PDA

View Full Version : Blender SoC Projects announced.


Nichod
05-26-2006, 11:21 AM
http://www.blendernation.com/2006/05/24/google-summer-of-code-results-announced/

"Nicholas Bishop - Interactive Sculpting with Multi-Resolution Models
Mentor - Jean-Luc Peuričre

One of the neatest tools to come about in open source modeling recently has been SharpConstruct (http://sharp3d.sourceforge.net/mediawiki/index.php/Main_Page), a powerful stand-alone brush modeler built originally for Linux and now also available for Windows. This summer, Nicholas Bishop, creator and head of the SharpConstruct project, will be bringing his knowledge and experience to Blender. He plans to create native Blender brush modeling tools (not Python plugins!) similar to those available in SharpConstruct, as well as a feature not yet in Sharp called Multi-Resolution Meshes.............

..............A very big thank you to Nicholas Bishop, Jean-Luc Peuričre, and Google for making the creation of this tool possible this summer! More information about the goals of this project can be found in the SoC proposal draft, found here: http://sharp3d.sourceforge.net/mediawiki/index.php/Google_SoC"


"Benjamin John Batt - Modifier Stack Upgrade
Mentor - Daniel Dunbar

In Blender 2.40 we saw the inclusion of the mesh modifier stack, and while this great new feature has enabled a lot of new customization in how a mesh is formed and rendered, there are still a number of features that it lacks. Looking to change that is Ben Batt, a Computer Science/Computer Systems Engineering student in Melbourne, Australia. His project will involve upgrading the modifier stack, upgrading the current modifiers to work with the new modifier stack, and adding 3 new modifiers...........

..............As you might imagine, we are also very excited about this SoC project as well. Thanks go out to Ben Batt, Daniel Dunbar, and once again Google, for their work and support to bring these much needed features into Blender. For more information, take a look at the project page: http://mediawiki.blender.org/index.php/User:Artificer/SummerOfCode2006"


"Dmitriy Mazovka - Sky Generator
Mentor - Kent Mein

The process of creating a sky for a scene is typically just throwing a big sky picture into the scene. But where is the control of the sky’s look and feel? How do you light the scene to mimic the sky image? How do you animate the scene with a static sky image? Dmitriy Mazovka has proposed to solve the riddle of the sky with his ambitious SoC proposal.

First on the list is a controlled, physics-based simulation of skylight, sunlight, and aerial perspective effects that comes with cloud simulation based on cellular automata (http://en.wikipedia.org/wiki/Cellular_automata), all of which will be able to be stored as an environment map. Additionally we’ll be given the ability to then animate the sky, and watch the sun walk across the sky as the clouds float by.

Not enough? Next is the creation of scene objects that can be used in constructing the lighting environment based on the generated sky properties. The proposal gives the example of light shafts being generated automatically from volumetric clouds based on the type of light you’ve added to the scene.

Integrated sky creation is certainly a sign that the Age Of Blender is upon us! Thanks in advance to Dmitiriy Mazovka, Kent Mein as the mentor, and yet once again to Google for bringing this functionality to Blender. A bit more information can be found here:

http://projects.blender.org/pipermail/bf-committers/2006-April/014374.html"

paintbox
05-26-2006, 03:34 PM
Interesting! Lets see what these guys can do :)

I read somewhere that N-gons where one of the SoC proposals...is that still going to be developed ?

KayosIII
05-27-2006, 09:09 AM
I think that Ton said that most of the proposals that didn't make SoC funding are going to be going ahead anyways.

PyRoT
05-27-2006, 09:25 AM
Could Blender become better than commercial 3D apps? I mean, a lot of these features seem very modern!

LetterRip
05-27-2006, 03:57 PM
ngons I don't think was proposed as a SoC project since the author will be doing school this summer, but he does plan to work on it.

LetterRip

VirgilioVasconcelos
05-27-2006, 04:47 PM
Could Blender become better than commercial 3D apps? I mean, a lot of these features seem very modern!

I think "better" is a very personal oppinion. I really think Blender is one of the best apps around (commercial ones included), and I think the fact of being an open-source app will push it more and more, because the growing community of developers and users.

Projects like "Elephants Dream" will help the CG professionals to take a better look at this great app.

Asgorath
05-27-2006, 10:48 PM
I think Blender is just not agressivily competitive, especialy in the capitalist sens (since it's open source dah...). But it is a great application none the less.

Papa Lazarou
05-28-2006, 01:28 AM
The big challenge for Blender is cutting into the commercial market. So far it's mostly being used by hobbyists. When Blender begins to be adopted by more and more professionals, it will improve exponentially. All signs are that it's getting closer and closer to that turning point. What will really make the difference is when studios are using it in their pipeline and contributing to the codebase.

private
05-28-2006, 02:31 AM
I think Blender is just not agressivily competitive, especialy in the capitalist sens (since it's open source dah...). But it is a great application none the less.

Blender UI Overall = competitive. Without an overall = stuck in hobby zone forever.

Zarf
05-28-2006, 03:58 AM
I think that Ton said that most of the proposals that didn't make SoC funding are going to be going ahead anyways.

While that would be nice, its also probably unrealistically optimistic. I am willing to wager that none of the proposals that recieved no funding will make significant progress in the next 3 months (if at all).

Good code takes time to write and programmers are real people who need to eat.


Blender UI Overall = competitive. Without an overall = stuck in hobby zone forever.


In your opinion, what is it about Blender's UI that keeps it 'stuck in hobby zone forever'?

Cheers,
Xarf

MooseDog
05-28-2006, 05:22 AM
i've tried to be fair, going back several times to try out new features (which are fantastic! and on-par with any 3-d program), and simply find the whole ui impenetrable. icons!? no thank you. just as difficult. blender needs to enlist a "user experience" expert. mho.

claybub
05-28-2006, 07:40 AM
blender, from what I've seen is a wonderful program that is capable of quite a lot.
But as stated above, the UI is daunting. Everytime I open the program the interface disrupts my creativity, which is why its purely a hobby for me. But the fact that its getting more and more support is very good and I look forward to following its progress :)

KayosIII
05-28-2006, 12:10 PM
While that would be nice, its also probably unrealistically optimistic. I am willing to wager that none of the proposals that recieved no funding will make significant progress in the next 3 months (if at all).

Good code takes time to write and programmers are real people who need to eat.

Yeah you are right, I think that the summer of code money will really help give said students time to really get into what they are doing. The thing though about Summer of code is that a lot of people end up getting payed for something they were going to do anyways. Most of the coders working on blender are doing so as volunteers. So we will just have to see.

