@brewsky said:

The most important thing for this plugin is, add as much useful information to a model as possible, but keep it simple!

Well said!

Just wanted to share an idea:

SKIM - More than Sketchup less than BIM

SKIM - Skimmed down version of BIM

SKIM - Sketchy BIM

PS: I still hadn't got the time to fully test the new version. So much to do, so little time.

@brewsky said:

@jql said:

I just downloaded it and it's impressive.

Thanks! And thanks for your input! I have already added a few of your ideas!

I´m very happy to give my so very small contribution and I'm amazed that you really implemented some of those ideas.

I've been away for a while, working hard on some short deadlines. I see you've also been working hard here and with some great improvements. The entusiasm you show is inspiring and it made take the time to read almost all posts (skipped code posts wich are transcendental to me). It was a very interesting read but it made me write an even longer (VERY LONG) post this time.

It seems to me there are two main concerns with your plugin (wich are also BIM core aspects):

1 - Building modeling;
2 - Information management.

As this has been a very free willed forum and though you have many valuable opinions already, I will had my own if you don't mind. I hope it makes sense to you and you see it as positive criticism.

This is a very ambitious plugin. Or is it not? What you have already done is impressive and is still very open. I think you should clearly define what you want from those two aspects of BIM so the concept behind your plugin is also clearer and you can clearly concentrate on what to do to get there.

My take on the matter is that Sketchup is not CAD or BIM. I gave up on CAD for a while now and tried going BIM on my practice. I didn't think CAD was fast, or intuitive enough to model with, or was it nice to create powerful presentations or even easily communicate with others. In the opposite corner wearing red shorts was BIM, wich was very powerful and could do everything (or so we're told), but was very dull and technical and had a way of putting itself in front of creativity (It basically takes the fun out of architecture and makes you move to an even more specialized working environment).

What I want is to have a tool that makes me work fast, enables me to richly communicate better with all people that surrounds me (clients, engineers, suppliers, contractors,...) and helps me acomplish the things I need to do so I can build my projects correctly and efficiently. I would also love to have a tool that did it all and that wouldn't require me to jump from one software to another, exporting and converting and making me loose time between tools. I wanted a tool in wich I could center all my working process and wich I could use up to the end of the project.

So what I needed was not BIM or CAD. What I needed was Sketchup and I'm really happy I found it.

Of course sketchup has flaws (many) but they are mainly corrected by some really useful plugins (such as this one will be).

Behind Sketchup the main reason I would use a Sketchup BIM instead of true BIM platforms would be price. Otherwise If I needed BIM I would really want a very solid solution without any flaws. From what I read I didn't clearly understand what are you trying to accomplish in BIM terms. Do you mean to give Sketchup the hability to export more useful info into BIM platforms? Or by the other hand do you really want to turn Sketchup into BIM?

I honestly believe that you shouldn't follow this second path too deeply. That would be a huge undertaking of course, but that is not the Sketchup target market. We are simpler...(?)

Nowadays most BIM software has already the built in capability of importing Sketchup models and if those models are built following certain rules, this works very nicely. That was the main reason I started exploring Sketchup. So if this project of yours is aimed at enhancing the level of information you can send from Sketchup to BIM, then I think it will be most useful for the ones that tend to use sketchup just until a certain point of their workflow, and then move on for more complex/powerful/technical software.

So what is Sketchup really useful for?

Modeling of course. The first thing that makes someone download a plugin for Sketchup is modeling. Few of us use Sketchup for added information. Sketchup is really powerful in conceptual stages, acurate modeling and presentations. The problem is we, as architects, need more than this in our work, but I found that Sketchup can be much more and I am trying to push it further into contruction documents. The main reason for that is that I tend to push conceptual stage to the last minute, really until final construction documents are finished.

The main problem with Sketchup for construction is that it deals with "written" information in a very opaque way and though I believe this has much more potential to develop than modeling, it is not the sketchup core, or is it my main concern in my design proccess.

So, on a technical workflow point of view, the thing I miss the most is a clear and effective way to have metadata inserted in my components and groups, and be able to export it to a spreadsheet, updating it regularly, along with useful info like areas and quantities. If I could easily do that, I could then easily relate my model to an accurate and fastly generated report.

I am already able to use Sketchup exclusively for most of my projects though. If any of you knows a way I could easily generate this reports, you could end my quest right now!

