Friday, April 24, 2009

Dear Autodesk

Let me try to explain the difference in the simplest terms. This is not about the discomfort of learning a "new" UI, but about learning one that is demonstratively and functionally inferior. It might test well in a well lit, refreshment stocked room for an hour or two but beyond that controlled environment, it is an extraordinary and inexplicable step in the wrong direction.

Your present missteps are gutting the value proposition of Revit by not only failing to develop long requested functionality, but by breaking the functionality that is there by forcing the user to jump through a maze of hidden, contextualized and subjectively displayed GUIs to get to it. Not just occasionally (as the users have become accustomed to) but every single time and with every single mouse click.

The sales manager that came to my office last week basically said, "The ribbon is here...get over it." Any other business (restaurant, bookstore, coffee firm...) that said this to its customers would lose them at the first opportunity.

So if you're going to make create a cross-product GUI - go for it! But make it wonderful and amazing and a joy to work with for 10 hours a day. Don't make it new. Make it better. Ultimately, it should be a pleasure to learn a new UI because of the functionality to which it elegantly provides access. Not as a burden with diminished returns. The cross-platform GUI that you eventually create should reflect a rigor, care and attention to detail that is expected of those that will use it to create their efforts. It should inspire.

Tuesday, April 21, 2009

Humpty Dumpty Sat On a A Wall...

The ribbon is fundamentally flawed for complex design environments. It incorrectly co-joins nouns (the thing being built) with the verbs (the action that you can preform on the thing being built). Every action is contextualized and the result is a confusing mental map of icons that subjectively expose or reveal themselves with each decision. What's it like? Try limiting your vision through one eye while staring down a paper towel roll. Now try to walk around. You get the idea.

I've received a number of phone calls, emails and instant messages from past associates inside the factory telling me the ribbon is half-baked and they're frustrated that they weren't able to think freely with regard to creating an elegant, cross-product GUI. Instead - the ribbon was imposed as a corporate mandate. While they're not able to say anything publicly, what is being said publicly ("Here's why the ribbon is better...") is a complete contradiction to what is said privately ("Look - you know we hate this, but it was forced upon us...").

Consider this: if the ribbon is a great metaphor for design rich environments, what did Microsoft (who developed the ribbon) come up with for the Expression and Silverlight suite? Not a "ribbon" in sight:


Sloppy seconds anyone? Looks like Microsoft isn't willing to eat their own dogfood - which kinda says something. Autodesk must expose the existing the UI in Revit 2010 and then do their own rigorous homework with regard to a cross product look and feel. Poor design aside, I consider it inexcusable that no business strategy was put into place that allowed companies time to adjust other than "upgrade and get over it."

And now ADSK's response? "Well, the ribbon is here to stay - so help us make it better." I've got even a better idea - stop disrupting your customers. The corporate vanity that believed a "common look and feel" across design applications was more important that what the users still can't do with those same applications is deeply irresponsible considering that no customer, no user group, no user forum, no wish list has put the redesign of the Revit GUI high on their request list.

The result? It'll take more years and more customer and shareholder millions to "fix" what is now more broken than what it was meant to improve.

What an unfortunate and enormous waste of customer and shareholder value and good will. So many millions, so many years, so much human effort has been spent making something "new" and yet not making it better.

Saturday, April 4, 2009


Had a long and much needed discussion with Matt Jezyk today (Saturday). Matt is the Product Design Manager in charge of the Massing in functionality in Revit 2010. And in spite of the earlier post, he's someone that I have a great deal of respect for since we toiled together at pre-acquisition Revit Technology.

Hopefully we can sit down next week via teleconference and discuss process and direction (not to mention any errors and omissions on my part). Even though I repeatedly tried to call him, he freely admits he refused to answer any of my calls. In part because as it turns out I was calling his office phone. And in part because he was in San Fransisco. My bad.