As for the UI I think it takes about two weeks of solid use before you crack it. (as opposed to about 2 days for wings3d)... It is a good UI is is just a pain to learn think of it as a car with only with 1st and 5th gear -- When you are travelling in 5th Gear everything is sweet just getting there is a pain.

kursad_pileksuz
05-28-2006, 01:30 PM
I have used all the major applications and use all the major applications in production (maya, modo, hexagon, xsi et) , and i have found blender ui better than maya when it comes to usability for example. So lets not present our personal opinions as universal facts. I am saying this because when it comes to blender everyone opens the app for a min and then give up and they conclude that blender ui is not very well. Blender ui is as weird as any other application, i think that because it is free to download and you do not pay for it you do not invest time to learn little more about it.



Blender UI Overall = competitive. Without an overall = stuck in hobby zone forever.

poly-phobic
05-28-2006, 02:14 PM
I have used all the major applications and use all the major applications in production (maya, modo, hexagon, xsi et) , and i have found blender ui better than maya when it comes to usability for example. So lets not present our personal opinions as universal facts. I am saying this because when it comes to blender everyone opens the app for a min and then give up and they conclude that blender ui is not very well. Blender ui is as weird as any other application, i think that because it is free to download and you do not pay for it you do not invest time to learn little more about it.

i dont think what he said about the UI is totally a personal opinion.

put 10 people in a room, and have 10 machines running xsi maya and blender, and just see which app they'll feel more comfortable in. im betting it wont be blender. [not saying blender is a bad application for its price ]
blender has some great features. and im betting if the UI and functionality was designed to look more familiar to other apps, im sure it would be taking off at a much more quicker pace.

efbie
05-28-2006, 02:57 PM
put 10 people in a room, and have 10 machines running xsi maya and blender, and just see which app they'll feel more comfortable in. im betting it wont be blender. [not saying blender is a bad application for its price ]
blender has some great features. and im betting if the UI and functionality was designed to look more familiar to other apps, im sure it would be taking off at a much more quicker pace.
Put 10 kids in a room, with bikes and tricycles, and just see which one will be the easiest to ride...

Blender's interface isn't intuitive at first, because it hasn't that third wheel that will slow you down in the future. If you've never tried blender for more than a week, then you can't understand why so many people swear by it.

Papa Lazarou
05-28-2006, 04:06 PM
Blender needs a UI overhaul. It's like learning to drive a car that has the gearshift in the steering wheel and the indicator next to the handbrake. You can probably get used to it, maybe after a while you will think that any other configuration would just be stupid. But try and drive a different car and you're pretty much back to square one.

If Blender had a good UI overhaul, it would attract so many new users that it's speed of development would change fast.

Bliz
05-28-2006, 06:04 PM
I haven't got 'too' much of a problem with the interface but I wouldn't claim it doesn't have problems or couldn't be improved. And I don't think people complain about the interface because the app is free as some people have problems with the ZBrush interface and zbrush is a $500 app.

I'm pretty excited about the inclusion of Sharp Construct concepts/code. I downloaded and had a play with SC last night because I saw it mentioned on this thread and I've got to say I was pretty impressed. Great for organic stuff.

SylvanMist
05-28-2006, 06:19 PM
These projects sound really nice, good luck guys and thanks! :)


Everyone who complains about the UI of Blender seem to just not want to take the time to learn it. Yes you do need to read the manual a bit, go through some tutorials. Other programs are the same.
I've used Maya, 3D max, and I also prefer Blender's UI over theirs once I got the hang of it. It is just more customizable and faster to work in for me.

Duhast
05-28-2006, 06:45 PM
Blender UI Overall = competitive. Without an overall = stuck in hobby zone forever.


As the Great Wise Man said:

"There are no bad UIs, only incompetent users"


If Blender had a good UI overhaul, it would attract so many new users that it's speed of development would change fast.


It does attract "so many new users" every day (you don't lurk through the blenderartists forum, do you?), and it's got the fastest "speed of development" when compared to any other 3D app (6 months more or less)

Zarf
05-28-2006, 06:57 PM
Yeah you are right, I think that the summer of code money will really help give said students time to really get into what they are doing. The thing though about Summer of code is that a lot of people end up getting payed for something they were going to do anyways.

I think you should modify this to say, "The thing about the summer of code is that a lot of people end up getting payed for something they *wanted* to do anyways." However wanting to do something and having the resources to actually be able to do it are two different things entirely.

Students are usually very poor, so between working the summer and getting paid actual money or working on a proposal that failed to get funding and starving which do you honestly think they will pick?

Most of the coders working on blender are doing so as volunteers.

I'm well aware of this since I actually am one of those volunteer coders.

Cheers,
Xarf

Zarf
05-28-2006, 07:08 PM
As the Great Wise Man said:

"There are no bad UIs, only incompetent users"


I hope your not serious about that. Theres plenty wrong with Blender's UI that *isn't* the fault of 'incompetent users'.


it's got the fastest "speed of development" when compared to any other 3D app (6 months more or less)

This sort of statement has so many things wrong with it I can't even begin to adress them all. I will say that the frequency of releases is no way to measure the speed of development though.

In my opinion this sort of talk is incredibly harmful to Blender's reputation.

Cheers
Xarf

Duhast
05-28-2006, 07:28 PM
I hope your not serious about that. Theres plenty wrong with Blender's UI that *isn't* the fault of 'incompetent users'.

Cheers
Xarf

I'm dead serious. There are no perfect UIs but in the end it ALWAYS comes down to the user. The word incompetent in not an insult, it just means lack of skill or aptitude to do something, which can be anything: in this case, the use of a particular brand of 3D software.

Period.

Nichod
05-28-2006, 07:44 PM
While that would be nice, its also probably unrealistically optimistic. I am willing to wager that none of the proposals that recieved no funding will make significant progress in the next 3 months (if at all).

Good code takes time to write and programmers are real people who need to eat.



With that in mind then Blender would have never progressed into what it is today.

And as far as Blender's UI is concerned...about 80-90% of the time you can avoid seeing most of the UI if you actually learn the program and its various shortcuts. While I agree the UI could be improved in parts (material for instance and the modeling tool setup) its not nearly as bad as is implied and shouldn't be used as an excuse for not learning the program.

JA-forreal
05-28-2006, 08:21 PM
Whoah yeah to the SOC! And what a summer this will be.

When I was a teen summers were spent hanging around doing whatever on my free time. If only we had opensource apps and boatloads of of free development tools a modem connection away...... Hey we may have been living “La Vida Matrix” right now. Hehehe. The olden daze were quite funky if you ask me.

Hummm, I often have to search for some of the Blender buttons. I rarely even use them. My favorite Blender hotkey is F12.

I wonder what Tons favorite Blender hotkeys are?


For of you who know and are aware of how cool it is to get down with it,
Blend on!

For the rest of you cool Cg folks,
have fun and have a great summer!

FunkyCowie
05-28-2006, 09:08 PM
I am trying to learn how to use blender right now and the UI is more complicated than it needs to be... when I learned Maya everything is neatly located in easy to find areas... and took much less time for me to find everything. I am not incompetent, or lack skill or whatever... it just is a pain in the ass... I think it could use a good overhaul to make it more user friendly. Unless you guys who are saying its not bad are just so used to it your afraid of it changing?

