Fix My *ADVANCED* Dynamic Block

[DanielEvalle]

So, I've put far more hours into this than I should, and am curious if someone else can come up with a more elegant solution and/or critique what I have done.  The attached file is revision 6, and I can share the other methods I've attempted, but 

The aim is pretty straightforward,

• Make a 3"x3" detail 'frame' with a 'leader' attached to a 1"x1" indicator 'frame'.

1. The 3"x3" detail frame should be static with a basepoint at the bottom left corner and attribute text at (0,-.0625")
2. The 1"x1" indicator frame should MOVE to any position

• Leader-line req:

1. The leader line 'head' should STRETCH with the indicator frame.
2. The leader line 'head' should incrementally adjust by .125"
3. The leader line 'head' should always be attached to any side of the 1"x1" indicator frame.
4. The leader 'landing' should adjust by .125" increments (INCLUDING corners)
5. The leader 'landing' should always be attached to the Left, Top, or Right side of the 3"x3" detail frame

As it is, for about 85% of the time this block works well.  BUT.  You will notice when using, if you stretch the leader landing position to the Top of the 3"x3" detail frame, selecting the flip action will wildly throw the "snapRot" (and all of its proxy movers/stretchers) out of position.  I understand that my "Proxy" movers are, ultimately, at the heart of my problem, but have been unable to come up with a better solution.

IDEALLY, there would be NO flip action and you should be able to incrementally adjust the leader landing to ANY side of the detail frame by just moving the 1 (polar parameter) grip there.  But again, I have not been able to find an elegant solution to this.  I also aim to have other versions of this block that allow the Detail and Indicator frames to have adjustable sizes...  But I feel like I'd like to nail down the "flip" issue first.

Thanks in advance for feedback.

[will_young]

Will this be used in paper space or model space?

The reason I ask is that if you are using in model space you could apply a wipeout to both boxes and then the line between the two can have its start point and end point at the boxes' geometric centres as per the attached. As long as you send the block to the back, the wipeout can obscure the line within the box and the model geometry can sit on top.

In paper space however, the viewport always gets obscured by the wipeout which would render my suggestion useless.

In that case the only thing I would suggest other than what you have done is that it might be less clicks to not include the line in the block and add it manually after inserting the block.

Also is it an option to get rid of the line altogether and have an attribute that displays under each box eg. '01' to match the detail to the callout?

[DanielEvalle]

It will be used in paper space so unfortunately the problem persists.

My goal is to improve my department's application of standard objects like these.  So while it is admittedly faster/more straightforward to just draw the two rectangles and a line, I am looking for more elegant solutions that remove guess work while still allowing for variability.  For example, even if your wipeout option worked in paper-space, it doesn't account for when a line (or any object) overlaps the connector line in a way that makes it less clear where the smaller rectangle is indicating.  Our goal is to make it blatantly obvious where/how our details apply.

[will_young]

I'm not suggesting drawing the rectangles, just the line between. I know it's not as neat as having it all in a block.

The other simpler way is to use polar parameter with move actions and have point snap turned on as attached?

Or is it essential for the increments on the indicator frame as well?

[DanielEvalle]

I want to make this thing fool proof.  If I wanted to use a polar parameter without bounds... does the complexity of my original block not demonstrate that I easily could have handled something like that? 🙃  Also, I think it'd be more elegant to just add vertices to my p-line (w/ colinear geometric constraints or something) instead of introducing 25 ungoverned point objects...?

The attachment to the indicator frame needs to be able to shift to any side regardless of the side that the line is attached to on the detail frame. 

Like so:

The actual increments aren't thaaat important, but again in keeping with making a block that maintains 'standard' usage, I find it helps to make things feel a little more uniform.

[DanielEvalle · Accepted solution]

Figured it out... Used a double lookup instead of my super complex rotation/stretch/move proxys.

[Brock_Olly · Accepted solution]

I know you've already completed it but I still wanted to give it a shot since it seemed like a nice block to learn double lookups for me.

[DanielEvalle]

Dude heck yeah.  Honestly love it much more than mine.

[DanielEvalle]

@Brock_Olly EXTRA CHALLENGE MODE:

I did have in mind a version where you could resize both the Detail Frame and the Indicator Frame to a certain degree. (i.e. +-1" in any direction)

See attached for example (and sorry for the clunkiness... I'm new to double lookups and am still not sure that I like them.)

[Brock_Olly]

I think the easiest way to do this is to add 4 stretch actions in each direction, you can limit them with increments/lists if you want.
edit: Should have used XY parameter instead

[DanielEvalle]

Maybe I'm tripping but do you ever chain actions?  I'm impressed at how much use you've gotten out of the blocks you make without chaining.  I think my block making philosophy has really suffered from not understanding or utilizing double/multiple lookups.

edit:  ALSO I just picked this up... you can use RANGES for number inputs in tables?????  WHAT SORT OF BLACK MAGIC DID THIS COME FROM?  WHY DID I NOT KNOW THIS? 
THIS IS SIMPLY REVOLUTIONARY!!!!!!!!!

But fr. does this work on all numerical inputs or only for angles?

[Brock_Olly]

Hi,

I only use chained parameters when necessary, in this case the block works so it's not needed?
Maybe there's a more efficient way with chained parameters but this is how my brain came up with a solution.

As for the ranges, see attached DWG 🙂

[DanielEvalle]

So everything I've learned about Dynamic Blocks is self-taught through trial and error (and Forum help! 😄 ) Where did you learn about range inputs? 

I took a course through NYIAD on AutoCAD to get my "AutoCAD User Certificate" and they only sparsely touched on blocks / block creation, much less dynamic blocks.  Is this actual confirmation that I should go to something like AU University?  It seems like so much of the content for dynamic blocks is so specific that going to a general "AutoCAD" convention or whatever you want to call it would only barely scrape the surface of the what and how that you can do with dynamic blocks.

[Brock_Olly]

Thanks, I actually don't think I'm that good, also still learning.

I read these forums a lot and have learned from their examples. The ranges I saw in a post by Ojuris whom I think doesn't visit anymore. Libbya seems to know everything about dynamic blocks and there are plenty more experienced and knowledgeable people here that have posted some amazing blocks over the many years!

Simply google dynamic blocks + your question and you're bound to find something that can put you on your way.