For specially complex projects I can always hire a BIM team wich would interact with my own Sketchup centered team and your plugin could really enhance that connectivity.

I don't need anything else too much and I certainly don't like to design interactively from spreadsheets like you can do in BIM platforms.

I am an architect and I am able to draw. Space relations are more prone to be tested by drawing/modeling (phisically or in 3d) wich are creative processes more similar to perception of space and building processes. For me it's harder to write/formulate/parametrize spatial relations. I also don't care much for parametrization because what I want is to see what happens when I model. I don't want to stop designing to think about the rules I use. We in our pofession, are already too much constrained with rules, codes, responsability and budgets so I'm trying to free myself from working processes chains. Some parameters are nice but, for me, those are the ones your plugin tries to solve (thickness, height, interconnection of elements). Dynamic components solve a lot of other cases (repetition, lenght, scale, etc...), and though I don't know much about using them, I think they will be enough when I need further parametrization.

So, to me, the main focus of this plugin should, right now, be architectural modeling. After that is solved I suppose that with what you have already done, you could "easily" jump to phase two - Information Management.

With that cleared from my mind I will focus on your new developments but I will cut my post in two because I want to do further testing so I can reach some more conclusions. I hope I've helped somewhat so far.

Sorry for the long post, best regards and keep up the good work.

@brewsky said:

New version 0.10.3 uploaded!

I wonder what you think, especially about the workflow!
Is it intuitive, sketchup-like? What must be improved?

I just downloaded it and it's impressive. I couldn't believe my eyes when I saw almost the same workflow I use in SU, reproduced with 3 buttons. The main difference is I create a group for all exterior walls/roofs and if I need to edit one or more specific walls/roof slabs, I then divide those groups. In this way it is simpler for me to edit everything... This plugin might change that though.

I also like the way face offset works. Setting it to zero makes it very simple to work volumes, extruding them in. However, when you toggle to element view, I think it should hide the source faces. Of course this is only visible when offset has a given value, and for me this would happen only when working on interior walls. Nevertheless it would be a very nice feature to implement if it is possible.

This is also true for corners. The plugin also creates border faces for each wall/slab corner wich should not appear when you create a section. It would be nice if these border faces could also be hidden. If the plugin could do this, then we could create a section only with the finishing lines active and no diagonals in the corners.

@brewsky said:

I made the interface as lean as possible, not a separate button for walls/roofs/floors, just thick faces with properties.
Is this the way to go?

Definetely the way to go. Simple, effective, clean. The way it should be in Sketchup. I would have a request for one more button though. It would be wonderful if there would be also a button to disconnect/reconnect the faces created. You could then disconnect the faces of a wall for instance, edit that wall, and then reconnect all the faces. This would be most welcome when you want to create special non-uniform walls, walls with unparalell faces or any kind of correction to geometry in dificult corners where some gaps and misajustments happen. It could also be used to correct the bugs you mention below like losing relations between source faces and elements. If one can disconect a face that is not working and erase all associated geometry (but not window components), one would also be able to rebuild it from scratch "debugging" the model. Of course some of the features of windows and dynamic thickness would be lost, but project freedom and overall organization would be kept.

@brewsky said:

I know it still has bugs(losing relations, sometimes crashing), "bugreports" are also welcome!

The main inconvenience I found is that if the user doesn't follow the specific steps to make it work right, faces and elements start to fail and some of them disappear. In a creative and interactive process, specially on early stages of a project, this might mean that sometimes you will have to discard the whole model and start over for it to work.

In the present state it is definetely a very useful plugin for simple geometry projects. However for more complex ones I think users would benefit if they had more flexibility with the disconnect/reconnect faces button.

Having said everthing I felt, I think this will be an extremely useful plugin if it gets a little more solid. It could very well be, one of those that really enhances SU. Congratulations!

Thomthom,

I'm sorry if you don't go further with this plugin. I have been using JPP but, sometimes and on certain shapes, (the ones i usually work on) it creates a shell whose faces are not all parallel to each other. I was going to comment on the JPP topic but gave up when I read the manual. In there Fredo clearly states that JPP was designed in a way that those unparallel faces are supposed to happen sometimes.