JA-forreal
05-28-2006, 09:22 PM
I hope your not serious about that. Theres plenty wrong with Blender's UI that *isn't* the fault of 'incompetent users'.



This sort of statement has so many things wrong with it I can't even begin to adress them all. I will say that the frequency of releases is no way to measure the speed of development though.

In my opinion this sort of talk is incredibly harmful to Blender's reputation.

Cheers
Xarf

Zarf, I think that buttons are not the way to go for future computer UI's. I kinda look forward to hands free computing in 2030 or so. Opensource computing is not about maintaining the status quo of common place computer interaction. That's better left up to development that partners with cpu hardware, input devices and GPU tech. Open source development is more of a home grown approach than mainstream. The "if it works well don't just fix it, make it better" approach has been good for Blender.

BTW man, your Path Select script tool has made my 3d design life so much better.

Blender's UI has been a question of great debate for years. As a user of software of all types I can find thousands of things to complain about. Often it's as good as it gets so I deal with it or I'm unproductive. Ah, 2030.

JA-forreal
05-28-2006, 09:30 PM
Unless you guys who are saying its not bad are just so used to it your afraid of it changing?

I'm saying that I pretty much go with the flow of what the Blender opensource development team dishes out. I make suggestions. Some vary useful features have been added to Blender due to user suggestion. But if you can't change a feature yourself you have to love it or leave it alone. With all of the options that 3d cg artist have now we can find an app suitable for our needs.

As far as learning to use Blender the best option that you have is to stick to the forums. You can even upload a .blend and let more experienced users kick it around a bit. I learned blender with these methods. Practice your hotkeys and learn where things are in the UI. Blenders UI is not that deep so it shouldn't take too long to get it.

If you use Maya and know it well, for a start only use Blender to UV map your .obj objects. Learn a bit about the materials (nodes now also), camera, lighting, rendering, etc., workflows. Don't try to take on everything. Only use the stuff that you need to use for now.

Blend on!

VirgilioVasconcelos
05-28-2006, 09:35 PM
Well...

As much as the Maya users are used to its UI, Blenderīs are too.

EVERY 3D app demands you some effort to learn it. Once you get into Blenderīs UI, youīll work faster than anyone (believe me). Some cool things like non-overlapping windows, context-sensitive actions have been there for years. Now, Silo and Modo uses some of those things and every one raves about...

Of course Blenderīs UI is not perfect. No appīs is.

Things like hotkey customisation are about to arrive (Tom Musgrove may know it better), I think just after 2.42. Maybe some presets to help users who arrive from other apps, like Maya, XSI or Max.

But trying to use a 3D app without reading the manual? Well... good luck for any app! =D

I would suggest this link for the ones who seriously want to try this great app out:

http://mediawiki.blender.org/index.php/QuickStart

Cheers!

Papa Lazarou
05-28-2006, 09:38 PM
It does attract "so many new users" every day (you don't lurk through the blenderartists forum, do you?), and it's got the fastest "speed of development" when compared to any other 3D app (6 months more or less)


I do lurk the blenderartists forum actually. Far too many new users get exited about the potential of blender, only to grow frustrated with it's steep learning curve. What baffles me is that there is a certain contingent among blender users that likes it that way. There seems to be the idea that any attempt to make it more artist-friendly would somehow threaten it's functionality.

JA-forreal
05-28-2006, 10:06 PM
I do lurk the blenderartists forum actually. Far too many new users get exited about the potential of blender, only to grow frustrated with it's steep learning curve. What baffles me is that there is a certain contingent among blender users that likes it that way. There seems to be the idea that any attempt to make it more artist-friendly would somehow threaten it's functionality.

Computers are generally not user friendly out of the box. You have to read a manual just to get things going. I'm all for a hammer and nail approach functionality for basic computing. Your more technical people will already be aware of "power nail gun" methods of use. 3d design is a bit more technical than drawing brush strokes and getting instant satisfaction. Apps like Maya and Max expect you to already be aware of what your getting yourself into...

So..... Blender's approach is not quite as deep for pro 3d production but somewhere in the middle.

But after you master 3d task, you thank yourself for sticking to it.

Have fun.

Duhast
05-28-2006, 10:26 PM
I do lurk the blenderartists forum actually. Far too many new users get exited about the potential of blender, only to grow frustrated with it's steep learning curve. What baffles me is that there is a certain contingent among blender users that likes it that way. There seems to be the idea that any attempt to make it more artist-friendly would somehow threaten it's functionality.

I'll quote Ton Roosendaal (Blender's Dad) from a recent interview:

"Many people expect that Free Software is an easy accessible mass audience product... / ...whilst the complexity of commercial products like Maya or Houdini is perceived as a confirmation of its "quality" "

http://en.wikinews.org/wiki/Interview_with_Ton_Roosendaal_about_Elephants_Dream_and_free_content_movies

Papa Lazarou
05-29-2006, 12:05 AM
I've done things with both Blender and 3ds Max, as well as many other cg programs. I'm not so foolish as to assume that there is not a considerable amount of time involved in getting proficient. But it is possible for complex programs to have an easier learning curve. You will always need to put in a lot of time, but there are better ways of doing things, and a lot of that comes down to the UI.

The first time I used blender, it didn't even have an undo. You'd ask on the forums, and the userbase at that time didn't seem to think it needed an undo, that it was better without it(since it encouraged better working methods, supposedly). I think Blender is a great piece of software and an incredible achievement, I just wish the blenderheads would not always take a defensive attitude to constructive criticism.

JA-forreal
05-29-2006, 01:37 AM
......................

The first time I used blender, it didn't even have an undo. You'd ask on the forums, and the userbase at that time didn't seem to think it needed an undo, that it was better without it(since it encouraged better working methods, supposedly). I think Blender is a great piece of software and an incredible achievement, I just wish the blenderheads would not always take a defensive attitude to constructive criticism.

I'm just a Blender user. If I get as far as being a Blender developer I'd be working out cool things in my spare time. I think that certain people who use opensource apps forget that any commitment to an opensource app is voluntary. People can criticize you all they want. If it's free and useful what can they really complain about?

I now use Ubuntu as my main Linux OS. I have had issues with driver support and whatnot. I can kindly ask other Ubuntu users and developers for help and make kind suggestions. But I know the deal, it's free stuff. I compile it myself, or apt-get it and configure it. I'm not expecting any opensource app to be offered up like a commercial app. I get yields from opensource apps and so do other people. We know that these apps are largely self serviced.

In the future a version of Blender may come out with full tech support for official non cvs builds. That would suit some people needs. It may have a price tag, etc. In this world you have to charge for this type of service. Other than that Blender is what it is. The other apps like Maya, XSI, etc are not free and come with high quality tech support, etc.