Without breaking confidences, I think he had some fine points:

1) Building Masses. There's some good reason to approach massing at the building scale different than geometry creation at the component scale. And because until hardware and priorities catch up with each other it's better that users not have the ability to carefully craft sink taps as multi-profile blends in Revit. I somewhat disagree. I'd rather users have the ability use a tool appropriately at the risk of them doing something stupid. But I understand his point. And it reminds me of the time a project refused to print or export due to the modeled vegetable sprayers (complete with holes where the water sprays out).

2) Present Geometry Toolset. There's likely some fundamental limitations, which required a significantly different approach in 2010 - and not just as a compliment to present processes. It was necessary to start over. His contention is that explicit Extrusions, Sweeps, Revolutions, etc. aren't necessary for creating, iterating and resolving form. And he reasoned that in the majority of use cases the new Massing approach can do the kind of iteration that I'm ranting about. And I'm hopeful that he's right.

3) Present Limitations in 2010 Massing. Yes - there are some. But this is a first pass in a succession of passes. Which brought up my rant about why more should and could have been done had time and resources not been focused on the UI "enhancements". To which he opined the UI changes are necessary. To which I emphatically offered that even if the new UI was "perfect" - there should have been a transition strategy in place that didn't disrupt the customers business.

My point was that users need an opportunity to get their heads around a new UI (or the New New UI which will certainly follow in 2011, just before the Newest Newer UI which will come out a year or so after that). Existing users need to be able to absorb a new UI in non-stressful, day to day activities. And they should have the option to switch back to the existing and familiar UI environment on a project if the deadline is today (or yesterday).

Perhaps I'm having a change of heart. Perhaps the new UI could actually be a great environment for the new Revit user that's starting to use a BIM tool for the first time. You see - the big, fat, brightly oversized icons kind of remind me of those thick starter crayons that are flat on one side. I suppose it is entirely possible that some kids just can't be trusted with the additional burden of learning to color with completely round and smallish crayons. So by extension, the new UI could help new users get up to speed before they're ready to move on to a persistent, minimal and adult UI that doesn't jump around with each button push or insist on entertaining (if not insulting) you.

Think of it as a kind of Microsoft Bob, you know? Back in the old'n days before the internet, most people weren't ready for prime time computing. And everybody loved it. So what we need is a BIM GUI for designers that aren't ready to make a building in a scary 'ol computer. Buttons the size of your thumb will surely help those less decisive:

"Oh sh!t...which button do I push?! I'm still not quite sure...."

Therefore, by the powers vested in me by this scotch and soda to the right of my keyboard, I hereby christen the new Revit 2010 ribbon GUI Autodesk Bob(TM). I can hardly wait to see "Hammy the BIM Hammer(TM)". Hammy will "swing into action(TM)" every time you do a search, or hover too long or push the F1 button (right next to the Ecsape key...perfect).

As for the discussion with Matt? Overall - we both had some valid points. So I'm looking forward to the Show And Tell later next week. And if I have to eat some crow? Well, at least Matt's the kind of guy who'd buy the beer to wash it down.

Friday, April 3, 2009

Autodesk Bob

Massing was supposed to be the silver lining to the cloud. With the new Massing tools, you were supposed to finally able to create forms that were impossible in Revit 2009.

Verdict? It's true: the old massing tools and flow are gone.

Wait for it....unfortunately. HD YouTube version here.

Yes - you can create some of what the old massing tools couldn't, but at the same time - you can't create all the forms that it could. And furthermore, you can't deeply iterate the results. So there's new stuff - but not all of the old stuff. To put it another way - it's a great 'answer' for someone that doesn't yet understand the 'question'.

What!? Let me explain. When you create form in in 2009 Massing and the Family Editor - the sketches and profiles to create those forms are not lost. You can still go back into "Edit" mode (the same still holds true in the 2010 Family Editor). But you can't go back and edit the sketch that was in Massing 2010 - only the sketch that is.

