Worth noting that Software AG has an integration product called webMethods Integration Server (now all owned by IBM, I believe) that has visual language called Flow[0].
Show HN: Flowcode – Turing-complete visual programming platform
41–50 of 82 posts
Re: Show HN: Flowcode – Turing-complete visual programming platform
#42What is the error handling story? I'm shocked that Control Flow doesn't contain Try-Except
Anyway, I ask because Write File returns nothing https://docs.getflowcode.io/docs/reference/file-system/write...> but there's also no discussion on that page of what happens when one cannot Write File (and the ways that can fall over are almost unlimited)
Also, please fix the inputs bullet list, that's a pretty glaring "no one reads these docs" bug. While you're there, never, ever, ever, ever, ever just have the documentation for anything be a repetition of the input variable's name. I can read, what I can't do is know what the domain of possible encodings are available (base64? base85? uuencoding? gzip+base64 https://docs.cloud-init.io/en/24.3/reference/modules.html#wr...>?)
Re: Show HN: Flowcode – Turing-complete visual programming platform
#43If you can't write the software that runs the flowcode with flowcode it kind of makes it simply a different notation and not a revolutionary paradigm shift?
Re: Show HN: Flowcode – Turing-complete visual programming platform
#44We use an internal 'no code' tool to do complicated data transformation logic. Our flow has become so large and convoluted, it is extremely difficult to put feature and debugging is absolute nightmare.
You have no idea how many times we had wished all this data transformation was in vanilla Java/Python. Our life would have been so much easier. But at this time, it is too large to ignore.
Re: Show HN: Flowcode – Turing-complete visual programming platform
#45Congrats on your launch! Looks very good! We are building a similar product at www.midio.com and it seems you have come to some of the same conclusions as us about visual programming.
Re: Show HN: Flowcode – Turing-complete visual programming platform
#46So far there are zero data points on the success side of the fence when it comes to visual programming. The exceedingly rare times I come across it professionally it's a toy project or something that's basically in the state of "We wish it was just code but we don't dare touch it".
Re: Show HN: Flowcode – Turing-complete visual programming platform
#47I love your premise here. I think visual programming has such untapped potential even though it has been tried and tried and tried so I hope you crack the code a bit (lol). I think you're exactly right on the idea of "I just diagrammed all the control flow why can't I just push run". It's kind of crazy to introduce a point of failure (and a menial translation step) by writing code to match the diagram. Another huge b…
Author here. Thanks! Lowering the barrier of collaboration is indeed a major goal for us. In the DRAKON analogy, Flowcode might help technical product managers collaborate with developers on LLM-heavy logic. Re: vibe-coding - that matches our vision 100%. Let AI build the nodes while a human (even if heavily AI-assisted) oversees
Great stuff, wish you every success with Flowcode.
It reminds me of Node-RED which seems very similar but is aimed at IIoT applications.
I’m working on a version that is based on Erlang and not NodeJS which isn’t a true concurrent language.
For me, flow based programming makes the most sense with languages which are truly concurrent and message passing is baked in, ie Erlang.
One thing I’ve come to realise is that visual programming is a hard sell, hence my effort to create a visual environment for Erlang - to introduce visual programming to a different set of developers.
Re: Show HN: Flowcode – Turing-complete visual programming platform
#48So far there are zero data points on the success side of the fence when it comes to visual programming. The exceedingly rare times I come across it professionally it's a toy project or something that's basically in the state of "We wish it was just code but we don't dare touch it".
Node based programming is pretty common for visual effects, like shader graphs (Unity), animation (Houdini), colour grading (DaVinci), etc.
Re: Show HN: Flowcode – Turing-complete visual programming platform
#49So far there are zero data points on the success side of the fence when it comes to visual programming. The exceedingly rare times I come across it professionally it's a toy project or something that's basically in the state of "We wish it was just code but we don't dare touch it".
What about Blender nodes? I’ve seen some pretty cool and fairly complex generative modeling using nodes.
For a simple example, here is Conway’s Game of Life. I’ve seen much more complex things implemented as well.
Re: Show HN: Flowcode – Turing-complete visual programming platform
#50Visual workflow tools, particularly those that are able to support a human-in-the-loop, MUST exist for the world to adapt and effectively use LLMs.
Literate Programming - coined by Donald Knuth and at the heart of Jupyter Notebooks - is powerful means of making software development a tool in the pockets a lot of people. I agree that just pressing "run" is critical to making a tool like this accessible.
I'm one of the core contributors to SpiffWorkflow. We are also visual programming platform and echo your belief of the power of combining visuals with programming. I shared this article here a few months ago that delves into our methodology. - https://medium.com/@danfunk/5-ways-bpmn-python-will-simplify...
conartist6 mentioned elsewhere in the comments of a need for a common standard for editors - I think BPMN could be that standard, or new revolutions of it, such as POWL.
collingree mentioned "good visual programming makes the software development process more accessible to other smart people" -- I think this maybe the MOST important long term benefit of these tools. More people in organizations need to understand what the hell is going on - and stop shrugging their shoulders when it turns out all the business logic is buried in 100k lines of code. Important shit should be easy to find and change.
As many others have mentioned here, visual tools are not likely to see wide adoption by software engineers for all of their work. But as middle groud in which to plant business rules, so those rules can be cultivated - it could change the world and the way we collaborate on software efforts.
Would love to talk to you both sometime, share war stories, dreams, and go to market strategies.