As a opensource developer you can focus on personal development goals. You can develop apps that help you get work done. You don't even have to be heavily concerned about what mainstream apps are up to. If the mainstream developments focus on film cg you can focus on a game cg app, etc.

A user of a film cg app comes over to the community of game cg app users. They complain about the lack of film quality image output or something. Do you really think that the game cg community cares that much about their complaints? If they are an opensource community they may listen to the film cg app user and get ideas. Film cg output may be available in the game cg app in a few months. Not that these users really need the output. Cool features are cool features.

So one fine day we see something the likes of "Elephant Dreams" from an opensource app that's not commited to film cg production. As a Blender user I think to myself, "oh snap, Blender can do that too". Then I smile and and continue to Blend on with ever so much of a vengeance of Blendering. There is nothing wrong with having more features than you actually need.....

Have fun!

flawedprefect
05-29-2006, 01:51 AM
I use the Mac version at work. I've gotten used to the interface from several versions ago, and from that background, each interface improvement makes my tasks easier and easier.

My boss has just bought Maya, and is encouraging me to learn it because it is "the industry standard" and if we get any freelancers in to carry on work, we're likely to find more Maya users than Blender users.

So I am plodding along, re-learning a new interface and about Maya's tools. I love the power behind the program, but at present, I can model, light and render a scene from Blender in a fraction of the time.

I've done animated fly-on CD cases for prominent artists; created a stock library of photorealistic DVD and CD cases; even created a city block of buildingsd that spelt out a logo of an Australian radio station all in Blender and absolutely wowed the clients.

(check out www.steam.com.au to see some of the results - any time you see a CD fly on - that's a Blender model)

I find some of the workflows in Blender don't work as cleanly as Maya, but what it lacks in those areas it compensates in speed. There is a shortcut key for pretty much any tool, and I can poly-model a CD case from scratch in 15 minutes using the simple Extrude, face cut and face fill tools. I've found that Maya works well when you take the time to set up all your shortcuts and shelves first, but I'm not yet proficient enough to figure out what I use more frequently - but Blender has an assigned key for most everything.

It's an uphill battle defending Open Source software, even when the results we can produce are as good as commerical software.

KayosIII
05-29-2006, 05:11 AM
Zarf: Yeah I am pretty happy to agree with what you just said (yes that was what I am trying to say). You don't have any idea when the input subsystem is going to be redesigned. (I know it is on the todo list)

Papa Lazarou: I think that some users take offense when a frustrated new user proclaims that the UI is ****ed and that it needs to be done from scratch. Some users overreact to this.. The basic UI is good - there are some parts which are nicer than anthing else I have used. Being able to switch complete layouts with a simple key combination is one of the better.

More often it is freatures that could be implemented better or that are not implemeted that frustrate me. I would however make the left mouse select by default and make sure that ALL tools work this way I would als make Y up and Z up user selectable with y up the default

Zarf
05-29-2006, 06:27 AM
Blender's interface isn't intuitive at first, because it hasn't that third wheel that will slow you down in the future. If you've never tried blender for more than a week, then you can't understand why so many people swear by it


Use it for 8 years, and you will see why some of use swear *at* it :)
The fundamental concepts behind Blenders interface are very good and make it worth using. There is a lot of 'nasty' stuff in there though.

For instance, when exiting/entering editmode some panels in the Editing Context of the Buttons Space just magically dissapear, causing your whole panel layout to slide around. Now the reason for this is that some panels only apply to Editmode and some only apply to Object Mode. However the automatic repositioning of the panels when switching modes wreaks havoc with the user's 'muscle memory' and makes it very hard to find your panels.

If this wasn't bad enough, the principle of 'hiding things not relevant to a given mode' is not consistent either; there are plenty of buttons in panels of the Buttons Space that shout 'ha-ha you dumb slob, you can't do that in this mode!' (this happens a lot elsewhere in Blender as well, such as the toolbox and drop-down menus).

I could list tons more nasties, but I think you get the idea.


I'm dead serious. There are no perfect UIs but in the end it ALWAYS comes down to the user. The word incompetent in not an insult, it just means lack of skill or aptitude to do something, which can be anything: in this case, the use of a particular brand of 3D software


Your just splitting hairs now. The quote you referenced said that there 'are *no* bad UIs', and placed the blame on the user. Thats the sort of sentiment that only a (poor)programmer could love.

The fact of the matter is that its much more complex than that. There *is* such a thing as a bad UI, and there are ways to evaluate this.

With that in mind then Blender would have never progressed into what it is today.


Blender would certainly have progressed at a much greater rate than it has if more developers were getting paid to work on it.

How many SOC applications that failed to get funding got worked on last year? (I will give you a hint, its a number lower than '1'). How many of the ones that DID get funding had significant development *AFTER* the initial 3 month period that Google paid for? I can think of one or two, out of ten.


And as far as Blender's UI is concerned...about 80-90% of the time you can avoid seeing most of the UI if you actually learn the program and its various shortcuts. While I agree the


I can list dozens of things that have no key shortcuts in Blender. For instance there are quite a few essential features that can only be accessed via the 'Buttons Space' (or 'Window' if you prefer the newer terminology). Furthermore, Blender has officially run out of key shortcuts for the View3d Space, yet new features keep getting added that affect the data that is visualized there. This means that it is becoming *impossible* to avoid the 'uglier' parts of Blender's UI '80-90 percent of the time'.


UI could be improved in parts (material for instance and the modeling tool setup) its not nearly as bad as is implied and shouldn't be used as an excuse for not learning the program.

The UI can be improved in lots of places, not just a few. The fundamental ideas behind it are quite good, however Blender has grown to a point where additional infrastructure needs to be added. These are real problems that do exist, otherwise the developers wouldn't be talking about things like an event system refactor or unified tool API.


BTW man, your Path Select script tool has made my 3d design life so much better


Glad to be of help, but I already rewrote that in C and committed it to CVS a couple months ago (in w-key menu). Its insanely fast now compared to the python version, and has new features, like allowing you to build compound paths easily.

Cheers,
Xarf

Apoclypse
05-29-2006, 07:12 AM
oops double post

Apoclypse
05-29-2006, 07:17 AM
What Blender needs is not a UI overhaul perse, but a rewrite. The Blender team themselves have admitted to this. They need to take a page from the K3d project which basically split the core from the ui leaving the ui as a seperate layer (think Maya but even more low-level). This is supposed to be worked on with blender 3.0, However the fact that they are adding features now as opposed to fixing things early doesn't bode well. OSS projects are usually a weird mix of slow/fast moving targets. When features are needed things get done and quick, however when it comes to fixing or rewriting core elements sometimes this isn't looked upon in a good light by developers since it means more work (but less work in the long run. ie. if the core and ui were to be seperated, we would most likely have a fully scriptable and customizable ui as it would be a fairly simple thing to implement).

