Questions, discussions, news and updates regarding the Thinkwise Platform.
Recently active
Ive been experimenting a bit with vector embedding and symantic search for a proof of concept of a customer of ours. The way it works in Thinkwise is clear to me, and I think I can get that working. However, all examples are in the documentation are based on small texts, like an email or chat history. In my POC I'm talking about legal documents of >80 pages in which I need to search. Based on that, some research all show that the best way is to chop the text in chuncks, which can be done in various ways. ChatGPT has a standard way of chopping text in 800 tokens with an overlap of 400. At the moment I use the vector embedding of chatGPT but this can become quite expensive, so I'd prefere to do as much as possible locally. I was wondering if someone already has experience with a best practice for doing so. Maybe some python service that can be called via api or something else? Or any other tips/ / tricks?
Universal GUI - Process Flow Triggers `bad_request` HTTP 400 error Dear community,When performing a process flow that has existed since 2017, but now executing it in the Universal GUI. It results in a `bad_request` error response. HTTP code 400.Within our team we have seen a set of occurrences where the process flow works as expected in WinGUI, however not in UGUI. This process flow opens a detail view, edits a row and prefills a set of fields. After the user presses ‘save’ on the form edit. A row is added to the detail of the element.See screenshot for reference of the process flow. The problem is that the underlying problem is not traceable, since Indicium does not report an error. Just the universal GUI responds with a bad_request HTTP 400 error on the `continue` request. /continue Status: 400 Source: Network Address: 212.121.123.213:443 Initiator: _app-830f5064ccd20668.js:204:2481511 Couple of important remarks:The error does not occur when running the application directly from wit
Hi you all! I would like to know a bit more about your experience with AI on Thinkwise up until now. The company I work for and me myself have interest in working and knowing more about AI in relation to Thinkwise. A few questions come to mind: Are there a lot of people currently using AI to develop? If you work with AI, what has your experience been like?Which current model (OpenAI) does fit best for Thinkwise (for each connector)?Are there any plans to add more AI providers to Thinkwise (I'd actually love to work with Claude (anthropic)?
I have a button in a screen tab that is derived from another screen. Now I saw something odd happening and I wanted to fix it so users can also add a fee here without having to spend time looking for the button. I tried changing different things within the related task hoping that it would somehow change positioning and looked into the original screentype, but with no results apart from minor changes (meaning I do have the right task, but minor change is about all I could do..). How I can make the button appear next to the other standard buttons above? What would be something I can try out?
This is a follow-up to the Testwise introduction blog, read that first for the necessary context:Disclaimers: While a Record & Playback approach is a great way to start from having no automated tests, it does have a couple of disadvantages, some of which can be mitigated by replacing recorded steps with Testwise-provided reusable logic and helpers. Some of below steps are going to change when we automate steps that are currently being performed manually Depending on the versions you install; steps might differ Prerequisites:Install Visual Studio Code on Windows Install Node.js — Run JavaScript Everywhere In your File explorer, create a folder on your local device → this will be your test project name Go into the folder, select the navigation, replace with cmd and hit Enter This will open up your terminal in the right folder. Type code . and hit Enter. This will open up Visual Studio Code. In there press Ctrl + ~ which will open up the Terminal within Visual Studio Code,
Hi community, We are running multiple tenants (application per tenant) in Azure with a file storage in an Azure Blob container. We have enabled the ‘Encryption used’ option. We are facing a problem that the client secret expires every 6 months and has to be reset using the ‘Set file storage key values (encrypted)’ (see the attached image)Any idea how we can update the encrypted client secret using the Indicium API of IAM? We are trying to automate the maintenance ;-)
Hi,I have a strange situation and can't figure out how it is caused.Our application can store documents in Azure files or on a file share based on a configuration setting.When a user uploads a document it is stored based on the physical filename and when not unique, Thinkwise adds a timestamp to the name.After the user has uploaded the document it resides at a temporary location. Using a process flow we determine the actual, unique filename based on some pattern defined in the application and then we move the document to it's definitive location (another sub folder) with it's newly determined filename. To place the document in the definitive location we use the move file connector. The origin and target locations are of the same storage type (from Azure files to Azure files or from a file share to a file share).When we execute this logic for a fileshare based document, it works as expected. The document is placed in it's new location with the filename we specified. The file contents di
Users are used to open an email in Outlook, so they can check it or add another attachment, before they send it. In Windows we use the TSFReportMailer for this. Will the same functionality also be available in Universal?
