Google is Listening!

thomthom

@thomthom said:

@unknownuser said:

In defense of SketchUp in general, not in particular v8, I find that when I turn off Edges and Profiles, or more generally speaking all Style related aspects,

Yes, style effects can slow down a model as badly as shadows. (even Colour by Material and Colour by Axis slows things down as SU has to draw the edges in multiple GL operations) I usually have a "Modelspace" scene that turns on my optimized modelling style with all effects off. I model with that style and only swap to scenes with effects on when I'm ready to export.

For maximum performance, you should stick to rendering styles that can be entirely handled by your GPU. For quick tips on optimizing SketchUp performance, see the "Making SketchUp run faster" Help Center article. For excellent and exhaustive detail, have a look at the "How do I make SketchUp run faster?" page on the SketchUp Sages site.

john
.

Noticed this from that page you linked:

@unknownuser said:

JPEGs are just fine. If you import images into your model, use JPEG images rather than TIFF images. TIFF images tend to have large file sizes and take more computing resources to display.

Now, I thought that the format did not matter when it came to displaying the image. I thought that once it was loaded the bitmpa was uncompressed in memory. But am I wrong? Are the GPU's optimized to make use of JPEG data?
I'd always imagined that compressed files where slower to load, as they had to be decompressed before they could be used. I imagine reading that for game optimisation tips...

dacad

@jbacus said:

You are being a bit disrespectful to me and my background, but my architectural education has gifted me with a very thick skin. I do think that you're confusing the specialized needs of archviz professionals with the generalized needs of all users of SketchUp. I talk to lots of heavy SketchUp users whose eyes would cross if I asked them about unwarping their UV's.

As i said i wasn't trying to be disrespectfull, just honest. My skin is as thick as yours also (architect too), and my area of work, in the last years, have been more wide than just archviz, so i'm not just talking for the sake of talk...

@jbacus said:

Also, you're assuming that there is one true method for dealing with "3D meshs" that the SketchUp team is either too stubborn, distracted or incompetent to implement for you. That's a little disrespectful, too– and it is kind standing in the way of moving this conversation forward as well.

I know there's not just one way to deal with 3D, right now if i'm not mistaken there's surface modeling, solid modeling, nurbs modeling and voxel sculpting (if you want you can add normal 3D sculpting too). And by the way, SK is still a surface modeler, even with "solid" tools, because calling it "solid" and add a volume calculator, don't turn the model to a solid, because if i section cut a SK "solid" it's still empty inside, just like a surface...
And just by saying something like this it should show you i'm not so ignorant or stupid in some matters too, so please don't assume that either.

@jbacus said:

There are certainly things we can add to SketchUp that will improve its usefullness for archviz pros without losing sight of the general needs of all SketchUp users. Itemizing those things and discussing them in detail would move this conversation forward.

So how, losing time doing better UV tools, wouldn't be good for the majority, or even all, that do 3D with sketchup, especially Pro users? For archviz, industrial design, characther modeling, game graphics, organic modeling? I don't think there's one area that didn't went trough a situation, on time or other, where better uv tools weren't need it...

So to sum things, i still don't have an answer to my first post related to new software architecture, or an opinion about the sugestions i made to SK beeing wide enough and usefull for everybody ( http://forums.sketchucation.com/viewtopic.php?f=10&t=30586&start=60#p268908 ), and i just discovered that spherical mapping it's a too wide and complex thing to most SK users, and so not need it.
Better forget about unwarp uvs and new or updated modeling and animation tools...

pep75

Jeff I do believe that there're benefits. But right now I just feel that Google puts more effort in making SU a tool for "Google Earth Content creation" than an architectural- or model tool. And you can't deny, SU started to sell their software in a box with "architectural" visualization tool on it. What is the real improvement? Come on...angular dimensions in LayOut and the "solids"-tools that they have bought from Whaats? I mean....I'm stunned. Where's the bezier tool? Where's the push-pull tool on non-flat surfaces? Where is the bevel/chamfer tool? Where is the bend tool? Where is the Subdivide and Smooth tool? All tools that are essential in SketchUp and that are made by plugin developers, but needs to be standard in SU to become a "decent" architectural tool.
Okay, I'm a teacher in architecture and at the moment I really doubt to suggest students to work with SU. It stucks to be a tool with a "freeware" abilities....I'm sorry to say, but organic architecture is still almost undo-able in SU...It's a shame because the UI of SU is UNIQUE...

Aerilius

@pep75 said:

...but needs to be standard in SU to become a "decent" architectural tool.

Those tools are surely very useful and should not be missing in an architectural program, but SketchUp is not [only] an architectural visualisation tool, if it is supposed to be an all-purpose 3D app for everyone, we have to accept that.
Would we like SketchUp to be bloated with tools that we don't need, because someone might need them? We should understand the plugins as part of SketchUp's featureset (although not developed by Google), but surely they are neither hacks nor workarounds but mature tools except that they are not in the default installation.