I keep mentioning k3d because that project has done so many things right yet they don't anywhere near the support that blender does and so its a very slow project. The project follows alot of things I would like to see in blender. Everything is a plugin (like most modern apps, specifically Maya, and 3dsmax). The whole gui can be written anyway you deem neccessary without having to rewrite the whole thing (this is not exactly correct as there are some things that are handled through the gui that can't be handled in the core). What I like about this concept is that essentially any company can create there own 3d apps without reinventing the wheel. That means they can create say a custom lighting app or a zbrush like program, maybe even something like a compositor using a core framework which already does things like geometry handling and putting things in a dag as well as assigning materials and rendering. You would be limited by the amount of plugins and type of interface you need. Essentailly blender could become a toolkit (with a widget set based on the ghost library with python hooks for full scriptability) and the core can be another project which focuses on things like geometry handling, scripting language intergration (a la mel in maya, except with python ofcourse). Other essential core type things would be exposing an interface so that 3rd party developers can intergrate there renderers, This would be a great testing ground for 3rd party renderers because with a proper sdk it should be very easy to write a new render for blender (currently its not, aqsis comes to mind). Thats why Max has so many 3rd party renderers. When the program first came out they actually provided an easy way for developers to create their own plugins and renderers, making it extremely easy for them.

Anyways I posted way to much.

private
05-29-2006, 07:41 AM
More often it is freatures that could be implemented better or that are not implemeted that frustrate me. I would however make the left mouse select by default and make sure that ALL tools work this way I would als make Y up and Z up user selectable with y up the default

Well said. This would be a good first start.

LetterRip
05-29-2006, 08:02 AM
Apocolypse,

What Blender needs is not a UI overhaul perse, but a rewrite.

Blenders event system and bindings is scheduled for an immenent refactor also there is pending UI work that was scheduled from before project Orange.

The Blender team themseleves have admitted to this. They need to take a page from the K3d project which basically split the core from the ui leaving the ui as asepreate layer (think Maya but even more low-level).

Instead of complete rewrites, which tend to be a very bad idea for projects (particularly open source) there is usually a refactor of one or two subsystems per major release cycle. If you do a complete rewrite it is easy to have it kill the momentum of an opensource project since only a small core of individuals can do any work, whereas incremental refactors everything but the item being refactored can continue to be developed. In the past year or so, the transform system, the underlying mesh system and display system, the animation system, the sequencer system, the rendering system, and the materials system have all had substantial refactors and rewrites. With these rewrites came major feature improvements which would have been nearly impossible under the preexisting codebase without the refactors and rewrites.

This is supposed to be worked on with blender 3.0, However the fact that they are adding features now as opposed to fixing things early, doesn't bode well.

I very much doubt that a monolithic refactor will ever happen. Also 'fixing' things and adding features aren't mutually exclusive processes (indeed the biggest feature spurts have tended to be from a refactor of a section of code).

When features are needed things get done and quick, however when it comes to fixing or rewriting core elements sometimes this isn't looked upon in a good light by developers since it means more work (but less work in the long run. ie. if the core and ui were to be seperated, we would most likely have a fully scriptable and customizable ui as it would be a fairly simple thing to implement).

Full scripting and customization is a goal, however, the work for the refactor is non trivial. Since other areas were more important limitations, those other areas have been tackled first.

LetterRip

efbie
05-29-2006, 04:12 PM
Use it for 8 years, and you will see why some of use swear *at* it :)
The fundamental concepts behind Blenders interface are very good and make it worth using. There is a lot of 'nasty' stuff in there though. I totally agree, that's why I have big hopes for the event and UI refactor. However, the problem with newcomers is not the lack of keys on the keyboard, but the fundamental concepts behind blender ui. Things like edit modes, datablocks and viewspaces. They want a "wizard" interface with buttons "click here to make a nice render". That's what I wanted when I first opened blender, that's what I complained about on the forums, and thanks god they didn't listened to me.

I think that the only way to improve this is better documentation, but what can we do for people who don't even read it ?

brkn
05-29-2006, 04:36 PM
I keep mentioning k3d because that project has done so many things right yet they don't anywhere near the support that blender does and so its a very slow project. The project follows alot of things I would like to see in blender. Everything is a plugin (like most modern apps, specifically Maya, and 3dsmax). The whole gui can be written anyway you deem neccessary without having to rewrite the whole thing (this is not exactly correct as there are some things that are handled through the gui that can't be handled in the core). What I like about this concept is that essentially any company can create there own 3d apps without reinventing the wheel.

While that's all well and good in theory and the code may be just lovely, without the functionality, it's all just a bit of a fun programming experiment. I don't see any specialist lighting applications or Zbrush tools actually existing in k3D. You can't create 3D art in any practical sense with a beautifully engineered plugin system alone, you create it with tools and features that are designed and implemented to work together.

Caffeinemonkey
05-29-2006, 05:44 PM
What about sharp construct? Did anyone play with it?

I just downloaded it, Its a great little sculpting program. Needs a lot more


As for Opensource 3D. they all seem to be products written buy college profesors. It seems like they put time and effort into the most intelectual paths but miss the forest for the trees. This is depressing, becuase its like the diffrence between Socialism and Capitalism. Ones a utopian society on paper but a disaster in practice, while the other, well just works.

Papa Lazarou
05-29-2006, 06:53 PM
However, the problem with newcomers is not the lack of keys on the keyboard, but the fundamental concepts behind blender ui. Things like edit modes, datablocks and viewspaces. They want a "wizard" interface with buttons "click here to make a nice render".


Actually no. There are certain expectations that experienced users with prior 3d experience have going in that are confounded. This is the problem. There are certain more or less standardized conventions that everyone is used to, but in blender they felt the need to reinvent the wheel. By all means you can get used to doing things differently if you put in the time. I'm not saying Blender's UI is useless or inferior. There are other programs that also have UI issues. All I am saying is that it is a barrier to wider acceptance of the software.

Zarf
05-29-2006, 07:18 PM
Actually no. There are certain expectations that experienced users with prior 3d experience have going in that are confounded. This is the problem. There are certain more or less standardized conventions that everyone is used to,

Yet very few people (if any) have actually *pointed out* what these 'standardized conventions' are that everyone expects and that Blender is lacking. The most I see people (usually) say is that 'the UI sucks' and nothing more.

So please, by all means share some specifics.


but in blender they felt the need to reinvent the wheel.

Blender was originally written almost a decade ago as an in-house tool for an animation house with very specific needs. I don't think one of the original design goals was to 'reinvent the wheel', but was more along the lines of creating a system that was tailored to them.


By all means you can get used to doing things differently if you put in the time. I'm not saying Blender's UI is useless or inferior. There are other programs that also have UI issues. All I am saying is that it is a barrier to wider acceptance of the software.