Transport companies always have their eyes on the road—quite literally. With fleets spread across countries or even continents, it’s vital to know where every truck is and whether it has reached its destination. While GPS trackers are now standard in every vehicle, simply knowing it’s position isn’t always enough. That’s where geofencing comes in.By assigning a digital boundary—known as a geofence—to each location, companies can automatically determine whether a vehicle has arrived or left. If it’s within the shape, it’s on site; if it’s outside, it’s not. Simple as that.But managing all that location data can quickly become a challenge. That’s why the latest 2025.2 update to the Thinkwise Platform introduces a powerful, user-friendly way to work with maps. Users can now easily add markers, and draw circles, rectangles, lines, or complex polygons—directly within the application. It’s a major step forward in making geodata management faster, more intuitive, and integrated into your Thin
Hi, We have a very basic, simple flow. In Windows GUI there is no problem, but in Universal GUI 2025.1.15.0.0, if i run the task succesfull 4 times i need to hit cancel 4 times, if i run it succesfull 6 times i need to hit cancel 6 times, and so on. Am I missing something obvious? is it a bug?You are meant to enter a palletnumber and click execute, and continue entering palletnumbers until you cancel the task.
As the latest Universal GUI (2025.2.10.0) is backwards compatible with 2025.1, we can already use this so enjoy some new features and a couple of bugfixes.Noting the option for the Filter form to also support filtering in Detail tabs and the older platform not having the settings for them yet (As far as I could find), should perhaps the Filter form listen to the extended property “DisableFilterOnDetails”?Especially considering the possible performance impact, I probably would want to disable detail filtering all together for now until we are on 2025.2 and can decide for certain details what we want to support.
We use barcode labels with the barcode font free 3 of 9 extended.In the windows GUI this works fine.With the thinkwise reporting service this works fine as well.Both this methods will become obsolete so I’m re-develop these reports to the Universal GUI.On which places do we need to install the barcode font we need in the report.Should it exists on each device running the Universal GUI?Should it exists on the webserver where Indicium is running.Till now the labels are coming out the Zebra Printer but not with a barcode. In the place of the barcode is now a human readable number, which means the ‘place’ where the report is generated doesn’t recognize the barcode font. Can someone help me out this puzzle?
Planning lies at the heart of every successful organization. Whether you're allocating resources, managing personnel, or scheduling production, having the right tools in place can make all the difference. With the release of Thinkwise Platform 2025.2, we’re excited to introduce a powerful set of enhancements to our Scheduler component—designed to help you plan faster than ever.This update brings hierarchical resource grouping, external drag & drop, and the ability to apply conditional layouts to both resources and time cells. These features give you much more flexibility and control over your Scheduler.In this blog, I’ll guide you through the essentials of setting up a Scheduler in Thinkwise, exploring each new feature along the way. This blog is the first part in a series on using the Scheduler. This first part is focused on setting up the scheduler. The second part focuses on interacting with the scheduler, and can be found here: Scheduler setup guideIn this blog, we’ll create a
I need to consistently show 2 decimals, including when the number is a full integer, or only has 1 decimal. Searching the forum I found references to custom_display_format, but I have no idea how/where to implement it. For example, here it should be 2.25, 1.00, 2.00 and 1.50:Example without trailing zeros Alternatively I could use a calculated field, but converting numbers to strings in sql for this purpose would be inefficient (and a lot of extra work and complexity), so I'd like to avoid it.
Hallo, I’d like to confirm whether the hyperlink lookup feature is included in the 2025.2 release? I had previously asked whether it's possible to add a hyperlink to each field value.I’m working with a table (or entity) displayed using a grid component, and I’d like the values in certain fields to act as hyperlinks that redirect to another table, subject, or subject variant.Those are the functionality I’m referring to. Additionally, is there any documentation or a set of steps available for configuring this feature? Thank you
Hello everyone,When we plot a route with the Maps component in our software, it plots the route correctly, it doesn't have any weird data in it and the route is generated satisfactorily. But there is a strange phenomenon: for every route it plots, it alwasy sets a dot on the coordinates (0, 0). That causes it to always zoom in on Null Island in the Atlantic Ocean. What could be going wrong?The screen when I open it with a dot on Null Island (long: 0, lat: 0): These coordinates cannot be found in the configuration of the route.The actual route it plots and should zoom into:
With release 2025.2, the filtering experience in the Universal UI has been significantly improved and expanded. The styling of the Default filter has been updated, making it more user friendly, and Advanced filter options have been introduced to support more detailed filtering capabilities.We strongly recommend placing the Filter button on the Action Bar, next to the Search field. This approach provides users with clear and immediate access to filtering functionality. From 2025.2 onward, an Enrichment can update existing Models and new Models have it by default.This blog outlines the thought process and design choices behind these updates, and aims to provide insight into how the filtering experience has evolved.Filter pop-upIt starts with the updated filter pop-up, which now combines default filtering, detail filtering, and advanced options in one central location. The default filter allows for quick and straightforward filtering, including detail filtering on specific fields. By togg
In previous releases you could modify the type of process step (always, successful, not successful) by double click the arrow in the process flow and change the type. This is not possible anymore (2025.1). I could not find something related to this in the release docs. Is this a known issue or is there a reason for?
