Hi ,

I'm using Talend Data Preparation 8.0.1 On-Premise on Windows.

When I navigate to Datasets, add Dataset, the only available option is "Local file". I would like to enable other connection types, particularly Microsoft SQL Server.

I have already configured the following properties in application.properties:

dataset.imports=local,tcomp-JDBCDatastore,tcomp-SimpleFileIoDatastore,tcomp-SalesforceDatastore,tcomp-S3Datastore,tcomp-AzureDlsGen2BlobDatastore

tcomp.server.url=http://host:8989/tcomp

Additionally, the Talend Components Catalog Service is running, and the jdbc_config.json file contains the SQL Server JDBC driver configuration.

However, the other connection types are still not displayed in the Add Dataset menu.

My questions are:

Is any additional configuration required to enable JDBC connections in Talend Data Preparation?
Are there any additional services, dependencies, or permissions required?
How can I enable the Database option to create datasets directly from Microsoft SQL Server?

Thanks.

Hi ,

I'm using Talend Data Management Platform 8.0 with Talend Administration Center (TAC) and Talend JobServer in an on-premises environment.

I would like to know if it's possible to trigger a task configured in TAC and execute it on JobServer through a REST API request.

Alternatively, is there another way to execute jobs remotely from an external application or interface without accessing the TAC web console?

Any guidance, documentation, or examples would be greatly appreciated.

thanks.

Hi to everyone,

After upgrading Qlik Sense for Windows to the release May 2026 Patch 3, when I try to open the Load Script for any published app, I can't. The option "Load Script Editor" is missing. This option was present for the release Nov 2025 Patch 6.

I tried to create new rules to permit it with no luck. 

I know, that the way can be to duplicate the app, and see the script, but I would like to know if this behaviour was set by design, and any workaround can't be set via security rules, or so on, or if it's possible.

Was anyone able to set the behaviour I'm missing? if so, can you explain me how to set it?

Thanks in advance

Thanks

Hi team,

I tried to set up this LLM but currently, to make the test with Swagger on Power Automate to authorize the OAuth session, it indicates that the redirect URI is not registered, and I made this in the before steps.

Some have the same problem or know other steps to work with Qlik Cloud MCP and Copilot.

Thanks

I'd like to check whether the architecture below is supported and recommended, or whether we should change it.

Environment
- QlikView Server (May 2026)
- Qlik NPrinting May 2026 SR1: 1 NPrinting Server and about 11 NPrinting Engines
- QMS and DSC on one server, with a clustered Publisher

Design
- Cluster A (2 QVS nodes) serves end users through AccessPoint on QVWS #1. Users open the documents and trigger NPrinting On-Demand from there.
- Cluster B (3 QVS nodes) is dedicated to NPrinting. The NPrinting connections point to Cluster B through QVWS #2 (qmss://QMS:4799/QVWS@web2/...), with CPU usage load balancing.
- Both clusters mount the same document folders on the same file share, so they share the same .qvw and .Shared files.

Please elaborate on your input, and thanks in advance!

Dear Support Team,

I am writing to request clarification regarding the behavior of the use_manipulation_pk_for_apply parameter and an apparent discrepancy between our test results and the current documentation.

1. Current Documentation

According to the manual, to enable this feature, the parameter should either be:

• Removed (unset), or

• Explicitly set to OFF.

2. Observed Behavior / Verification Results

Based on our internal testing, the actual behavior appears to contradict what is stated in the documentation:

• When use_manipulation_pk_for_apply is unset or set to ON: If a configuration exists to arrange primary key (PK) values in a specific order, the system appears to attempt to reorder and apply them according to that sequence.

• When set to OFF: This reordering process does not seem to be triggered.

• New Task Comparison: Based on our observation, the reason this unexpected behavior does not occur in newly created tasks seems to be simply that no specific PK ordering has been configured for them.

Therefore, our verification suggests that the feature is actually enabled when the parameter is unset or set to ON, rather than OFF.

3. Questions / Confirmation

Could you please confirm the following points?