Yet you havn't even described what this barrier is, only that it exists. Unless you do this, your not really giving constructive criticism, your'e just complaining.


I totally agree, that's why I have big hopes for the event and UI refactor. However, the problem with newcomers is not the lack of keys on the keyboard, but the fundamental concepts behind blender ui. Things like edit modes, datablocks and viewspaces. They want a "wizard" interface with buttons "click here to make a nice render". That's what I wanted when I first opened blender, that's what I complained about on the forums, and thanks god they didn't listened to me.


As far as I can tell no one is completley sure what people find wrong with Blenders UI, because 99 percent of people who object to it simply say 'it sucks' without describing what they mean. Unless more specific critiques are given its impossible to deduce what exactly people find so distasteful about it.

Cheers
Xarf

Milky
05-29-2006, 07:19 PM
Well...

As much as the Maya users are used to its UI, Blenderīs are too.

EVERY 3D app demands you some effort to learn it. Once you get into Blenderīs UI, youīll work faster than anyone (believe me). Some cool things like non-overlapping windows, context-sensitive actions have been there for years. Now, Silo and Modo uses some of those things and every one raves about...

Of course Blenderīs UI is not perfect. No appīs is.

Things like hotkey customisation are about to arrive (Tom Musgrove may know it better), I think just after 2.42. Maybe some presets to help users who arrive from other apps, like Maya, XSI or Max.

But trying to use a 3D app without reading the manual? Well... good luck for any app! =D

I would suggest this link for the ones who seriously want to try this great app out:

http://mediawiki.blender.org/index.php/QuickStart

Cheers!
I fully agree with you. It's a bit disturbing to see the excitement about Modo and Silo with *their* innovation. People seem to assume that just because a program is open source, it means that it sucks. I'm not a person who would hesitate to "try out" Maya or Max by visiting a "website", and I have tried them, but frankly, Blender is just so much more effective.

Anyways, a few years ago, while I still believed this was the case (about open source, or free stuff, rather), I tried out both Max and Blender. I just stumbled upon Blender actually, and I didn't like the UI, because I wasn't used to having to understand it in order to use it. In Max you can much more easily acess stuff even if you don't know. And half a year ago, I decided to get into 3d-modeling. I decided to try out blender, and I'm happy I made that choice. I am far more productive in Blender than I was in Max, and it can only grow better, eh?

I just have one comment for the hesitating. The best things in life are free.

Womball
05-29-2006, 07:25 PM
What would be nice would be a way of recording workflows. Something like a zscript. I think this could be more useful than a manual. If the users can see how things work from a script.

Aww they create a sky generator after I could have used one. :-( What are the mod stack updates?

KangarooMan
05-29-2006, 08:34 PM
What would be nice would be a way of recording workflows. Something like a zscript. I think this could be more useful than a manual. If the users can see how things work from a script.

Aww they create a sky generator after I could have used one. :-( What are the mod stack updates?

Yeah, I think so too. I found the ZScripts that came with ZBrush very helpful to get to know the basics of the application and to get comfortable with its interface.

You can read up on the Modifier proposal here: http://mediawiki.blender.org/index.php/User:Artificer/SummerOfCode2006

Matt :)

Papa Lazarou
05-29-2006, 08:52 PM
Yet very few people (if any) have actually *pointed out* what these 'standardized conventions' are that everyone expects and that Blender is lacking. The most I see people (usually) say is that 'the UI sucks' and nothing more.

So please, by all means share some specifics.

Blender was originally written almost a decade ago as an in-house tool for an animation house with very specific needs. I don't think one of the original design goals was to 'reinvent the wheel', but was more along the lines of creating a system that was tailored to them.

Yet you havn't even described what this barrier is, only that it exists. Unless you do this, your not really giving constructive criticism, your'e just complaining.

As far as I can tell no one is completley sure what people find wrong with Blenders UI, because 99 percent of people who object to it simply say 'it sucks' without describing what they mean. Unless more specific critiques are given its impossible to deduce what exactly people find so distasteful about it.

Cheers
Xarf

Well for starters, some basic things like the lack of standard edit menu functions next to the file menu. That's fairly standard. Another one would be using left-mouse-click to select objects rather than to position the '3d cursor'. While I'm sure some blender users find the 3d cursor quite useful, I fail to see that it is of such importance that it requires the left-mouse button at the expense of other functions.

lukep
05-30-2006, 01:21 AM
Well for starters, some basic things like the lack of standard edit menu functions next to the file menu. That's fairly standard. Another one would be using left-mouse-click to select objects rather than to position the '3d cursor'. While I'm sure some blender users find the 3d cursor quite useful, I fail to see that it is of such importance that it requires the left-mouse button at the expense of other functions.

Blender UI is non-standard yes. this can be disturbing if you play with it 10 mn, yes.
But blender strength is also that it is non standard.

The whole UI is designed for fast workflow, and quite succeed on that. For example, the 3D cursor is really useful when you know how to use it and need indeed a very fast access. the non-overlapping layout (way before modo and co), the total absence of dialog boxes, the multiple scenes, the flexibility of the layout, the strange but very efficient buttons windows make blender a software really fast when you overcome the oddities. What it means is that you wont easily graft a more standard UI on blender without loosing the benefits in workflow that you have from those. now that means too, the learning curve is a bit step.

there is still plenty of quirks in the UI showing its age and lack of global design (or more acurately, bad evolution of an initial good design), some examples are that buttons have no disabled state but sometimes disappear or reappear, there is hidden modes, every command need to be coded 3 times (for buttons window, toolbox and menus) and is sometimes different from one access to the other, there is serious limitations in some modes like vertex paint, the separation of edit mode from object causes some problems, etc ... things are adressed as modules are refactored. the animation tools rebuild did wonders on the pose mode.

But i have tried many other 3D softwares (both artistic and CAD, i'm a draftman), and all are complex to learn even when they seems more friendly while in the end blender is one of the fastest workflow around. In my experience, you can almost equate "user-friendly" with "bloated workflow". On the pure modelling front, Wings3D and Silo are a bit more efficient but they do a lot less and have problems of their own.

What really lacks and will be adressed shortly is some "getting warmed" tutorials to understand these famous non-standard conventions.

Apoclypse
05-30-2006, 02:01 AM
While that's all well and good in theory and the code may be just lovely, without the functionality, it's all just a bit of a fun programming experiment. I don't see any specialist lighting applications or Zbrush tools actually existing in k3D. You can't create 3D art in any practical sense with a beautifully engineered plugin system alone, you create it with tools and features that are designed and implemented to work together.