Hallo,I have a module in my web portal under that allows adding, editing, and deleting data.In this model, I'm working with two fields: Company and Liners. Both values are derived from the same table called Company. To differentiate them, I created expressions where:company_name includes companies starting with "NL" liners includes companies starting with "L"My goal is to display these as two separate dropdown lists—one for Company and one for Liners—even though the data comes from the same table.I initially tried using a company_id lookup from the Company table, but that displays all records, which isn't what I need.So my question is:Is using expressions the correct approach for filtering the dropdown values? If not, what is the best practice for achieving separate dropdowns for Company and Liners from the same table based on different filter conditions?Below is the webportal page and the current data model table. web portal data model
I have a cube view with expanded cube fields (row and column). The behavior in Universal seems different from Windows. In Windows columns are expanded (year → month). In Universal the year is collapsed and shows year totals. What could be the reason for this behavior?
This is the second installment in our blog series on the Scheduler.In Part 1, we walked through the initial setup of the Scheduler. Now, in Part 2, we’ll explore how to effectively interact with the Scheduler to enhance usability and user experience.The Thinkwise Scheduler supports a variety of interaction methods, such as creating activities by clicking on empty time slots, editing them through double-click actions, and assigning external resources via drag-and-drop. In this blog, we’ll demonstrate each interaction technique and show you how to integrate them into the Scheduler we configured in the previous part. Interaction MethodsClick to create a new ActivityThis feature allows users to click a time cell to initiate a task for quick data entry directly in the scheduler.Start by creating a new task ‘project_planning_scheduler_add_activity’. Add task parameters for all fields that are needed to create a new activity:Assign this task as a table task to the scheduler view. To make thi
As a QA team within our Product Innovation department, we understand the need for efficient, reliable ways to test applications that are built to last forever. Quality is an essential pillar of our brand promise; we want you to trust our technology and enable you to develop your functionality 10X faster. See this old blog for some background reading on our role:We place a strong emphasis on end-to-end, user-focused testing—a shift-right approach—because it provides the level of confidence that we need to say we're delivering a quality product. For a system as complex as ours, investing in test automation is essential. It allows us to rely on consistent, repeatable checks, so that any manual testing efforts can be focused on where they’re most valuable: ensuring that new features and ongoing development meet our quality standards.Since we are shipping a low-code Platform -with which all of you build amazing Applications-quality is a shared responsibility. Both the quality of the Platfor
Hi everyone,In our application (using Universal GUI), we have some custom components. We experience some problems when we want to style the custom components depending on choosen mode (dark / light) in the user preferences by the user.The custom component is presented in an <iframe> in the browser, however, within the <iframe> the css class ‘dark’ is not applied, causing the custom components not to be adjusted to the selected mode (dark/light). This makes UI/UX of application not as good as we want to.Any ideas how to solve this?
We are using the Windows GUI (2 tier). Recently users reported a problem that they cannot save their own created or changed cube views. Users can make the changes and save their cube views with no errors or messages showing, but afterwards the change has not been applied or saved to the IAM database (MS SQL server). As a admin user myself i can save the cube views with no problems as i have more rights on the database itself.We also recently upgraded our environment from 2023.1 to 2025.1. I suspect this has to do with the change in cube view preferences where cube views are split into cube views and cube view classic preferences. Also a change has been done in the upgrade where public rights have been revoked and assigned to specific roles. Our users have the end_user role. I have checked execute rights have been granted on below procedures and the task_delete_usr_pref_classic_cube_view. This is the case. Also the rights on the underlying tabels i have set to inserted, update, delete t
Hi, I have been using the topic: Upload multiple files using the Universal GUI for drag and dropping files in my test environment. The test environment is hosted on Azure and uploaded the files to an azure storage. However, the Production environment is hosted on premise, therefore I wanted to test the drag and drop functionality on the acceptance environment (also hosted on premise). However, I can't get the drag and drop to work. I have changed the URL in the html to the right url. The HTML file is saved on the universal folder, same as the test environment. However, I just keep seeing the 'no preview available’. The file name column has a file upload control type to a file system storage location. Uploading and opening the files from the local storage works. Kind regards,Mark
Already have an account? Login
No account yet? Create an account
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.
