link domain tag value was not used when forcing a term.
The fix consist of a confirmation message first being displayed: now, when a thesaurus term without the proper link domain is selected, there is first a message box asking confirmation to add the domain to the term. If you confirm, the term is selected and the domain is added to the thesaurus term. If you deny confirmation, the thesaurus term can't be selected.


Saving the record after marking a checkbox, did not save the ticked status of that checkbox.


An output format ADAPL errorm message was not shown.


Turned on views (aka widgets), like the Geographical map view for example, did not automatically turn off in databases where not applicable and you could not switch them off manually because the icon was missing.


After a new Workflow record was created, the Refresh button in the Workflow list did not refresh the list with the new record when the list was sorted.


After editing a linked record via its zoom screen, the merged-in fields in the main record weren't updated.


It was impossible to enter a decimal separator character in numerical field input (with default settings) for interface language Dutch: an entered comma would be removed from the entered value on saving the record.
The Presentation format option for numerical fields is not relevant for Collections as Collections would refer to the locale of the web server, which is not relevant in a multi-user web environment. Therefore, Collections ignores the setting and uses the currently selected language/country of the current user to derive the locales. Thus, all numeric formats are presented in a format that is natural to the user’s UI locale. In the case of Dutch (NL) this will be using commas for the decimal separators and periods as thousands separators. This only concerns the presentation format though. Which decimal separator is used in the stored record in the database (not visible to the user) depends on the database settings.


Newly created documents in application fields could not be written or downloaded.

2020-04-20: updating existing records with the import functionality

The Import functionality in Collections has been extended with the option to update existing records, instead of just importing new records.


The options are now as follows:

Leave the new Match and Update only options empty: all records are imported as new records (errors excluded of course), no existing records will be updated.
Only select a indexed field from the Match field list and leave the Update only checkbox empty: existing records will be updated with values from the import File if the imported value from the Match field is found in the relevant field in an existing record; if that value is not found, a new record will be created for the imported value row.
The selected Match field must always be a uniquely indexed field, like priref (the record number field) or object_number (if object numbers must be unique in your application) because otherwise the import procedure doesn't know which record to update.
The selected Match field doesn't need to be selected in the field list to import too (this is the large box in the right bottom corner).
In existing records, all field content that is not being updated, will be left as is, so no data is removed from those fields.
In existing records, when a field to be updated has more than one occurrence already, then always the first occurrence will be overwritten by the imported value. So if the existing record has two occurrences x and y, and the imported field only contains the value z, then the resulting two occurrences will z and y.
Importing multiple occurrences of a field is possible if the relevant field name is repeated in separate columns in the import file. So if object_name appears in two columns then the import file may contain two object names. If it does, the two first occurrences of the field in an existing record will be overwritten. If only the first occurrence in the file is filled while the second is not, then the first field occurrence in the existing record will be overwritten while the second existing occurrence will remain as is. If only the second occurrence in the file is filled while the first is not, then the second field occurrence in the existing record will be overwritten while the first existing occurrence will remain as is. If both occurrences in the import file are empty then both field occurrences in the existing record will remain as is. A similar logic applies to three occurrences or more. Do check in the Collections interface if the field is indeed a repeated field.
It is not possible to have the imported value(s) added as new occurrence(s) and it is also not possible to have all existing occurrences emptied before importing the value(s) from the import file.
Note that the field names are data dictionary field names, and those might be different from the field labels in the Collections user interface. To be sure you have the right field, just hover the mouse cursor over the entry field in the Record details to show the system field name in a tooltip.
Instead of scrolling the Match field list to find the field you are looking for, you can also start typing the field name in the Match entry box to limit the displayed list to fields starting with what you typed.
If you want to empty the Match entry box (to remove an earlier selected field), scroll all the way up and select the empty value on top.
Mark the Update only checkbox (next to selecting a Match field) if you never want any new records to be added if no existing record to be updated can be found for a record from the import file.

To import all fields from the import file, simply leave the field list on the right empty.