Marian

@pep75 said:

Where's the bezier tool? Where's the push-pull tool on non-flat surfaces? Where is the bevel/chamfer tool? Where is the bend tool? Where is the Subdivide and Smooth tool? All tools that are essential in SketchUp and that are made by plugin developers, but needs to be standard in SU to become a "decent" architectural tool.Okay, I'm a teacher in architecture and at the moment I really doubt to suggest students to work with SU. It stucks to be a tool with a "freeware" abilities....I'm sorry to say, but organic architecture is still almost undo-able in SU...It's a shame because the UI of SU is UNIQUE...

Well we have the plugins and they work and we should be glad we didn't have to wait for the SU team to implement them.
I do agree that some measure of integration for some of those vital functions should be added to the core SU, but knowing some of those plugins you should be more aware that organic modeling can be achieved and I don't think having the same exact functions as the plugins, natively inside SU, improves modeling in any way, maybe only stability and relibility.
I'm not an architectural expert by any means but I don't see any insurmountable obstacle in doing organic architecture with SU, it may not be perfect or 100% accurate but I think most stuff is doable and you should try harder or let your students find they're own solutions, I think you'll be amazed.

I do understand your view, I too feel a little disappointed at the level of innovation in the new releases and also the direction they took, but having said that, the new improvements are ok and dare I say it...a bit cool.

jonorman

I love SU, been in love with it since the day it found me, back in version 4 or somthing.......

SU is a great "sketch and modelling" software wich has a lot of great features, and with the superb induviduals that makes all the amazing plugins, taking sketchup to new heights.

I want to reverse the question and say: Where do GOOGLE want to take Skectchup ?? A better 3d modelling program for Design, architecture and engineering, or a free software easy to use for everyone so we can make a model of our houses and post it on google earth ?? Or are trying to have it both ways ??

I understand i write this a little black/white and sarcastic, but U`ll get the idea of the question....

I dont want to offend anyone with this question, i just want to know google`s point of view of SU future.

I still love SU and use it every day for work and for fun, will allways use it due to the greaaat plugins showing up... It`s like christmas every so often. so i cant lay it to rest, just have to wait and see what comes up next in the plugin forum.

RickW

@aerilius said:

Does the Ruby API allow to set a color value to a pixel of a texture? Or to reload a texture image very fast?

#1 - Yes and no. I worked through the PNG specification and have an .rb library (that works with zlib.so) to read .png pixels. It's a matter of converting XYZ to UV to get the (approximate) pixel at a pick point in the model. That, and time for Ruby to read a .png

#2 - I have an as-yet-unreleased plugin for this. It needs a bit of polish yet.

jeff hammond

I guess this post is aimed at two people- thomthom and jbacus.

Using uv tools for instance. Thomthom, I know you've been messing around with uv mapping in su ruby a lot but as I understand it, there isn't enough access(?) for you to fully do what you need to do.
What would you need in order to develop the necessary plugin to deal with proper uv mapping?

Then at JB, suppose Thomas would like something added to the api, is it just a matter of you allowing more access or do you have to do more than that to let him have at it? I'm sure there is a point where you reach sensitive material limits on how much access you can give but I'm just curious as to how this process works.

Phil Rader AIA

The tone of this conversation has now reached the correct balance. Developers and end users agreeing to mutually work towards a common goal. I clearly remember sitting in a bar in California watching Justin Chin (MonsterZERO) show some of his amazing work and his workflow to several of us beta testers. Todd Burch was present and you had an end user showing his "made up" methodology to accomplish a desired task (rigging) and a developer Todd saying things like we could do that series of manual tasks with ruby. It was great to be present when the people who have the experience and knowledge on both sides of a teeter-totter are working together >>SU is the support in the middle<< The users are on one side and the developers are on the other. It's a balancing act The developers must not be too heavy and the users must not be too heavy. John put it very eloquently when he noted that there is a continuous need to improve hardware and software to meet the demands of the user but as soon as a user has the software and the hardware to meet his current needs he or she promptly pushes the boundaries beyond their current needs and hardware and software are forced to expand yet again (it's a viscous cycle).

Lets just be thankful that we are all here and have a voice in this cycle. Are all the ideas, requests, improvements, etc. going to happen...did all of your plans for your life come true...NOPE. I for one have about 999,000 dollars to go to make my goal of becoming a millionare...hehe.

Maybe John could release a SU retrospective video showing snippets of the development of Sketchup through the ages. I Started using it in version 1 so I remember many of the milestones that were made. It's helpful to know where you have been in determining where you want to go.

DIEGO-RODRIGUEZ

sr. jbacus

I just want to ask a question.
why not make more complete the four most powerful tools of sketchup?

in sketchup 8 pushpull 2.0 updated, but very few improvements.Why?
why not add more features?
plugins exist, but works best native pushpull

for example:
conical pushpull
pushpull to multiple preselectd faces
vector pushpull
Pushpull outside group
use "Shift+one click mouse" for Redo same Push Pull

[flash=853,505:11pd0pbv]http://www.youtube.com/v/y0GiUThzwoI&fs=1[/flash:11pd0pbv]

please make sketchup tools most complete, would help me work better and faster

would appreciate your response.

