Curved Monolithic Stairs – Am I Doing Something Wrong?

One of my Revit Architecture users hit a snag the other day when making a curved set of stairs to get up to a stage in an elementary school.  They were going to be cast-in-place, so we opted for monolithic.

Now, I have never been a huge stair fan in Revit.  Most of the time, in fact, I hate them (I despise their close relative, the railing, who has become my nemesis).  So when something started going wonky, I wasn’t surprised.

Maybe we’re doing something wrong.  Maybe we have a parameter set incorrectly.  But let me tell you, I tried it on multiple projects, with multiple combinations of parameters, and different materials, and heights, etc etc.  Always had the same bizarre situation:

In plan, everything looked nice and played well together.  So we thought we were good. 

Looks great!

A section cut is where something told us things were angry.  Where are the stairs?!

Something is amiss

When in doubt, go to 3D.  A quick camera dropped in, or switching to your 3D view can answer many questions that other views cannot.  That’s why we model it!  Unfortunately, this was not the case.  Going to 3D just was more bizarre.

Where are my stairs?!

The only thing we can figure is that Revit is refusing to let the “section” of a monolithic stair go past its calculated point, so it just slices it.  What?  And this is only for stairs whose type is checked as “monolithic”.  Non-monolithic curve fine.

Non-mono looks goooood…

This is crazy.  We opted for an in-place family sweep, but even that doesn’t make me happy, because you cannot use the STAIR category for in-place families.  Another bizarre and severe limitation.

If you have seen this issue, and you know how to work past it, feel free to post a comment and let everyone know.  Otherwise, just be aware of the insanity, and cross your fingers that stairs (and their jerk cousin, railings) get an overhaul soon.

Gridline Gridlock

A question I find myself answering a fair amount is “Who put the cat in there?!”; more Revit related, I get asked “what the heck does that little ‘3D’ by my grid bubble mean?” a lot.

What the heck IS that?

The “3D/2D” button by gridline heads is often a source of confusion, partially, I think, because of the nomenclature they used for the system.  It’s a good system, but really has nothing to do with the traditional definition of 3D and 2D.  Semantics has often been an issue in Revit (looking at you, “Save to Central”) that often the team does a good job of fixing (looking at you, “Synchronize with Central”).  So far, the 3D/2D toggle is not one of them.

So, if you click a gridline, you will see a tiny “3D” up by the grid bubble.  The terminology is flawed.  Essentially, what the “3D” means is that this grid bubble position is shared throughout the project, and if you move it, you will move it in other views as well, and wherever you put it, that’s where it will show up in any view where that grid bubble is marked “3D”, or just outside the annotation region, depending on which is shorter. 

When you click it, it changes to “2D”.  This basically means “View Specific”.  So, if you move the grid end now, it won’t affect any other views.  This view is it.  You will see a ghost grip where the “3D” of the gridlines is.  Dragging the head back to this will reset it to “3D”. 

Now, let’s say you click the 2D on for some of your grid bubbles and you move them and you really like the positioning in this view, so you want to copy it to other views.  That’s where the “Propagate Extents” button comes in to play. 

Select the gridlines you want to copy and click “Propagate Extents”.  You will be prompted with a list of parallel views.  Select the views you want to copy to and then the gridlines position settings will be copied onto those views.  It’s a one-time thing for the “2D” gridlines, so if you were to tweak the position after propagating, you would need to re-propagate to get those new settings pushed over.

Do I have a solution for the 3D/2D terminology fiasco?  Not really.  The only better icon I could think of would be a little chain to replace the current “3D” and a broken chain or chain with a slash through for “2D”… for “linked” and “unlinked”, get it?  Is it better?  Who knows.  Will it ever change?  Who knows that either.  But hopefully the above clears it up.

Help Your Project Manager – Before He Kills you

To our chagrin, we have discovered that our number one hurdle with transitioning to Revit is not the software learning curve, it’s not the user’s need to shift to a new workflow mindset, it’s not even convincing PICs or the board that it’s time to buy new licenses.  It’s the Project Managers.

You need to be sure you fully educate your non-production PMs about how a Revit project will go.  Best bet is to have someone on the team who has already gone through a Revit project.  Someone who is willing to gently, or not so gently, let the PM know what is possible, what isn’t possible and most importantly what makes sense.

Your typical PM of this nature simply knows that they usually get plans at this week of the project, elevations here, and then sections here.  With Revit, the workflow and document production is shuffled all around, and that’s a good thing!