Your shell, with those same shapes makes them always parallalel (at least in the preview). It really works well in the rest of the time and the only "flaws" I noticed were the ones I pointed. If only those previews could be materialized in "real" lines and faces...

If you give it a second thought, I do think it's worth putting just a bit more effort on this.

I don't want to sound ungrateful though, because in its current state, this is already one of my favourite plugins.

Best regards and thank you.

What a bomb "shell" Thomthom. Thank you very much for this beta that is already a great plugin.

I've been long waiting in the shadows, looking for a way to make myself useful around here. And if you don't mind, as I'm no coder I offer myself as a crash test dummie. I made some observations too and I hope you may be able to take some conclusions from them.

1 - As was already mentioned before, I noticed that with some geometry the preview is being shown but the shell isn't created with double click or enter. In my case It usually happens when the preview draws extra faces that intersect the original group/component's faces, getting outside. Besides that the preview is accurate. (If there was already a way to force the previewed geometry to become real geometry, this could be an awesome plugin.)

2 - This preview geometry thing is more prone to happening with closed or revolved shapes. Sometimes It works but creates strange intersections inside. If those faces are open and the shell isn't too deep, it works just fine and it's also very fast. When it does work it is flawless with all faces parallel and at the same distance from each other! I made some (almost) manual shells, just to compare and they were always perfect matches.

3 - I also tried negative shelling just to test if I could get the same shape back. When tt_shell works, the negative shell also works. I also noticed that once I inserted the negative value, SU remembered that direction as being positive. So if I wanted to make a "regular" shell after, I had to input a negative value again.

4 - In the file below there is an example of an object that did not create a shell but got some interesting results:
a) I tried creating a 30cm shell and it failed.
b) I created the shell above using Normal Push Pull and the usual intersect faces method. I got the resulting 30cm shell I wanted.
c) From this resulting shape I tried using the plugin to create a second 30cm shell and it worked, wich I found strange.
d) So I tried to make a negative shell wich would grow the shape back 30cm. This again worked and I was back with the shape I got in b).
e) I picked back the smallest shell from c) and I tried to make it grow 60cm so I could have the initial shape in a). It failed to do so.
f) I tried again, but with a 59cm value and it worked.
g) I noticed that there were some edges that were very small and that if I would have tried to reuse the 60cm value that I did before, they would become vertices.

What this suggests me is that maybe the plugin isn't able to manage edges that shrink until they become vertices. The reverse should also be true and that means the plugin also can't handle vertices that must grow into edges. Of course I'm not absolutely sure but probably there's something down this line.

This is all illustrated in the scenes from the attached file.

If it is ok I'll keep testing the plugin until I find some more examples of this.

I really hope I made a small contribution and thank you again for all the hard work. It's much appreciated.

shell_test.skp

Why don't you try quadface tools? I saw that plugin solve a similar problem in an armchair, though I don't know if it was rendered after...

Also, have you tried different mapping options inside Kerkythea? It seems to me that if you set one material only (without photomatch turned on) you could easily tweek the UV cilindrical map in KT to get the right effect.

@tig said:

Best to do a second JPP, intersect and erase unwanted bits?

Thought so... If it isn't easy to do "by hand" it should be very hard to automate. The work around I used was to insert a volume inside the window component in a separate layer. Then I intersected the inner wall with the component and removed what I didn't want. I then hid the separate layer of the volume. To close the walls I had to make the component unique and draw the necessary lines. Repeat the process for the other windows... Hate complex shapes.

It will be an headache to change the window positions later and not being able to drag the inner hole with it.

I will try to find a tool that "drapes" a hole in a surface in the direction of a given face/component. It would be cool if it could also "drape" a fill in the hole after, so I could easily heal the face again at a much later time if i need to... maybe by extending all the lines in the surface. There should be something like that.

Any way thanks for your excelent tools and your unbelievably fast answer! That I just didn't expect!

I'm new to sketchup but obviously I realized already that half its potential is due to people like you. So thank you!

This tool immediatly struck me by its usefulness, (well it sort of punched me in the nose, actually).

It still lacks one thing to be perfect though, (not that I am complaining or anything, I don't wanna be punched again):

You clearly state that the inner walls must be paralel to the outer ones. However, would it be very hard to make the tool work even if they're not?

To put in other words, could one use this tool in a wall made of Joint-pushed-pulled loft surfaces?