Phil Rader AIA

Diego-Rodriguez Thanks for the informative post. You show some useful Rubies I was not aware of.

I think the Idea of extending core functionality of SU to incorporate Successfully implemented Ruby programs is a great Idea. I'm not sure about the Process and or the ability of Google to simply adopt functionality that was created by third party developers though.

John are there examples of Native SU functionality that originated from outside companies. I believe that the whole Dynamic components addition came about in this fashion if I am not mistaken.

jeff hammond

[edit] @diego

And do that to everything then we'll have a thousand tools to navigate

I'd much rather see this done through plugins. Maybe in a more manageable set up than now but please, don't add all that stuff to core sketchup.
I'm willing to guess that even the su programmers are surprised with what's happen via ruby. A management system or depot etc weren't even being thought of as necessary at the beginning but we're still doing things the same way as when ruby was first implemented. I'm sure the suteam has already began discussing ideas etc about plugin development but just in case, I'll put my 2 cents in saying I think they should.

JClements

At LadyBugz (Carolyn):

Why are postings tagged as inappropriate and the author not informed of why and by whom?

Does anyone have the capability of tagging a posting as inappropriate?

thomthom

@unknownuser said:

Using uv tools for instance. Thomthom, I know you've been messing around with uv mapping in su ruby a lot but as I understand it, there isn't enough access(?) for you to fully do what you need to do.
What would you need in order to develop the necessary plugin to deal with proper uv mapping?

As for UV mapping directly there isn't much that we don't have access to. But UV mapping isn't all that trivial to get right.
However, we do lack the ability to determine if a material on a face is set to Projected, and we also lack the ability to set the Projected property - which would make things easier, and cooler.
Then there is the thing around it, it would help if we could draw bitmaps onto the SU viewport to make good previews of the texture, or make a UI like that native SU texture tool where you move pins around. Can do without it, but it's just no fun imagining the images...

I'll see if I can post a preview video of my experiments soon.

thomthom

@unknownuser said:

Then at JB, suppose Thomas would like something added to the api, is it just a matter of you allowing more access or do you have to do more than that to let him have at it? I'm sure there is a point where you reach sensitive material limits on how much access you can give but I'm just curious as to how this process works.

I don't think it's always as easy as connecting a wire from A to B to hook up Ruby API access, some features could be hidden deeper inside some other code. The Ruby API was retro-fit into SU at version 4, wasn't it?

thomthom

@jclements said:

At LadyBugz (Carolyn):
Why are postings tagged as inappropriate and the author not informed of why and by whom?

Does anyone have the capability of tagging a posting as inappropriate?

Postings here at SCF or at Google's pages?

pep75

Everybody is right. But that was not my point. I think, that Google IS developing more and more tools(what also can be done by plugins....?!) for their Google Earth project and put them in SU.
I think Ruby IS a great option to make a good piece of software better, and I'm glad we have that option. But what I wanted to say is that I'm missing just ESSENTIAL modeling tools, that can also be integrated in the existing tools(for instance; selecting a line an use PushPull on it..). And, off course (to Marian), I know that you can model anything with Sketchup...but there are tools that can do it a LOT faster. I think Google must come up with different versions, that's all. I LOVE plugins!!! But the core of the concept must be good to...especially when you market your software on architectural academies....

DIEGO-RODRIGUEZ

@unknownuser said:

[edit] @diego

And do that to everything then we'll have a thousand tools to navigate

no. More tools not.
I just meant the possibility of integrating more functions to the same tool. pushpull to be more complete.
I think skechup not have to add more tools only support to developers and improve the current core.
but, the tools rotate, scale, follow me, move, pushpull and offset. have to be more complete. may be the video options or other.
sketchup all functions must be improved. but not add new tools

jeff hammond

It'd be neat for Instance if say joint push pull could be available via plugin but it ties itself in some way to push/pull.
ie- if I choose to add jpp to my system, I could assign a modifier key to regular push pull in order to activate it. Or maybe a flyout toolbar that If I click and hold push/pull, my other options pop out but the UI in general remains clean unless I decide to clutter it up with options I choose.

If someone makes a selection tool ruby, it could attach itself to select etc.

jeff hammond

Right Diego, I was talking about what you are. Not making new tools per se but adding new functions to the existing ones. You're still going to have to activate the new features in some form or another be it a toolbar, menus, modifier keys etc. However it would be done, there is going to be a whole lot more to navigate. It's going to be a mess.
I would like exactly one of the functions you mentioned to be in SU. Some might want more, some even less.

Point being, I would like to customize to suit my needs. That's why the plugin system works so well (or, could work so well)

I mean, I have a few ideas that would help me work faster but I don't expect them to be in core sketchup. You probably wouldn't like my ideas so why should they be cluttering up your screen?