We have had more than one Revit models that were on the edge of failure because a PM demanded X, Y, and Z at very specific times.  Why?  Because it’s what he always had gotten before.  He didn’t want to hear about what made sense.  The team was too new to Revit to argue, so they built the model to support X, Y, and Z.  Unforunately, in Revit, they should have been focusing on A, B, and C at that time.  So the model suffered, and the team spent more time picking up pieces later on simply to accomodate this PM’s backward demands.

Every firm has at least one PM like this.  There is no easy way to deal with him or her.  Just keep a close eye on the model and work hard to make sure that there are no drastic flaws in the workflow that can bite you down the road.

Winter Blues and Online Classes – AU Virtual

Wow!  It’s been very long since I posted something!  Do you care?  Possibly.  If you do, that’s very sweet of you.  Winter is always tough; I don’t like the sun setting right after lunch.  And to top it off, I didn’t even get to go Vegas this year for Autodesk University.  Luckily, Autodesk brought the best parts of AU to me!  (Note: well, maybe not the BEST parts, but some pretty good parts)

According to this article, attendance at AU was down about 30% this year.  That’s physical-in-person-I-can-touch-you attendance.  Now stop touching me.  This was also the year that Autodesk introduced AU Virtual, which (according to the same article) was “attended” by about 20,000 folks.  First off, this number is hard to quantify and verify.  However, I’m not gonna argue with it, because I bet it was a LOT of folks who watched some AU classes online.

My experience with online classes has been hit or miss over the years, mostly miss.  The content is sometimes questionable (no, not inappropriate, just not what I was hoping for – get your mind out of the gutter.  And I told you to stop touching me!) the presenters are not always engaging, but the worst part is the format.  You are stuck staring at someone’s screen, quite often a PowerPoint, and you hear them talk.  That’s it.  You stare at the screen for an hour listening to someone else go on and on and on.  Eyes glaze, minds wander, snoring ensues.

I haven’t done much research into how the human mind learns and how it assimilates data, but I bet staring at a screen with someone talking at you is low on the list of effective ways to learn something.  Luckily, the folks at Autodesk agreed with me.

A+ to the format of the AU Virtual classes.  You have a screen on the left showing a PowerPoint presentation or an application, and on the right is a human.  A real live human being.  Looking up at the camera and talking.  This tiny little human addition made the learning far more engaging than most other online classes I had seen.

Another issue with most online classes is, man, you REALLY want to show something to someone afterwards, but you can’t record the presentation!

A+ to letting folks download the ENTIRE screencast and “handouts”.  This is an absolutely remarkable feature.  If I could do a backflip, I would.

I won’t put an overall grade on the presenters or the specific classes.  I will say this – the large majority of the classes we watched were very much worth our time and the presenters were clearly knowledgeable.

The only drawback was the selection of classes.  I’ll give this a C.  I imagine this was being handled on a more tentative and experimental basis, but there were certainly many more “real” classes that I would have loved to have seen “virtually”.

Lastly, A+ to the price.  For those folks on subscription, the price was very very low.  I don’t want to use the word “free”, because that’s a big check I cut every year to Autodesk.  Having AU Virtual as complimentary for the subscription folks made the value of subscription much greater, however.  And frankly, I sometimes scratch my head as to what exactly why I am paying for subscription fees (aside from the license agreement), wondering what extra I am getting from this.  Bonuses like this make me scratch my head a little less.  It is my understanding that the cost wasn’t crazy for non-subscription folks.  I bet well worth it.

Obviously, there are features to the “in-person” conference that just cannot be experienced online.  I have attended twice now and I look forward to going again, even with the excellent option of seeing the classes online.  And no, not because it’s just Vegas.

So, a big kudos to the team who put together AU Virtual this year.  If you missed it this year, be sure to take advantage of it next year.  Hopefully, Autodesk will continue to expand the offerings and keep the excellent work they did for the first AU Virtual.  Hopefully, they will keep the same cost for subscription users, or have a VERY low cost to offset some of the expenses.

And hopefully next year, I’ll get to back to Vegas.

It’s All In The Details

There are many misconceptions out there about Revit.  Some of these are holdovers from early versions, some are just bad ideas that people latched onto some time ago that they haven’t let go of.  Like Crocs or calling chicken sandwiches “burgers”.  One of those “Revit Rumors” has been popping up on my radar – Twitter, user comments, etc.  The idea that Revit is bad at details.  I have seen several tweets along the lines of “I like Revit, but it is lousy at doing details” or even worse “Revit is great, but I still do my detailing in ACAD”.

