there are a few people here who seem to have strong opinions on this, so i thought i would turn to you guys to get some feedback that might direct my future ambitions in ways to kill time when i'm between things. i'm thinking of xenakis, who seems to think programming is the future, or starchitectalpha, who almost seems to be offended when i point out the shortcomings in revit. back in the overtime thread i had asked an offtopic question, so i thought it would be better to pull this into a new thread.
also, null pointer linked to a designanalyze? podcast in donna's design rubbish thread. input from anyone else knowledgeable is appreciated too.
first off, i accept that revit does a lot of things pretty well. aside from that though, there are some things it doesn't do well, which really should be developed a bit better. most of the remaining problems as i see it are in documentation rather than modelling (though i think there are a few lingering problems with modelling as well). lack of numbered/bulleted lists and tab stops in text being the biggest, but i also think the way revit handles schedules is really not as robust as it should be.
my first (1) question is, why all the talk about dynamo rather than c# or even c++? does dynamo give you the same access to the database in revit and the same tools to pull information out of and put information into the model? is the difference mostly with the gui, or is dynamo mostly only good for repeating and modifying model elements? i don't know much about either, so i'm hoping to hear thoughts from people who might have some experience with both, and why they prefer one over the other.
my second (2) is, how does the revit database work? it seems a lot of the potential in revit revolves around that database (what with 'parameters' being database elements right?). i haven't had a chance to get into it too much, and probably won't for a while, but can i use c# or dynamo to access and modify the data in doors and walls and everything else, or do i have to pull the database out of revit and into something like iis/access/apache/something and modify it with some other software program?
mostly my thought is whether it's better to put energy into learning how revit works with .net or dynamo.
Non Sequitur
Nov 4, 15 11:44 am
Can't help you on those Curt but I am also interested on this topic. I'm shouldering the Revit learning curve in my office and we're just a few weeks into construction on our first Revit project (300k sqft 9-storey office building). It's been my baby for over a year now. They grow up so fast! sniff, sniff
I agree with you points of the pros and cons of the software but there were times during the modelling and CD phase where I needed to pull certain data from the model but could not without external means, ie export to another software.
I do however disagree with the general statement (mostly coming from fresh grads) that CAD is dead or that revit is first coming of christ. Will stay tuned to this and see where it goes.
zonker
Nov 4, 15 11:49 am
but can i use c# or dynamo to access and modify the data in doors and walls and everything else -
You can use Dynamo to read and modify data then write it back into Revit - the trick is to manage the indices - keep them straight using the categorize function nodes - that is to say read from the revit file, modify those indices refereing to the the params, then write them back. it works like C#, uses C# or .net
zonker
Nov 4, 15 12:00 pm
BIM and programming is the way of the present and future - no stopping it - this is the 21st century folks - just like long ago when Prop powered airliners were replaced by jets - you dont see any old DC-7s flyin aroun - cept in historic airshows
I just dont see a majority of designers (read: visually oriented people) learning C++, C#, Vbscript , python or what have you to the facility that someone with a computer science background. Same goes with overly complex programs like Digital Project, it will be a specialty demanded by only a few firms. That being said, Dynamo/Grasshopper are great bridges, and can accomplish most of what we need (dynamo needs something like hoopsnake to create iterative loops...).
Revit definitely has its shortcoming - the massing environment is crap, and will never be as facile or iterative as massing in Rhino/Sketchup. I strongly agree with Xenakis - BIM is here to stay, and it will well serve you to learn it... dinosaurs- it's too late for you, just grab a junior designer and throw rolls of sketches at him. But if you're graduating university now and turn your nose up at Revit, you can expect that to adversely affect your chances at most major firms.
Also, BIM is not well geared towards early conceptual design- more information is asked prematurely which can bog down the quick iterative process. I don't care about micro level information and data - whether something is a CMU wall or a GFRG ceiling - if i'm just defining general massing. Also, NURBS is king for free form modeling. A friend of mine toured Zaha's office in London a year ago and most people ditched using maya for most of their design work, and now use Rhino.
in conclusion... learn all you can, BIM is the future, Parametricism for 1,000 years (Patrik, Patrik where are you?). JK on the last part. Maybe.
null pointer
Nov 4, 15 1:07 pm
Ivory: I don't know about that first statement. Once you grasp object oriented programming, the act of coding feels an awful lot like sketching. Once you know architectural design is about dealing with agents and constraints; if you can represent the concepts visually, and procedurally, you're making use of the same faculties that a good architectural designer uses.