Well, k3d doesn't have anywhere the support that blender has. The reason is lack of features and its depedency on gtk, which is cross platform yes, but is nowhere near as small and compact as blender, because these have to be installed before k3d. Not really a problem as the devs have packaged everything pretty nicely, gtk on windows is just really buggy though. The oher reason being features is basically a chicken and egg syndorme. There is no interest because it lacks features, features are lacking because there is no interest. Don;t get me wrong I've been using since they first realeased it to the public (back when it wasn't opensource) and I've always admired the what the devs have done with it. Moreso now that things are goignso quickly in terms of feature adds. My only thing is that blender is not the end all be all of oss 3d packages. K3d has real potential but don;t have enough programmers to make it really shine. What it needs is to take a page from blender itself and create a gui that doesn;t depend on the gtk library and use something like the ghost library (or some other opengl crossplatform library) and lots of pics and showing what it can do. Blender has alot of word of mouth and marketing behind it so it attracts users and contributors alot more than other projects.

private
05-30-2006, 04:32 AM
Well for starters, some basic things like the lack of standard edit menu functions next to the file menu. That's fairly standard. Another one would be using left-mouse-click to select objects rather than to position the '3d cursor'. While I'm sure some blender users find the 3d cursor quite useful, I fail to see that it is of such importance that it requires the left-mouse button at the expense of other functions.

I agree...and that's just a start.

VirgilioVasconcelos
05-30-2006, 12:36 PM
Blender has alot of word of mouth and marketing behind it so it attracts users and contributors alot more than other projects.

Iīm sorry... but I only agree with you in the "word of mouth" thing.

Marketing is something I believe Blender lacks. If it had a good marketing plan, people would rave about itīs innovative non-standard UI instead of complaining about.. =D Again, see Silo and Modo who have good marketing and promote their 10-year-old "innovations".

The Blender Foundation probably knows that and Elephants Dream (Orange Project), http://www.blendernation.com and http://www.blenderartists.org are the beginning steps toward filling this gap. =D

Cheers

TroutMaskReplica
05-30-2006, 01:39 PM
people would rave about itīs innovative non-standard UI instead of complaining about.. =D Again, see Silo and Modo who have good marketing and promote their 10-year-old "innovations".

we do rave about it, or have raved about it. we just get tired of arguing with people on forums who've used blender for 10-30min before giving up, without completing a single tutorial.

how do i know these people haven't consulted the manual or the tutorial series? because all one needs to do to become familiar with the interface is complete one tutorial, called 'gus', available in the first pages of the docs or the tutorial series. it takes about 30 min.

so we have a situation where literally hordes of people who don't know anything about how blender works, coming onto forums to basically talk out of their ass.

one gets weary of it, and at the end of the day, i don't give a rats ass what software you use.

btw, i downloaded the demo of maya, and i couldn't make a beautiful picture in the first 20 minutes of using it, so i must conclude it's an absolute piece of garbage.

VirgilioVasconcelos
05-30-2006, 02:00 PM
we do rave about it, or have raved about it. we just get tired of arguing with people on forums who've used blender for 10-30min before giving up, without completing a single tutorial.

I totally agree with you, TroutMaskReplica. I do rave about it, also! =D

What I meant is that the raves are only "word-of-mouth". If, PDI/Dreamworks or Bluesky raves about it like they do for Maya, thatīs marketing!! Itīs obvious there are backstage actions involving money for this to happen, and itīs results to the mid/low-market (most of us!!) are huge! Itīs easier for Autodesk/Alias to offer advantages to this great studios in change to this advertising.

I have to make sure to you all that Iīm not comparing Blender to Maya as 3D apps, but the marketing strategies that their organizations have. The Blender Foundation, as far as I know, is a non-profit organ.

If something like the use of Blender in the Spider-Man animatics happens, is just because one have realized itīs great qualities, not for backstage action.

As I said on a previous post, where Iīve also included the link to the quickstart, itīs a shame complaining about itīs UI without even reading the manual! And this applies to all apps! =D

Cheers

VirgilioVasconcelos
05-30-2006, 02:30 PM
For all the ones who complain about... =D

RedSquirrel
05-30-2006, 03:37 PM
With regards to the SoC Projects there seems to be some nice developments for the future but to be honest they're not really needed right now, they are not necessities. I have used Sharpconstruct and its a nice tool but to get the detail you are looking for in your models it requires your mesh to have a high face count, it comes up with nice results but the high mesh takes away your systems resources.

I think a more logical path is for the blender team to look into "Micro poly displacement" (think Zbrush) which would required them to rewrite Blender Internal renderer or export to a renderer that can handle micro poly displacement. Not many Blender users can afford programs such as Maya or Max, but they could be able to afford Zbursh and to be able to import a zbrush model into blender and rendering it with all its detail would bring a nice smile to many beginners and Professionals alike. Not to mention the reduction in the memory requirement would be less to render model with "micro poly displacement" that it is to render a highly subdivided raw mesh.

There are other things that the blender community is always screaming about and requesting which I think are more import such as proper Normal mapping (blender only allows Normal mapping on flat surfaces, although there has been some rumours that this issue will soon be resolved), updating the Metaballs, adding Ngons, Voxels, caustics for blenders internal renderer and the list goes on and on.

Apoclypse
05-30-2006, 04:24 PM
I agree with the comment about there being more important things that need to be done for blender. Sharp construct ias great but so is esspresso which has a really nice intergrateds system. its s script so its not fast but it works most of the time. If sharp construct can intergarte half as well then e could see something special otherwise why bother, instead resources should be given to the esspresso guy. The renderer needs either a rewrite or it needs to be replaced. It output is bland and has that chalky feel to it that Max had with it internal renderer before they added different filters and crap in 3.x. I like the diretion its going though, that whole pytexture thing is great and could be very powerful. If they made a fully pluggable renderer with pytexture for shader writing and c plugin for features we could really see something special. I'm going to do some research on this to see how viable this is. I might not contribute code as that stuff is way out of my league, but I definately would be interested in maybe creating a document outlining stuff, if not for anyone else at least for myslef. However the whole render intergration backend needs to be redone though. That way the render can be put into a seperate process or commandline option ala maya. If the renderer were written correctly it could pipe its output directly to blender through a diplay plugin . Renderman should definately be a source of inspiration.

Papa Lazarou
05-30-2006, 04:35 PM
we just get tired of arguing with people on forums who've used blender for 10-30min before giving up, without completing a single tutorial.



So everyone who has issues with the UI can just be dismissed as a clueless newbie?

Zarf
05-30-2006, 08:35 PM
Well for starters, some basic things like the lack of standard edit menu functions next to the file menu. That's fairly standard.


This seems fairly minor to me, but I took a look at what the 'Edit' menu for C4D, Silo and Motionbuilder all contained. The set of features 'common' to all those would be:

-Undo
-Redo
-Delete
-Select All
-Deselect All

A couple of things about this. The 'file' menu now currently resides in the header of the 'Info' Space type (or window). This is 'ok' since the Info Space type deals with the 'Global' enviorment within Blender. Undo and Redo are 'global' functions as well, however Delete, Select All and Deselect All are all functions whose scope is 'local' to the data being visualized in a specific Space type. To understand the problems inherent in this, imagine that I have a Blender screen with two 3d views, an F-Curve (IPO) Editor and NLA window open. Now if this 'Edit' menu were added right next fo the 'File' menu and I decided to choose 'Delete', 'Select All' or 'Deselect All' from it, which Space or 'window' would it operate on?

