Running into trouble with B* input

Hey all -

I am having some problems with the boolean that is supposed to trigger the end of a loop. Even when the boolean turns false, it is still read into the component as True, as you can see in the screenshot.

Because it's not properly registering the point at which to halt, it continues to loop and fails. This happens to me in both versions 0.5.1 and 0.5.5.

Any idea why this might be happening or what I can do to work around it?

Replies to This Discussion

Permalink Reply by Pieter Segeren on May 31, 2011 at 3:41am

I had the same thing happening, but don't know a workaround. Thought I would let Yiannis and you know.

Attachments:

trueisfalse.png, 35 KB

Permalink Reply by Yannis on May 31, 2011 at 6:25am

Hi Andrew,

Does the same happen if you connect it directly to the component instead of the parameter?

I currently cannot reproduce the behavior myself, would it be possible to share the definition?

EDIT: Just saw Pieter's answer too...so I guess it does happen with components too..

If any of you guys could share even the part of the definition where the problem appears would be of help.

Thanks

Permalink Reply by Andrew Heumann on June 5, 2011 at 12:43pm

Hi Yiannis -

Here is the file where I was running in to trouble. You can see that the loop almost always fails, and when stepping through, there comes a point where the boolean turns to false but the hoopsnake component reads it as true.

Attachments:

Bstar_Fail_ex.3dm, 62 KB
Sort_And_Unfold.gh, 249 KB

Permalink Reply by Yannis on June 6, 2011 at 6:19am

Andrew,

Can you try with the file I attached and let me know if it's ok?

Oh, also try disabling flattening the B* input.

Thanks.

Attachments:

HoopSnake.gha, 31 KB

Permalink Reply by Andrew Heumann on June 6, 2011 at 2:01pm

Flattening didn't do anything, and the gha you posted fails to load in my Rhino -

"Could not load file or assembly 'file:///C:\Users\AndHeum\AppData\Roaming\Grasshopper\Libraries\HoopSnake.gha' or one of its dependencies. Operation is not supported. (Exception from HRESULT: 0x80131515)"

Permalink Reply by Fred Becquelin on June 6, 2011 at 2:22pm

Hi,

I had the same problem.You can see in this video, the tolerance is set to 1.5mm but it always go one step further.

I wonder if it could be some timing issue, because my solution needs about 2s to complete but the boolean trigger is evaluated near the beggining.

Another thing, sometimes it looks like data is not cleared properly, and loop stops after one step.

Hoopsnake is great, ty so much! Now I need you to figure a way to nest loops!

Permalink Reply by Andrew Heumann on June 6, 2011 at 2:38pm

While we're making wishes, it would be great if there were a way to trigger a loop or a step from an external component like a boolean toggle, or cause it to automatically re-run the loop when the input data is updated.

Permalink Reply by Yannis on June 6, 2011 at 3:09pm

Hi guys,

The cause of the 'one-step-too-far' issue is located and the file I posted supposedly would correct it.

Andrew, I re-downloaded the file (in case sth was wrong) and re-tested and it looks fine (see screenshot). Could it be that you are using another version than 8.0010? This is the only version that I test with. Can you open any of the Hoopsnake examples using this file?

In any case, I am re-uploading a completely fresh binary, tested with example files as well as Andrew's file.

Fred, the new file should fix the issue. As for the clearing issue I recently experienced it also and I am currently investigating it.

Please just note that the input will still mention 'True' even though the internal state of the component is false. In other words, visually the 'one-step-too-far' issue still exists.

Nesting/triggering are features that I would love to see implemented too, but they are still in the design phase, since the implementation should be done carefully in order no to mess with Grasshopper's timing.

Attachments: