For a while now I have been somewhat obsessed with the notion of building a parametric geodesic dome in Revit. I was originally inspired by David Light’s winning DesignByMany submission for Bucky’s Dymaxion House (which, by its geometry, can take advantage of standard divided surfaces. I quickly found that geodesic domes cannot be built using stock divided surfaces.) Also, I just love geodesic domes. I visited Bucky’s own Biosphere in Montreal this summer and committed myself to figuring out a way to build one in Vasari without writing any code.

The following video shows my technique, including an aspect of 3D snapping to surfaces that was pleasant surprise to me (and Zach as well!)

Coda: After building this model, I came across George Mokhtar and Iwan Peverett at BimAcadamy  who took a very similar approach. Interestingly, they generated their icosahedron as a pseudo solid using formulae. This seems a bit more efficient than my technique of stitching the intermediate components across the golden rectangles.

Another Coda: And just last week the BIM Troublemaker made some trouble with a stellated dodecahedron in Revit (sweet!) Yet another (probably more efficient) way to explore building this dome.

_tom    

As journals used VBscript we've successfully modified them to read from text files to automate sheet creation, model geometry, and upgrade families.

In this post I'm going to share an idea to use a journal to visualize command usage and sequence - something of great interest to a software designer as it could uncover patterns of usage and common workflows.  I'll mention that Autodesk does collect command frequency usage via our CIP program but this can only tell you how often a command is used.  More meaningful usage data remains ellusive but not in journals.

Process a Journal File

To get the data I need I just want a list of all commands in the order they occur and nothing else.  A journal file contains this but its lost in the noise. Here is a line that contains a command:

Jrn.Command "Internal" , "Create a default 3D orthographic view. , ID_VIEW_DEFAULT_3DVIEW"

What I want is the last command ID or description - in this case the creation of a default 3D view.

To condense a journal file a there is a great text processing tool in Perl.  I downloaded the free version of Activeperl and here is a good perl reference.

I created "process_journal.pl" which contains the following code:

open(OUT,">JournalProcessed.txt") or die "Can't open $filename to write";

open(INPUT, "journal.txt") or die "Can't open journal.txt";

while (<INPUT>)

{

  $this_line = $_;
  if (/Jrn.Command\s*".+?"\s*,\s*"(.+?)\s*,\s*(\w+)"/) {

  # $1 is command description, $2 is ID

  print OUT $2 . " " ;

  }

}

close INPUT;

In pseudo code:

• open journal.txt
• read a line
• look for the lines that have "Jrn.Command" and extract the description and command ID storing these in two variables $1 and $2
• Print out the variable $2(Command ID) to a new file JournalProcessed.txt and add a trailing space
• loop through  journal.txt

The output of JounralProcessed.txt looks like this:

ID_STARTUP_PAGE ID_FILE_NEW_WITH_TEMPLATE1 ID_FAMILY_NEW ID_REVIT_FILE_CLOSE...

It's a long string of command IDs separated by spaces.

Visualization

With a little searching I found someone has done the heavy lifting for me on the IBM Manyeyes website.  On this site you can upload tables and choose many pre-baked visualizations.  In my case I didn't even need a table!  Their "Phrase net" visualization only requires only a block of text.  The Phrasenet renders words with the font size relative to the word frequency and connects these using directional arrows that are also weighted by the relative number of connections.  A lot of data is encoded using this simple method.

Real World Test

So now I have a tool that can process a journal and another that can help me look at it.  The next question is "What to look at?"  Every journal file is different depending on the task at hand.  I could look at journal files from different disciplines or compare files from different phases in a project (creation vs. editing)

For this post I'll process a journal donated by co-conspirator Zach over at Buildz.  He gave me the journal file used to make TROGDOR the Burninator.  Here is trogdor in all his incendiary glory:

Here is a snippet of the processed journal:

ID_REVIT_FILE_NEW ID_FAMILY_NEW ID_OBJECT_3D_CURVE ID_ZOOM_IN ID_BUTTON_SELECT ID_OBJECT_POINT ID_OBJECTS_CURVE_CIRCLE ID_SKETCH_PLANE_TOOL ID_BUTTON_SELECT ID_EDIT_SCALE ID_BUTTON_DELETE ID_EDIT_SCALE ID_EDIT_SCALE ID_BUTTON_SELECT ID_SKETCH_PLANE_TOOL ID_BUTTON_SELECT ID_SKETCH_PLANE_TOOL ID_BUTTON_SELECT ID_ZOOM_IN ID_OBJECTS_CURVE_LINE ID_BUTTON_SELECT ID_OBJECT_POINT ID_BUTTON_SELECT ID_SKETCH_PLANE_TOOL ID_OBJECTS_CURVE_CIRCLE ID_BUTTON_SELECT ID_EDIT_UNDO ID_EDIT_UNDO ID_EDIT_UNDO ID_EDIT_UNDO ID_IMAGE_WIREFRAME ID_BUTTON_SELECT ID_BUTTON_DELETE ID_BUTTON_DELETE ID_BUTTON_DELETE...

and visualized in ManyEyes:

We can see heavy use of Select, ESC, Points, Sketch Plane, and Zoom.  Points are by far selected the most and the sketch plan tool has a high affinity with drawing lines (id_objects_curve_(circle))

The visualization allows me to select nodes and move them around.  Here is a different arrangement:

And a link to the live visualization so you can play yourself:

Happy visualizing!

_erik

We have compiled a short survey for Revit Server 2011 & Revit Server 2012.   If you have used either version or perhaps plan to implement within the next 6 months, your feedback is highly important.   Even if you evaluated Revit Server but decided not to implement, that particular feedback is also helpful.  I know your time is valuable, so I very much appreciate your participation and thoughts concerning Revit Server.

Your feedback helps us create better solutions for you.   Please feel free to forward this link to any team-member or consultant that may have also implemented Revit Server.  We are looking for general end-user feedback as well as IT or BIM Manager support feedback.

https://www.surveymonkey.com/s/RevitServerSatisfaction

As always, a huge THANK YOU!!    The survey will remain open for 2 weeks, and close on February 24^th.   But , why wait?  Have fun today!

Richard

Underlay

Underlay has always been slightly out of alignment with user mental models.  In 2D processes underlay had one drawing placed under the other yet in Revit we are looking deeper into the model at parallel levels - not placing one view under the other.  Further there is the additional behavior that underlays are open for selection which is also often unexpected.

This can come in handy when using Linework to reference items above yet can also cause accidental changes do to inadvertant selection.  You may note that the system currently prevents box selection from working on anything visible via the underlay and this was implemented after receiving user feedback.

My question to everyone is when is selection of an underlay useful?  MEP workflows or reading property values? What if underlays were never selectable?  Is the work around OK?

Another potential mitigation strategy would be to allow underlays to be selectable in specific tools, such as the Component or Linework tool, yet disallow selection for the Modify (Arrow) tool.

I'm very curious to hear your thoughts.

Links

Much like underlay when is the selectability of a Link a problem and when is it advantageous?  Specific examples can help us acheive the best default behavior and/or allow flexibity if required.

_Erik

Microsoft Vista support will be discontinued for new 2012 Autodesk products (Revit included)

Windows XP & 7 are still supported.  See the following link for more information: http://autode.sk/xY3PHO

In this post I want to briefly talk about Chain Select  This is the behavior that allows you to select a contiguous set of connected lines.  Revit does not have a polyline element so this is one way to speed multi-segment selection.

In the Massing environment the preference is to always select the chain or loop.  Tab will select individual sub elements of the chain.  This is 180 degrees opposite from the project environment where individual elements are given preference and you must tab to select the chain.

Each behavior has its benefits yet having two opposing behaviors can lead to frustration and slow learning.  Is this the case in the field?  Do you prefer one behavior over the other? Why?  Do we need a setting to control this?  can one behavior serve both environments?

I look forward to the discussion.

_Erik

A product like Revit that is moving from adolescence to early maturity needs to revisit its roots periodically to make sure that it is not coasting on early glory. To illustrate, let me turn on the time machine.

The scene: An architect’s office late 1999.

The action: I am demonstrating a Beta build of Revit 1.0. On screen we see a floor plan, an elevation, and a window schedule. In the floor plan, I add a window. It appears immediately in the elevation and a new line appears on the schedule listing an instance of Window Type 2. I delete another window from the elevation and it disappears from the schedule and plan.