I remember being in graduate school and starting my projects by drawing adjacency diagrams with bubbles and labels. By the second year, those adjacency diagrams had morphed into procedural diagrams. By the third year, a lot of those bubbles had become functions and classes in python (for maya).
zonker
Nov 4, 15 1:58 pm
null_pointer
exactly - only I learned programming before studying architecture - the same thinking -
zonker
Nov 4, 15 2:09 pm
Most of the students in grad school like null_pointer are learning Grasshopper, Python and C# - whats amazing is the facility some of these people have when going from Grasshopper to Dynamo like its all one and the same - on my weekends, when I don't have to be at work that is, I am learning dynamo from the many tutorails taht exist on both RevitKid and the Dynamo website
curtkram
Nov 6, 15 9:59 am
so if starting today, c# would be a more versatile and useful platform to learn, whereas dynamo is more for people who want to easily visualize puzzle pieces sticking together?
i put ms visual studio on my computer, and it looks like revit has it's own ide type thing that's pretty robust, so even with c# you have stuff in color and autocomplete for commands, so it's not really that hard to keep track of what you're doing.
i guess the answer to my second part is that it's all there and just as magic as it sounds, and once i dive into the rabbit hole i will see it for myself, just as alice did.
archiwutm8
Nov 6, 15 10:53 am
I think I am slightly qualified in answering some questions, I worked with a autodesk reseller and had regular contact with autodesk and many people who have developed many workflows and helped many firms start their BIM journeys - That's my background.
To the questions, I apologise if I'm repeating stuff as I haven't bothered to read the whole thread yet.
Dynamo much like grasshopper is a visual coding blocks created to help those without a coding background to get into and design, it allows you to use Python for the more code savvy designer. I is extremely powerful and allows you to control pretty much any aspect of Revit. So in that case Dynamo is still a "program" whilst C# or C++ are languages and should be considered completely different.
As stated above you can control Revit aspects using Dynamo, so you can pull out the parameters from Revit to Dynamo and do what you please. I don't believe that every designer needs to learn Dynamo at all, it should and will always be a specialist skill in design, many firms and studio already employ designers with a computer science background to help them through their complex designs. I've met many studios/firms who are only going to AutoCAD now.
curtkram
Nov 6, 15 11:49 am
when you say 'completely different,' do you mean they do completely different things, or they just have a different way to go about doing the same things? i guess my first thought is that they both have the same api, just a different way of showing it?
so if i am equally comfortable with c# and dynamo, is there a benefit or detriment to choosing one over the other? i suppose we should also assume that i'm pretty sure i will never design one of the loopy curvy buildings like schumacher wants to do.
dynamo has a gui where i can have a slider to input numbers, whereas c# probably won't, so maybe accessing information in a database and doing something with it would be better with c#, whereas having a graphic slider to modify a number in a wall parameter would be better with dynamo, due to the gui? can i build some sort of input/output gui thing with c# like i could with vba?
i have some experience with vba, javascript, code editors, and stuff like that. i don't have experience with grasshopper or dynamo, but i messed around with mit's scratch for an hour once, and it seems to me sticking with the code editor just makes more sense than some weird puzzle piece thing.
zonker
Nov 6, 15 12:17 pm
i suppose we should also assume that i'm pretty sure i will never design one of the loopy curvy buildings like schumacher wants to do.
I get to because it's something I do on weekends for fun - I get to play shumacher and occasionally enter into an Evola competition - My day job is your standard Multi-family buildings - those get built and vanilla Revit, skp. and acad along with trace is all we ever need for that. got enuff challenges with codes and stuff
curtkram
Nov 6, 15 12:22 pm
i guess i could clarify that i don't intend to go down that path because i don't want to ;)
in a perfect world where i had all the money and all the client and could do anything i want, it would still be a box. just maybe a better box.
aside from that, there's a lot of information hidden in a revit file. might be able to make some of that code stuff a bit easier if you put it together right.
zonker
Nov 6, 15 1:26 pm
When I was at SOM, they had people with programming backgrounds Digital Design Specialist as they are called do programming - Dynamo will be mostly a big office type tool
MikeJarosz
Nov 6, 15 2:27 pm
I was at SOM for almost 25 years. I started out as a designer, but as computers started to appear in the studio, I picked up on them, eventually learning to program, first Acad, then Revit. For the last 5 years of my time with SOM, I became one of those "Digital Design Specialists".