I’m not sure where this misconception about Revit being bad or difficult at detailing started, but it’s just plain false.  And those that claim that it is easier in AutoCrap, I would actually argue the opposite.

Revit has cleaned up and simplified the process of detail work in the computer.  We no longer have to worry about layers.  Or even worse, color, and if that color is going to impact the lineweight (I know some of you are still plotting with CTB, you cannot hide!).  There is a selection of Detail Lines, and what I see is what’s going to print.  It’s like my old empty plastic orange juice concentrate can that held my Koh-I-Noor pens, a black sharpie, and some grey markers.  Easy.  There are also fill regions and a masking region.  And if you happen to use the wrong “pen” or fill, select it and change it.  There is an excellent selection of already created detail components.  Fill patterns and line patterns automatically with scale.

And that’s just the pure detail elements.  Don’t forget that you can constrain detail elements to model elements, so if you’ve detailed a joint on a wall with detail lines, you can constratin it so if that wall moves or changes its thickness, those detail lines will update as well.  On top of that, there is the oft overlooked but VERY powerful Edit Cut Profile tool which allows tweaking the edges of model elements in my view.  So I know and have confidence that cut patterns and fill patterns are consistent and proper.

It’s change and it’s different and I understand that.  Over the last two years as we have been doing our deployment in earnest, a major new part of my job has been therapist.  Change is scary, especially change on such magnitude.  People like the tools they are used to.

But that doesn’t make them good tools.

The people who think AutoCrap is easier to detail in, are simply used to that piece of software.  Revit has taken a little of the software out of detailing.  Don’t let that scare you.  Come along with us.

But first take off those Crocs and put on some real shoes.

Revit Arch 2010 Advantage Pack and 2009 UI Issue

I haven’t been able to extensively test, but it looks like the aforementioned 2009 UI “hack” that you can do in 2010 gets confused a wee bit with the Advantage Pack on Revit Architecture 2010.  I plan on spending some time in an upcoming post discussing the Advantage Pack (I know – you can’t wait), but wanted to get this out there now since I haven’t seen it in the Intertubes yet.

We always knew the 2009 UI tweak wasn’t fully supported, but it looks like the Advantage Pack has set it a couple more steps out of alignment.  First up, the newly introduced (and much loved) Structural Ribbon loses the Truss tool in the non-ribbon interface.  Just can’t find it anywhere.

The “Tools” toolbar is… weird.  The tools seems to have been shuffled.

In order from left to right, I have listed what the icon is, and what the command acutally is:

What it looks like	What it really does
Show Work Plane	Work Plane
Linework	Work Plan Visibility
Spell Check	Spell Check
Join Roof	Find Replace
Some kind of mini Work Plane button	Tape Measure
Measure	Match Type
Show Hidden Lines	Linework
Hide Hiden Lines	Show Hidden Lines
Paint	Remove Hidden Lines
Split Face	Paint
Edit Cut Profile	Split Face
The Hammer from Demolish	Edit Cut Profile
Join Geometry icon next to the word “Demolish”	Demolish

So, that looks like one tool (Spell Check) that does what it looks like it’s supposed to do.

Align, Split, Trim and Offset all seemed to be saved from this madness, but the next batch of tools fared worse.

What it looks like	What it really does
Unjoin Geometry	Join Geometry
Cut Geometry	Unjoin Geometry
Don’t Cut Geometry	Cut Geometry
Copy/Monitor	Don’t Cut Geometry
Edit Wall Joins	Edit Wall Joins
CENSORED	Edit Beam Joins
Match Type	Join/Unjoin Roof

Again, we have one tool that matches its icon.  Weird.  Mousing over any of these will tell you what it’s going to do, you just have to ignore the picture.

A couple other items:

• If you have installed the Model Review, that just throws a nice error on bootup and won’t even load.
• The DB Link tool seems to be MIA
• The wonderful Convert Lines tool is also not to be found in the 2009 interface.
• I have not installed the Framing Tool, so cannot comment on its availability

That’s all that really jumped out at me.  I imagine I am missing some.  I also was not able to test/verify if keyboard shortcuts to these missing commands will work.  Also, I cannot verify if the other Revit flavors will have these same hiccups.

