What are the top 3 things that ticked you off the most when opening your graphs in Revit 2023 and/or Dynamo 2.13 for the first time?
Please reply below with the top 3, and be as nice as possible. Also, please make it only apply to upgrading graphs to Revit 2023. (If you use Dynamo in another product like Civil3d and have upgraded to Dynamo 2.13+, highlight those below as well)
Background:
At Autodesk University 2022, I taught a class that highlighted all of the issues one might encounter with upgrading your Dynamo graphs to Revit 2023. (I Shipped My Scripts) It got a lot of attention and we are trying to figure out if we are missing anything.
It may or may not be obvious but we’re looking to get @john_pierson to help us resolve some of them So the more noise a problem makes, the higher chance it will make the short-list!
The three you mentioned were the first to come to mind. Node sizes and dropdown selections are a little annoying but understandable.
The thing that probably annoyed me the most was the fact that even though Dynamo would warn you of Python changes or missing versions, it wouldn’t specify which nodes. So if you had a graph with a bunch of custom nodes using python, you had to check each node (or at least each package) to know which ones auto-switched versions if you didn’t have the IronPython2.7 package.
A dialog with warnings about packages versions. It is not mentioned if you have installed a newer or older version. There is only one good solution in this situation and that is to keep the currently installed version. After that, the user could check if there is a new version of the package and walk through the process of uninstalling the package, restarting the host application, and installing the latest version of the package. Then update the graph if nodes had changed.
There are more than 3 for me… but top 3 are probably…
The size! Not just the nodes. Nothing fits and it’s so messy and I can’t see as much as I used to when I’m working (because of the size) and it’s more difficult.
The search seems to work a little faster … but it scrolls off the screen and the library defaults to the bottom of my list which is irritating because I use the top the most.
I love the fact we have real Python now… BUT loads of my Python has broken and it’s going to take aeons to fix (and I’m not entirely sure how to fix all of it yet).
The biggest problem for me is that Cpython doesnt support enums. Another issue is the version 3.8. Why not 3.10?
So a lot of the api functionality is lost by exiting ironpython 2.7.
For example the placement of colums through the api is not possible in Cpython but was possible with ironpython. This is not fixed yet.
So now this is only possible with nodes.
The strong point of dynamo was for the many possibilities of it. But this is now become very limited.
I would have decided to wait for ironpython 3 or solve the problems in cpython first before throwing away 2.7.
Hi @wouter.hilhorst - We unfortunately couldn’t wait for IronPython3 due to the lack of security patches to IronPython2 meaning we needed to remove it from Dynamo Core. Given security means different things to different people, we did make the package available on the Package Manager for all users who deem the risk acceptable So you can freely download that and keep it around as long as you wish. The flexibility that you had in IronPython2 out of the box is fully available there in the package manager.
We will be periodically updating CPython3 versions - with the current latest version being 3.9.12. Do note that this isn’t just Python, but also the PythonNET project that we use as a bridge between Python and .NET.
In time, updates to both CPython and PythonNET will solve some of the deltas that exist between IronPython and CPython. Fundamentally, however, they are implemented differently and there will never be 100% parity between the two.
#1 Node sizes - I want to see as much as possible without zooming a lot #2 library sizes - I don’t want to scroll a lot to find a node #3 Color scheme is not suitable for printed handouts or books
I have posted these images at the 2.13 release post, but they are still valid (how it is vs mock-up):
All of the above.
I remember mentioning a couple of years back in a customer research session with someone from Autodesk, that the waste of UI real estate started to get worse. Now it is not worse. Just bad.
A good UX, IMHO, implies a UI with:
No waste in it’s real estate (and this is the pain point here)
clarity, logic in it’s layout
the least search time of a function,
the least moves, (›see 0)
the least clicks to get things done
expected ways of interaction with the software
On the python side and a bit off the track of the main topic: i wonder when you expect the move to cpy and pythonNET to clash entirely with ‘us et coutumes’ of dynamo users.
I imagine a lot of custom packages dev’s have moved on and don’t necessarily have the time to migrate or learn how to migrate their code.
A ‘python for dynamo bible’ like the one of Kevin Himmerlich came out in English not so long ago, people are learning with ipy today, they cannot really learn with cpy quite yet as there is very little content available. What’s your take @solamour
I don’t mind the new look but i really dislike the lack of clarity what nodes are generating an error.
Before we had yellow nodes. I need a different color node when it gives an error.
I need to see that when zoomed out.
Size is fine, we have monocle to resize the graph.
Great other new features I’m just beginning to discover. Love the new player.
If you all can be a bit more specific regarding nodal (in-canvas) based changes that would greatly help with this effort. All of the comments around general changes and UI are useful (keep them coming!), but if you can submit items along the lines of:
Why yes, I upgraded my graph to 2.13 and none of my dictionary nodes worked anymore.
@john_pierson The book (the spanish version I got) is 600+ pages of text+ python code explaination of most of the Revit api. Oliver did a great job but the book is definitely a deeper dive into python+Revit API, try it.
Good evening, the English version is already sold out (no luck), do you know if there will be a new draw?
The Spanish version is still accessible, are there many major changes on the api between 2020 (date of publication) and 2023
So the more noise a problem makes, the higher chance it will make the short-list!
and I can’t see as much as I used to when I’m working (because of the size) and it’s more difficult.
So you can freely download that and keep it around as long as you wish. The flexibility that you had in IronPython2 out of the box is fully available there in the package manager.