1. Is the statement in the manual incorrect, and should it state that the feature is enabled when set to ON (or unset)?

2. If the manual is correct and setting the parameter to OFF is indeed the intended specification to enable the feature, is the observed behavior where ON triggers this process considered a defect/bug?

We would appreciate your clarification on the intended behavior and documentation.

Best regards.

This Techspert Talks session covers:

• AI in Qlik Cloud
• Data prep best practices
• How Answers works
• Upcoming improvements

Chapters:

1. 02:21 - Qlik's agentic architecture
2. 03:58 - Different data model examples
3. 06:33 - Improving field names
4. 07:07 - Field visibility
5. 07:55 - Semantic Understanding
6. 09:02 – Synonyms
7. 09:30 - Impact of app size
8. 09:41 - How Answers works
9. 10:41 - Writing better prompts
10. 12:05 - Library and new apprentice analogies
11. 13:01 - Learning.Qlik.com
12. 14:05 - Roadmap: Data Readiness Score
13. 15:26 - Roadmap: Skills
14. 16:13 - Roadmap: Multi-Application Assistants
15. 18:45 - Q&A: What technology is behind Qlik Answers?
16. 20:33 - Q&A: Can we use skills with Answers?
17. 21:31 - Q&A: How to create a Knowledge Base?
18. 22:18 - Q&A: Will Insight Advisor go away?
19. 22:50 - Q&A: How can data be restricted with Answers?
20. 23:22 - Q&A: How to teach that prompts matter?
21. 23:57 - Q&A: How to provide LLMs with full context?
22. 24:32 - Q&A: Should every field be a Master Item?
23. 25:03 - Q&A: When will LLMs be able to create scripts?
24. 25:31 - Q&A: Why not use LLMs directly with Snowflake?
25. 26:36 - Q&A: How to monitor Qlik Answers usage?

Resources:

• Qlik Learning course: Agentic AI with Qlik Answers and Qlik MCP Server
• Qlik Learning course: Artificial Intelligence and Machine Learning Foundations
• Free Learning: Agentic AI Adoption Awareness

Click here to see the video transcript

I have a table that currently works - it allows the user to select WeekStartShipDate, which is loaded in the load script with a Floor value, filtering the table to just one week:

The variables vStartSixWk and vEndSixWk determine the date range for the 6 Week column. 

vStartSixWk  =(WeekStartShipDate) -42

vEndSixWk  =(WeekStartShipDate)

The 6 Week column is then calculated as follows:

Sum({<WeekEndShipDate={">=$(vStartSixWk)<=$(vEndSixWk)"}>} EquivalentUnits) / 6

The % of 6Wk Total column is calculated as follows:

Num(
Sum({<WeekEndShipDate={">=$(vStartSixWk)<=$(vEndSixWk)"}>} EquivalentUnits) / 6
/
Sum(
TOTAL
Aggr(
Sum({<WeekEndShipDate={">=$(vStartSixWk)<=$(vEndSixWk)"}>} EquivalentUnits) / 6,
ShipTo
)
),
'##0.0%'
)

So, that all works fine until I add another dimension column, such as SalesOrder. Once I do that, the date filter breaks, and the 6 Week expressions return zeroes.

Please advise.

• A tree view of your application content: Sheets and the objects on those sheets are shown as a hierarchy, similar to the layers panel in design tools.
• A logical overview: See the full structure of your app at a glance, instead of moving through sheets one at a time.
• Search and find: Look up sheets and objects by name and get to them quickly.
• Built for understanding and organizing: Whether you inherited an app or built it yourself, the tree gives you a clear picture of what’s in it.
• Drag 'n' drop into layout containers, and change z-direction within them.

• How the tree view works on your complex applications
• Whether the structure matches how you think about your application
• Whether it helps you find sheets and objects faster than you do today
• Content or actions you expected to find in the tree but didn’t
• Anything that feels different from what you’re used to, even small things

• Bookmarks are not included
• The application styling property panel, map object layers, and fields/measure used by charts are not part of this preview.