I think the lack of warning here is a major issue. Simply removing it with no transition period is completely unreasonable in my opinion.
We have probably 50 workflows that I now need to migrate, many that use local storage.
This has created an enormous amount of work for me
Typical bait and switch play we’ve seen many times over in OSS space : Piggy back off the community for feature development and then start locking the features behind the paywall.
But I don’t think knime has what it takes to execute it in a commercially viable manner (user base too small, they are more niche than they’d like to admit, but the product is great of course).
Wonder if the new SaaS vision did not play well with the original founder/ceo and why he was ousted (one can only speculate). Why can’t we just have an option to pay for a managed service if it’s so much hassle for users to set up batch. Some will choose it and some will prefer the hassle.
I think next on the horizon will be cloud only access, per-execution pricing models or some ridiculous row-based caps in the free tier (if there will be free tier).
From my pov knime is feature complete + new extensions can further extend functionality. I hope new extensions will be backwards compatible and a reliable OSS fork will appear. To me this is a clear signal to migrate off knime for any greenfield projects and maaybe use it as a SQL wrapper to make my workflows more vendor agnostic (but I do find myself using Marimo notebooks for everything more and more often). I used to argue with low code dev haters saying knime is clearly different (“open source” in big letters on KNIME’s home page) but unfortunately they were right - major vendor lock-in is on the horizon .
Apologies for the rant, I really love the product and would hate to see it ruined by a short sighted silicon valley/VC bro. In the SaaS game you only need to position your company for the next sale to the highest bidder and long term business health is not the objective here sadly.
yep.
and quite a few more things. Interestingly, there was also a sudden wave of newsletter e-mails in the last 2-3 weeks. Guess the pressure was now increased.
Good thing AI programming is becoming more viable and as @Add94 wrote, there a quite a few very interesting tools showing up like Marimo
This was posted in Dec 25. Just a regular post saying it was still supported with helpful instructions how to install. Never before have a seen such a feature removed with no warning without even considering that people rely on this functionality and giving them time to migrate to Pro or leave KNIME for an alternative.
I’ve been an ardent supporter of KNIME for years but this has shattered that impression. I really hope you reconsider.
@mlauber71 Does NodePit Batch depend on the KNIME batch executor to function or us it using some other method to run that will continue working?
@bobpeers I don’t know. Maybe @danielesser can speak to that. The knimepy package does rely on the current batch.
NodePit Batch does not depend on the KNIME Batch Executor. Based on our current understanding, the announcement in this thread should therefore not affect NodePit Batch.
NodePit Batch supports KNIME versions from 4.7 up to 5.11, as well as the current nightly build (5.12). That said, we will continue to monitor the situation closely and evaluate any potential impact that further changes in KNIME’s core might have on NodePit Batch going forward 
Yes I’m currently using the knimepy package that is now unsupported.
@danielesser Thanks for the quick reply! If I have questions is it best to mail you or book a demo, I see there are not many time slots available for demos so guess you are busy 
That’s great news @danielesser , if that’s the case. I predict you will have an influx of people interested in this!.
I think I can safely say that you are not alone!
Right now I’d happily pay for this since the alternative is both lost functionality and a huge amount of work. KNIME Pro simply doesn’t support the flows I use extensively (local storage access, SQL database not accessible from the internet).
Having had 24 hours to digest this a few things occur to me:
more importantly now is the question if other features will be cut too, like the nodes to read and execute local workflows, or the call workflow nodes that work on projects in the same workspace.
(and how much trust to put into any statement we receive now)
Bob,
I am accessing my local Postgres database from KNIME team plan data apps. Can you tell me why it does not work for you?
@Marc_Goossens Honestly I haven’t tested, but our SQL db sits in Azure and is not exposed to the internet so a flow runnnig on KNIME’s hardware using Pro is not going to be able to access this unless I open the firewall to KNIME which I’m not in favour of doing in this case.
KNIME also say themselves:
You can upload and download data directly from your personal space on KNIME Community Hub. In addition, you can connect to cloud-accessible data sources using KNIME’s connector nodes, such as Google Cloud, Azure, Microsoft sharepoint, box, and many more.
Please note that since KNIME Pro is a cloud-based service, workflows can currently only access systems and data sources that are reachable from the cloud environment. Access to on-premises or private network resources is not supported at this time. However, we’re actively working to expand connectivity options in the future.
@fe145f9fb2a1f6b That’s the thing with trust, hard earned, easily lost.
and it doesnt help that the nested calling within the open workspace using Knimes intended nodes is riddled with bugs
I’ll just add that after some testing NodePit Batch does work as a drop in replacement to KNIME Batch Executor (tested on 5.11) The only option missing for me is the –nosave option but it’s not a blocker as –reset makes sure it starts with clean data on every run.
Yep…it works, but indeed you would have to tell the firewall to permit access the database server port.