I’m not too surprised by this.  There was no reason for Autodesk to maintain support for a tweak that was completely unsupported.  The moral is just to pay attention if you are using the 2009 UI and have or plan on installing the Advantage Pack, you may be forced to hop in the Ribbon to see the full benefits of it.

I Hate It When I’m Wrong – Revit 2010 UI

We waited until the first service pack was out (web update – whatever they call it) before we deployed Revit 2010.  I, for one, was singing the praises of the much maligned Ribbon.  “See how clean it is?” I would say.  “See the nice big icons?” I would point out to the nonbelievers.  “See how it’s organized so well?” I would show my cats – who frankly didn’t care.

The first indication of something VERY bad showed up when one of our more experience Revit users reported crashes on his brand new 2010 model.  Not much content in there at all.  Very few views.  This was a tiny file that should have been able to run on a PC that was four generations old.  But he was crashing.  A LOT.  Up to seven times a day.

Soon after, another experienced user was reporting the same thing.  Different project.  Different user.  Virtually same hardware.  A recent multi-core XP 32bit workstation, 4GB of RAM, and a nice video card.  About 18 months old.  The hardware shouldn’t have been the problem.

We spent a LOT of time back and forth with Autodesk support.  They looked at the model, they told us to downgrade the video driver (yeah, you don’t hear that one too often, do you?) they said to not run anything along with Revit.

Same results.  Numerous crashes each day.  And these weren’t gentle “Revit is about to die – let’s save a recover file for you” crashes.  These were “POOF! Revit is gone!” style crashes.

We took two approaches.  For User A, we wiped his PC, upgraded him to XP 64bit and threw 8GB of RAM in (the max the motherboard could handle).  About 2 total hours of work stretched out over 2 days waiting for updates and installs, plus around $170 for the RAM.  He reports that 2010 is running great now!  Whoo-hoo!

For User B, we edited the ini file that allows 2010 to run in “debug” mode and use the 2009 interface.  About 2 minutes of work stretched over 2 minutes, plus around $0 for buying nothing.  He reports that 2010 is running great now!  Whoo-hoo!

Wow.  I am scratching my head over this.  I am so frustrated that the user interface was designed so poorly that it alone causes enough memory to be sucked from resources that are essential to the software running in a stable state.  I can’t be the only person out there.  And I’m not.  I think the 2009 UI tweak was one of the most re-tweeted Revit items on Twitter in a while.

We have been training new users on the Ribbon interface for months now.  Do I go back and spend my time showing them the old interface?  This is crazy.  This is shameful that Autodesk let their product ship with such a major memory drain.  Two updates later, still a big hole.

I certainly hope this is a major priority for the team.  I love the Ribbon.  I just wish it didn’t suck so much… memory.

Spinning! 3d View Quick Tip

Sometimes it’s the small things that we forget or overlook.  I was surprised to find out that someone who I could consider one of my top Revit users was unaware of the following little nugget.  So I thought I would jot it down here.

Navigating through your 3d view can get a little tricky when your model starts to grow.  Holding down the SHIFT key while center-button dragging will rotate your model around to let you get a better view of another side.  By default, Revit rotates around the center of your model.  This might lead you to rotate-pan-rotate-pan-rotate-pan until you see the angle you want.

Here’s the nugget.  Wanna know a nice trick?  Select an object before you rotate.  The object you select will then become the center of your rotation.  No more rotate-pan-rotate-pan-rotate-pan!

We’re All Getting Twitchy

Warning: I’m a geek.  A pretty big one.  I try to keep these posts to be on the low end of the geeky spectrum and relatively accessible to those of us in the design industry who like and use technology.  This post may…scratch that… this post will leak far over onto the geek end of the spectrum.  You have been warned.

So, we’ve all seen Twitch now.  Or we’ve read about it.  Only those that live on the west coast, or apparently somewhere in the Pacific can actually test it out.  No.  Not the islands.  Actually physically in the Pacific.  This is what I saw when I tried:

Do you have a boat out there running Revit?  Neither do I.

But I digress!  Basically, it’s remote processing and hosted computing for some of Autodesk’s apps.

And it’s about damn time.

I don’t know when the guys at Autodesk Labs first started thinking about this, but my brain started getting tickled to the idea as a possibility several months back when I heard about the announcement of the OnLive video game service.  If these guys have actually worked out a way to avoid the latency and possible speed hiccups of the Internet, and push really really high end video games to my screen, then there is absolutely no reason it can’t be done with high end BIM and modelling software.