You can certainly manipulate (rotate, stretch, push, pull) what you've created. You can even add additional points, edges and profiles (with some exception). But once you finish a sketch - that's the starting point. No more sketch mode. The results? In the image below - the blended form "A" can be modified via sketch into blended form "B". Note how fillet arcs have been added in the process. This can be done in 2009 Massing.

This process of reiteration is lost in the 2010 Massing Tool. The closest you can get is to push vertices to get from "1" into "2". You can push/pull vertices and faces - but you can't turn a point into a curve. I can't imagine the reasoning behind this limitation. And in order to allow for the most flexibility, there's heavy use of splines (for both paths and profiles). Splines will certainly allow for the creation of more morphic forms.

Yet in the strangest of decisions - if you import a 2009 project which contains masses, these forms may be manipulated via the present rules in 2010. You may still edit the paths and profiles/sketches. AND you can even create new forms using the existing metaphors using the present rules. The engine is still under the hood! But inexplicably - this approach and functionality is not exposed if you're creating new Masses in 2010.

In Revit 2009, all of the forms in the above image may be derived from any of the other forms. This flexibility gives a designer the wonderful ability to intuitively and quickly move between creation, iteration, rationalization and eventually resolution. Inexplicably, this design flow has been removed from Massing in Revit 2010.

In Revit 2010:
  • If you start with a circle, but later discover that you meant ellipse? Start over.
  • If you start with a square but meant a circle? Start over.
  • If you start with a line, but meant a curve or spline? Start over.
  • If you start with an acute angle, but meant a radial fillet? Start over.
Digital clay? Digital concrete.

What seems incredulous is that this geometric toolset wasn't designed to compliment the present approach to Massing - but to replace it. And not unlike the approach to the proposed UI changes, the customer is again faced with a decision that is more "either/or" rather than "both/and". Yet another imposition without a transition. And ironically, the new method of form creation is not available in the 2010 Family Editor. Does Autodesk really believe it necessary that Revit users should have one approach for creating form when designing the whole building, but another approach for designing the pieces and parts that go in them?

Creating 3D forms in the computer used to be a significant task. But now that task is rather trivial. Many elegant and inexpensive (if not free) applications can create complex geometric shapes. The deeper challenge is within the ability to further rationalize those forms into surfaces, shapes, etc. that can leverage cost effective production and assemblage methods.

It's really wonderful that Revit users will finally be able to transition spline based profiles across spline based paths from within Revit. But have you ever tried to dimension a spline? Ever try to communicate a spline in order to resolve construction? It's essential that the rationalization be a compliment to the iteration. Why? Because great design is more than an exploration of geometric possibility - but of geometric plausibility.

Autodesk needs to take particular care in two areas:

First: Autodesk should stop exciting customers with a five-minute demos of a massing tool that on further investigation actually turns out to be quite limiting. By doing so they run the risk of building resentment when they could have otherwise built a trusting business relationship.

Second: Revit is by far the best BIM tool available. The integration of Building, Content, Documentation and Multi-User / Multi-Discipline work flow are well complimented with being highly implementable. But so long as there remains little competition, it seems that Autodesk has begun settling for better rather than striving for great. For example, it apparently takes Autodesk two years to bring new functionality to market even when the customers aren't even asking for it (cough...ribbon...ahem). But it can take even longer to develop the functionality customers really do care about ( tools...sputter). If this continues, ever so slowly a vacuum will begin to form.

Before the internet-thingy, a well-heeled technology company could control marketing, distribution and FUD through a global channel of distributors and resellers. Only a decade ago, Information Age companies enjoyed being complimented by Industrial Age business practices (dutifully shipping actual boxes - think about it!).

Now? Someone can read a Twitter about a new application, download the demo the same day, purchase a copy the next and blog about it by the end of the week. No distributors or resellers. Immediate access. Training? Podcasts.

Think about it: two years. Two whole years. Add to the mix a small, highly motivated, enthusiastic team that doesn't know when to take "no" for an answer. What could possibly happen in two years? It's not like it's happened before, right?

Think about it.