Saved the day! But also some bugs.

First of all, you guys saved my day / week/ month.

I was running an unreleased multi-objective optimization plugin, based on adaptive, relaxed Pareto front formation combined with the non-dominated sorting genetic algorithm NSGAII.

But, after weeks of work I found out it wasn't compatible with the latest versions of Rhino/Grasshopper (d'oh).

However, in one hour I had your program running with my generative model, great!

Now, the sad part: often after saving a Grasshopper file with some optimization history embedded and reopening the file I get the following error and the component is gone.

Could someone give some insight in this problem, and maybe provide a fix?

Thanks!

Replies to This Discussion

Permalink Reply by Robert Vier on December 11, 2013 at 2:44

i ll try to fix this the next days, suddenly occured with the last versions ..! no idea why so far. all im doing is serialize my class using the binary formatter, then compress it with c# s built in methods..
lets see!

Permalink Reply by Robert Vier on December 11, 2013 at 8:26

could you eventually provide an example where this happens for you?

best

Robert

Permalink Reply by Daniël Smidt on December 12, 2013 at 3:22

Great to hear there is a follow up.

I'm using a model with a lot of dependencies and parts which cannot be published as of now.

My study is on freeform architecture and non-standard tessellation in relation to structural performance and some architectural objectives. I'm using EMO (MMO) to search the solution space and a fuzzy neural tree for performance evaluation.

Here is (bit boring) example of what happens now, maybe nice to see for you as an author.

Some insights in when the error appears.

When I just save the GH file after running Octopus, then close the document and reopen it, Grasshopper finds a recovery file. But the stored data is still there in Octopus after ignoring the recovery file.

If I save, close the document, and unload Grasshopper. Then restart Grasshopper and open the file, the error occurs.

More problematic: after this happened also the GH files with Octopus who could be opened normally previously (without the data) now produce the error (Octopus component gone from canvas).

Restarting Rhino entirely and opening the initial file with Octopus in it (without generated data) seems OK, the same goes for the GA files with the by Octopus generated data.

For some reason, even without running Octopus, but by only opening the Octopus GUI, and closing the file again. The next time I open the file Grasshopper finds a recovery file (while nothing should have changed).

The only option to (sometimes?) preserve the data from Octopus between file opens is to:

1) Run Octopus

2) Save the GH file

3) Close Rhino entirely (GH unload doesn't seem to be enough)

4) Reopen the file with Octopus data

For now it just looks like, I shouldn't use GrasshopperUnloadPlugin, even after a file save. But I thought to have seen the error previously even after closing Rhino entirely.

Some other findings (maybe by design):

Opening the GH file without my Rhino file open with referenced content. Octopus tells me there are changed objectives and Octopus doesn't show previously found solutions.

After opening the Rhino file with the referenced geometry Octopus shows the solutions, but still gives me a warning about changed objectives.

When I close the GH file, open my Rhino file, then open the GH file with octopus again, there is no such warning.

I hope this gives some insight, although not really technical and maybe a bit confusing.

Please let me know if I can be of any further assistance.

RSS