High end PCs to run Revit and other apps are expensive.  I could argue that in the business world, aside from these production workstations, there isn’t much need for the typical user to have a PC that costs more than $500.  Our typical workstation runs us about $2500-$3000, and we don’t even get the uber high end stuff.

But if Revit could move to the hosted model, businesses could keep their hardware costs down.  IT departments wouldn’t have to waste time deploying and maintaining software on hundreads of PCs and could focus on what they are really good at (namely, napping and drinking Mountain Dew).

And Autodesk would benefit as well.  Piracy would be a thing of the past.  The subscription model that is the Holy Grail for all software companies would be a given.  They could start charging per hour for little spurts of licenses needed, like in the summer when we always get a bunch of interns, and then proceed to run out of licenses on our FlexLM server.

And for the really big firms that don’t want to share the time?  Autodesk partners with a hardware company, and rents out a Revit Rack.  A preconfigured, preinstalled series of rack servers that the local network users login to.  IT does a quick setup for IP address and the like through a nice web interface, then the Rack gets monitored and maintained and supported remotely by Autodesk.

Can you tell how giddy I am?  I am giddy.

If the latency issues and file sharing questions are resolved, then performance for the end user could be better as well.  If the virtualization for servers movement has shown us anything it’s that PCs and servers are wasted power.  For the majority of time, your PC sits there and twiddles its thumbs.  We spend the big bucks for gobs of RAM for the 15 minutes a day that your PC has to think really hard.  Aggregating the hardware resources into a shared server means more power for everyone.  Chances are you are not going to be rotating that 350,000 sf model in 3d view the same time that someone else is.  So for that minute, you get more RAM and horsepower from the shared resources.  And if more people come on board, you don’t need to spend $3000 in hardware and unknown costs in IT setup time, you just need to spend a couple hundred bucks to add some more RAM and maybe another hard drive to the server that hosts the application.  Oh, and you want to work on your model on the road or from home?  Piece of cake.  You just need an Internet connection.

This makes so much sense.  This SCREAMS it makes sense so much.  It’s good for firms, it’s good for Autodesk, it’s good for users.  It’s not often that everyone wins.

The Autodesk Labs guys are pretty darn smart, but they hit it out of the park with this.  I have no idea if anyone from Autodesk reads this blog (I bet “no”), but I hope that the business side of the company realizes that this is the future and they need to make it happen.

/nerd hat off

Placing a Tag Outside its Host

Seven times out of ten, you place a tag and you want it to be near that item that you are tagging, be it a door, room, window, whatever.  The tag wants to be near the host, otherwise, what’s the point?  You might as well be placing tags by tossing them in the air and letting them land wherever, like as much confetti.

Two times out of ten, you tag something and there just isn’t enough room for that tag to live there.  So, you enable a leader from that tag.  Excellent.  A nice clean arrow indicating precisely to what item that tag is referring.

Then comes the oddball.  That one time out ten (probably even more rare) that you want a live tag to pull info, AND you don’t want it near the object you’re tagging, AND you don’t want a leader.  The out-of-the-box tags in Revit don’t want you to feel like you can accomdate this.  As soon as you move a tag too far from an item, Revit tells you that you MUST activate leaders and have that connection back to the object in question.

We ran into this situation… and I won’t keep you in suspense, we found a way around it.

Our firm standards for interior elevations indicate that we show the room number JUST BELOW the actual elevation.  We wanted to keep the tag “live” so it was pulling the number from the room, in case the room number changed.  “Dumb” tags are SO un-Revit.  But, every time we dropped the tag down, it asked to turn on the leader, which we didn’t want to do.  How to fix? 

The question really was, how does Revit know if a tag is inside or near the item that’s being tagged anyway?  Opening up the tag itself gave us our first clue.  The only thing in there is lines, text and… some reference planes.  Quickly checking the reference planes indicated that these planes were of the “origin defining” type.  You probably see where I’m going with this.

Turns out, Revit doesn’t care at all where the “graphics” of the tag is at all.  The origin point of the tag is how Revit determines if a tag is inside or near an object.

Solution?  Create a new type that had the graphics droppped down below the origin.  As long as that invisible origin was inside the room, the tag was below the elevation, and it didn’t need a leader line.

This can obviously be applied to any tag type, with even more clever placement being defined by parameters and types.  It doesn’t happen often, but it’s nice to know that you can keep the “automagic” updating and not lose out to the last 10%.