The reaction: “That’s amazing, I love it. Checking window schedules is so tedious. Now how do I put the window type symbol in the schedule? No? I’m going to hate that. What about adding comments or shading every other line? That can’t be hard. I’m sure you’ll fix it soon.”

Now: It’s 2012. Revit schedules can do much more than they did in 1999. You can add new properties, perform some calculations, and group header cells together. Unfortunately they don’t provide full control over cell format, won’t schedule a generic model component, or display a window tag. “You’re killing me” I hear.

Schedules and related elements such as Legends, Takeoffs, Lists, or Reports are powerful and highly useful features of Revit. They do, however, have a number of limitations both in the data that can be presented and how it is presented. Some are the result of constraints imposed by current interface tools, others reflect deeper issues in the Revit data structure and regeneration engines. To build a better picture of how the tools should be working, the AEC User Experience team is studying schedules and related non-graphical data display tools.

As one part of this research effort, I am gathering information on how you use or want to use schedule and schedule like methods in your work. If you are interested in helping us in this, I have provided a link to a short survey focused on your usage and a request for you to share image samples of what you are doing.  Samples of real work are an important part of our research. They will be used as test cases and help us illustrate themes as we explore methods to increase flexibility and capability. They need not be confined to what you can currently produce in Revit. Examples of what does not work well or appear correctly in Revit are more useful than what worked well.

https://www.surveymonkey.com/s/66BVMW3

_David Conant

This control defines what happens when you click on an element and drag while holding the button down. It defaults to being enabled.

Behavior

• When enabled the element under the mouse down action will be selected and then moved as you drag.
• When not enabled the mouse-down/drag/mouse up will select elements only and will not move any selected elements in the same action.

I'm interested in how many people use this setting.  I know one may disable it to try and pick elements that are located over another larger element such as an imported image or import.  The aim is to avoid picking and dragging the background element while making a window selection of a smaller group of elements.

Lets imagine this control is un-checked and removed from Revit.

• You can still select elements then drag them -  just not in the same action.
• Accidental drags are prevented which is good as this can often cause long regen times and slow tasks

What are the downsides?  Please share any you see.  Is drag/move even useful given the lack of accuracy?

FYI, I am looking into this as part of research into face selection.  It is a well known issue that having to pick elements by their edges and tabbing to do so is a usability and learning issue.  Allowing face selection, however,  will increase the potential to accidentally drag elements on selection.

BTW to preview face selection it currently works on the Part category.

_erik

For this specific effort I desire to better understand current usage and context. To this end I'm on the search for screen shots of the following:

• Browser Organization "Folder" and "Filter" tabs for Sheets and Views. (4 images)
• Project Browser Screen shots showing all the nodes expanded.  I realize this may be many as long lists are a known issue in the current design.
• MEP Users feel free to send an image of your System Browser

Anything sent will be for internal use only.  If you use a third party product please mention it as I'm looking at products offered by our partners as well.

From this information I hope to get a better picture of how data is organized (by discipline, project parameters ect..) and also what constitutes a realistic data set (size and diversity) for use in testing new concepts and ideas.

If you can help please send images in any format via email.

I'll follow-up in the weeks ahead with more related topics to help validate known issues, common tasks and requirements.

_erik

1. Go to the Dynamo repository on github.
2. Click on DynamoInstall or DynamoInstall_32bit (depending on your Revit install)
3. Click on Release
4. Click on both the msi file and setup.exe. For each, click “View Raw” and that will allow you to download the file
5. Run setup.exe and restart Revit. I will post instructions soon on getting it going in Vasari.

Ian clearly admits: It’s incredibly buggy, feature incomplete, and completely undocumented. Regardless, this little science project is really cool and has generated a lot of interest internally as well as at Autodesk University last month. If you were were not lucky enough to attend Matt Jezyk and Zach Kron’s “Energizing Supermodels” course at AU, stay tuned. We will soon share some more cool stuff you can do with Dynamo. In the meantime, I put together a short video to introduce the add-in and show how to get started.

_tom]]> Design Prototype Tips & Tricks Video Tom Vollaro Tue, 20 Dec 2011 14:02:16 -0500 http://insidethefactory.typepad.com/my_weblog/2011/12/revit-spaghetti.html