The important thing is that I did all this past the age of 40. I did not and do not make sketches and hand them to the interns. In fact, much of what I do today is fix the Revit models those interns screwed up, trying to make Revit behave like Acad.
gwharton
Nov 6, 15 3:17 pm
Computer programs are tools, just like pens and pencils or chipboard and glue. To the extent you master the use of more tools, you will be able to do more. But there is no one magic "swiss army knife" for doing everything architectural. Such a tool does not and probably never will exist.
Use the right tool for the job. With more tools, you can do more jobs. With only one tool, you can only do one thing, regardless of how well it's done.
Related to that, Revit seems to be limiting itself by trying to be too many things and not doing any of them very well. Grasshopper is better than Dynamo. Sketchup is better than Formit. ArchiCAD is better at straight BIM tasks. etc.
null pointer
Nov 6, 15 4:34 pm
A few weeks ago I took on a couple of small projects. I priced them according to my CAD time. Did them in Revit after having broken in my standards on a few tougher projects (retail stuff with lots of custom millwork) well... I can't divulge my effective hourly rate on those two projects (fixed fees are magical)... but... easily tripled my typical hourly rate. Not having to coordinate plans and elevations is such a huge time-saver.
l3wis
Nov 6, 15 7:20 pm
Null pointer, was your set better or worse than if you had used CAD?
null pointer
Nov 6, 15 8:51 pm
probably the same. marginally better in terms of graphic consistency.
gruen
Nov 7, 15 12:22 pm
Id love to learn how programming could speed up revit use for me, but all of my projects are very vanilla so no need for complex shapes. I use AutoCAD for really basic projects that are mostly plan based (interior renovations, tenant fit outs, etc) and REVIT for anything that needs exterior elevations and lots of sections-mostly this is new ground up buildings. I do save time for both project types but find the initial time to do the revit model is about twice the time as in CAD
archiwutm8
Nov 7, 15 8:28 pm
Mike could you elaborate on more of what you did? Because that's the route I want to take, I want to be a digital computational design specialist.
there are a few people here who seem to have strong opinions on this, so i thought i would turn to you guys to get some feedback that might direct my future ambitions in ways to kill time when i'm between things. i'm thinking of xenakis, who seems to think programming is the future, or starchitectalpha, who almost seems to be offended when i point out the shortcomings in revit. back in the overtime thread i had asked an offtopic question, so i thought it would be better to pull this into a new thread.
also, null pointer linked to a designanalyze? podcast in donna's design rubbish thread. input from anyone else knowledgeable is appreciated too.
first off, i accept that revit does a lot of things pretty well. aside from that though, there are some things it doesn't do well, which really should be developed a bit better. most of the remaining problems as i see it are in documentation rather than modelling (though i think there are a few lingering problems with modelling as well). lack of numbered/bulleted lists and tab stops in text being the biggest, but i also think the way revit handles schedules is really not as robust as it should be.
my first (1) question is, why all the talk about dynamo rather than c# or even c++? does dynamo give you the same access to the database in revit and the same tools to pull information out of and put information into the model? is the difference mostly with the gui, or is dynamo mostly only good for repeating and modifying model elements? i don't know much about either, so i'm hoping to hear thoughts from people who might have some experience with both, and why they prefer one over the other.
my second (2) is, how does the revit database work? it seems a lot of the potential in revit revolves around that database (what with 'parameters' being database elements right?). i haven't had a chance to get into it too much, and probably won't for a while, but can i use c# or dynamo to access and modify the data in doors and walls and everything else, or do i have to pull the database out of revit and into something like iis/access/apache/something and modify it with some other software program?
mostly my thought is whether it's better to put energy into learning how revit works with .net or dynamo.
Can't help you on those Curt but I am also interested on this topic. I'm shouldering the Revit learning curve in my office and we're just a few weeks into construction on our first Revit project (300k sqft 9-storey office building). It's been my baby for over a year now. They grow up so fast! sniff, sniff
I agree with you points of the pros and cons of the software but there were times during the modelling and CD phase where I needed to pull certain data from the model but could not without external means, ie export to another software.
I do however disagree with the general statement (mostly coming from fresh grads) that CAD is dead or that revit is first coming of christ. Will stay tuned to this and see where it goes.
but can i use c# or dynamo to access and modify the data in doors and walls and everything else -
You can use Dynamo to read and modify data then write it back into Revit - the trick is to manage the indices - keep them straight using the categorize function nodes - that is to say read from the revit file, modify those indices refereing to the the params, then write them back. it works like C#, uses C# or .net
BIM and programming is the way of the present and future - no stopping it - this is the 21st century folks - just like long ago when Prop powered airliners were replaced by jets - you dont see any old DC-7s flyin aroun - cept in historic airshows
http://dynamobim.org/
http://formit360.autodesk.com/blog/
http://formit360.autodesk.com/blog/
I just dont see a majority of designers (read: visually oriented people) learning C++, C#, Vbscript , python or what have you to the facility that someone with a computer science background. Same goes with overly complex programs like Digital Project, it will be a specialty demanded by only a few firms. That being said, Dynamo/Grasshopper are great bridges, and can accomplish most of what we need (dynamo needs something like hoopsnake to create iterative loops...).
Revit definitely has its shortcoming - the massing environment is crap, and will never be as facile or iterative as massing in Rhino/Sketchup. I strongly agree with Xenakis - BIM is here to stay, and it will well serve you to learn it... dinosaurs- it's too late for you, just grab a junior designer and throw rolls of sketches at him. But if you're graduating university now and turn your nose up at Revit, you can expect that to adversely affect your chances at most major firms.
Also, BIM is not well geared towards early conceptual design- more information is asked prematurely which can bog down the quick iterative process. I don't care about micro level information and data - whether something is a CMU wall or a GFRG ceiling - if i'm just defining general massing. Also, NURBS is king for free form modeling. A friend of mine toured Zaha's office in London a year ago and most people ditched using maya for most of their design work, and now use Rhino.
in conclusion... learn all you can, BIM is the future, Parametricism for 1,000 years (Patrik, Patrik where are you?). JK on the last part. Maybe.
Ivory: I don't know about that first statement. Once you grasp object oriented programming, the act of coding feels an awful lot like sketching. Once you know architectural design is about dealing with agents and constraints; if you can represent the concepts visually, and procedurally, you're making use of the same faculties that a good architectural designer uses.
I remember being in graduate school and starting my projects by drawing adjacency diagrams with bubbles and labels. By the second year, those adjacency diagrams had morphed into procedural diagrams. By the third year, a lot of those bubbles had become functions and classes in python (for maya).
null_pointer
exactly - only I learned programming before studying architecture - the same thinking -
Most of the students in grad school like null_pointer are learning Grasshopper, Python and C# - whats amazing is the facility some of these people have when going from Grasshopper to Dynamo like its all one and the same - on my weekends, when I don't have to be at work that is, I am learning dynamo from the many tutorails taht exist on both RevitKid and the Dynamo website
so if starting today, c# would be a more versatile and useful platform to learn, whereas dynamo is more for people who want to easily visualize puzzle pieces sticking together?
i put ms visual studio on my computer, and it looks like revit has it's own ide type thing that's pretty robust, so even with c# you have stuff in color and autocomplete for commands, so it's not really that hard to keep track of what you're doing.
i guess the answer to my second part is that it's all there and just as magic as it sounds, and once i dive into the rabbit hole i will see it for myself, just as alice did.
I think I am slightly qualified in answering some questions, I worked with a autodesk reseller and had regular contact with autodesk and many people who have developed many workflows and helped many firms start their BIM journeys - That's my background.
To the questions, I apologise if I'm repeating stuff as I haven't bothered to read the whole thread yet.
Dynamo much like grasshopper is a visual coding blocks created to help those without a coding background to get into and design, it allows you to use Python for the more code savvy designer. I is extremely powerful and allows you to control pretty much any aspect of Revit. So in that case Dynamo is still a "program" whilst C# or C++ are languages and should be considered completely different.
As stated above you can control Revit aspects using Dynamo, so you can pull out the parameters from Revit to Dynamo and do what you please. I don't believe that every designer needs to learn Dynamo at all, it should and will always be a specialist skill in design, many firms and studio already employ designers with a computer science background to help them through their complex designs. I've met many studios/firms who are only going to AutoCAD now.
when you say 'completely different,' do you mean they do completely different things, or they just have a different way to go about doing the same things? i guess my first thought is that they both have the same api, just a different way of showing it?
so if i am equally comfortable with c# and dynamo, is there a benefit or detriment to choosing one over the other? i suppose we should also assume that i'm pretty sure i will never design one of the loopy curvy buildings like schumacher wants to do.
dynamo has a gui where i can have a slider to input numbers, whereas c# probably won't, so maybe accessing information in a database and doing something with it would be better with c#, whereas having a graphic slider to modify a number in a wall parameter would be better with dynamo, due to the gui? can i build some sort of input/output gui thing with c# like i could with vba?
i have some experience with vba, javascript, code editors, and stuff like that. i don't have experience with grasshopper or dynamo, but i messed around with mit's scratch for an hour once, and it seems to me sticking with the code editor just makes more sense than some weird puzzle piece thing.
i suppose we should also assume that i'm pretty sure i will never design one of the loopy curvy buildings like schumacher wants to do.
I get to because it's something I do on weekends for fun - I get to play shumacher and occasionally enter into an Evola competition - My day job is your standard Multi-family buildings - those get built and vanilla Revit, skp. and acad along with trace is all we ever need for that. got enuff challenges with codes and stuff
i guess i could clarify that i don't intend to go down that path because i don't want to ;)
in a perfect world where i had all the money and all the client and could do anything i want, it would still be a box. just maybe a better box.
aside from that, there's a lot of information hidden in a revit file. might be able to make some of that code stuff a bit easier if you put it together right.
When I was at SOM, they had people with programming backgrounds Digital Design Specialist as they are called do programming - Dynamo will be mostly a big office type tool
I was at SOM for almost 25 years. I started out as a designer, but as computers started to appear in the studio, I picked up on them, eventually learning to program, first Acad, then Revit. For the last 5 years of my time with SOM, I became one of those "Digital Design Specialists".
The important thing is that I did all this past the age of 40. I did not and do not make sketches and hand them to the interns. In fact, much of what I do today is fix the Revit models those interns screwed up, trying to make Revit behave like Acad.
Computer programs are tools, just like pens and pencils or chipboard and glue. To the extent you master the use of more tools, you will be able to do more. But there is no one magic "swiss army knife" for doing everything architectural. Such a tool does not and probably never will exist.
Use the right tool for the job. With more tools, you can do more jobs. With only one tool, you can only do one thing, regardless of how well it's done.
Related to that, Revit seems to be limiting itself by trying to be too many things and not doing any of them very well. Grasshopper is better than Dynamo. Sketchup is better than Formit. ArchiCAD is better at straight BIM tasks. etc.
A few weeks ago I took on a couple of small projects. I priced them according to my CAD time. Did them in Revit after having broken in my standards on a few tougher projects (retail stuff with lots of custom millwork) well... I can't divulge my effective hourly rate on those two projects (fixed fees are magical)... but... easily tripled my typical hourly rate. Not having to coordinate plans and elevations is such a huge time-saver.
Null pointer, was your set better or worse than if you had used CAD?
probably the same. marginally better in terms of graphic consistency.
Id love to learn how programming could speed up revit use for me, but all of my projects are very vanilla so no need for complex shapes. I use AutoCAD for really basic projects that are mostly plan based (interior renovations, tenant fit outs, etc) and REVIT for anything that needs exterior elevations and lots of sections-mostly this is new ground up buildings. I do save time for both project types but find the initial time to do the revit model is about twice the time as in CAD
Mike could you elaborate on more of what you did? Because that's the route I want to take, I want to be a digital computational design specialist.