For applications like Cinema 4D, the main 'edit menu' just 'assumes' that you mean for the select/delete operation to be performed on data in the 3d viewport, while other window types (such as the timeline) have their own seperate 'Edit' menus with 'delete/selec' operations in them. So it really seems that Blenders approach in this case makes more logical sense, but the approach that other applications like C4D take have the weight of 'convention' behind them. Which is more important, logical consistency or farmilarity? Hopefully the event system refactor will allow these sorts of issues to be adressed with custom menu layouts.

Another one would be using left-mouse-click to select objects rather than to position the '3d cursor'. While I'm sure some blender users find the 3d cursor quite useful, I fail to see that it is of such importance that it requires the left-mouse button at the expense of other functions.

While I'm not going to argue with you about the usefullness of having an easy to use auxilarlly pivot/transform point built into the UI of the app, I would point out that you can switch the behavior of the mouse buttons in the info space (window) so that left mouse selects and right mouse positions the 3d cursor.


we do rave about it, or have raved about it. we just get tired of arguing with people on forums who've used blender for 10-30min before giving up, without completing a single tutorial.

how do i know these people haven't consulted the manual or the tutorial series? because all one needs to do to become familiar with the interface is complete one tutorial, called 'gus', available in the first pages of the docs or the tutorial series. it takes about 30 min.

so we have a situation where literally hordes of people who don't know anything about how blender works, coming onto forums to basically talk out of their ass.


I hasten to point out that many Blender users, over the past 8 years, have been very obnoxious in how they 'promote' Blender. It's not uncommon to see Blender users saying things like 'Who needs to spend a lot of money on software like maya/max when you can get blender for free? You suckers are just wasting your time with crappy closed source software.'

It's very rare to see a Blender user actually defend the application/interface in a reasoned and intelligent manner, and therefore these peoples 'opinions' are rightfully ignored.


With regards to the SoC Projects there seems to be some nice developments for the future but to be honest they're not really needed right now, they are not necessities. I have used Sharpconstruct and its a nice tool but to get the detail you are looking for in your models it requires your mesh to have a high face count, it comes up with nice results but the high mesh takes away your systems resources.

Part of the SOC project proposal of the SharpConstruct author was that the Blender integration would include multi-resolution sculpting, similar to Zbrush, so face count won't be as much of an issue.


I think a more logical path is for the blender team to look into "Micro poly displacement" (think Zbrush)


I don't see why its 'more logical' since Micro-Poly displacment is a seperate issue from the sharpconstruct integration. It's not an 'either/or' proposition, and the idea is to eventually have both. However you have to understand that for an Open Source project its not as simple as just 'assigning' a task to someone, and people tend to work on what they need/want first unless money is involved.

BTW, AFAIK Zbrushes renderer does *not* use micropoly displacment at all (Aurick from Pixologic explained this on Zbrushcentral's forums a long while ago). REYES based renderers (most renderman ones, including opensource ones like Aqsis) can however.


which would required them to rewrite Blender Internal renderer or export to a renderer


A full rewrite of the internal renderer may or may not be nessecary. Ton did a lot of work on it for 'Elephants Dream' that may have made it easier to implement micro-poly displacement.


that can handle micro poly displacement. Not many Blender users can afford programs such as Maya or Max, but they could be able to afford Zbursh and to be able to import a


A large number of Blender users can't afford Zbrush either.


There are other things that the blender community is always screaming about and requesting which I think are more import such as proper Normal mapping (blender only allows Normal mapping on flat surfaces, although there has been some rumours that this issue will soon be resolved), updating the Metaballs, adding Ngons, Voxels, caustics for blenders internal renderer and the list goes on and on.


If the Blender community wants something bad enough, they can start coughing up the money to drive development in the direction they want, or start coding things themselves. You can't complain too much when your not paying for a product that other people spend their free time developing.


So everyone who has issues with the UI can just be dismissed as a clueless newbie?


Absolutley not. Blender really needs well thought out constructive criticism in order to grow. Criticism that isn't constructive however, should and will be ignored.

LetterRip
05-30-2006, 10:27 PM
Tangent space normal mapping there is a private patch for, which I hope/expect to be in 2.42. Micropoly displacement - it was a wishlist for Orange, but there wasn't enough time. I suspect that the next movie project will have it as a wishlist (and given that in general the ultra critical needs were all addressed in Orange, other more 'wouldn't it be nice' wishes can be addressed in the next movie), it would especially make sense given that the tangent normal maps and sculpting will be in.

I agree with zarf that we tend to have too many advocates that don't represent Blender in a very productive and persuasive manner. I think that is improving though.

LetterRip

VirgilioVasconcelos
05-30-2006, 10:58 PM
I hasten to point out that many Blender users, over the past 8 years, have been very obnoxious in how they 'promote' Blender. It's not uncommon to see Blender users saying things like 'Who needs to spend a lot of money on software like maya/max when you can get blender for free? You suckers are just wasting your time with crappy closed source software.'

It's very rare to see a Blender user actually defend the application/interface in a reasoned and intelligent manner, and therefore these peoples 'opinions' are rightfully ignored.




I agree with zarf that we tend to have too many advocates that don't represent Blender in a very productive and persuasive manner. I think that is improving though.

I agree with both of you on this. I only should add that this may happen because of the passion involved. They want to fight for their beliefs. Most of them donīt know much what theyīre talking about and thus donīt have consistent arguments, but want to say something like: "hey, cīmon! donīt talk like that about my favorite app! I like it! It works for me!"

Thatīs not only for blender users. Maya ones. Max ones. Christians. Muslims and so on.

If we had a religionsociety.org forum, I have no doubts this would happen the same way.

Before criticize, please, try to study and know better what youīre talking about.

Cheers

RedSquirrel
05-30-2006, 11:28 PM
Originally Posted by Zarf

I don't see why its 'more logical' since Micro-Poly displacment is a seperate issue from the sharpconstruct integration. It's not an 'either/or' proposition, and the idea is to eventually have both. However you have to understand that for an Open Source project its not as simple as just 'assigning' a task to someone, and people tend to work on what they need/want first unless money is involved.


I don't consider it a separate issue due to the fact that the end result will be rendered on screen. Micropoly displacement mapping excels with complex scenes the amount of detail that can be added with this method is astonishing. Multi Resolution meshes is similar in a way to how the decimate option works in blender (of course its a more advanced version of the decimate function) it just tries to keep the detail high by reducing the face count in some areas without it being noticeable. I can see the advantages of multi resolution meshes on static object but what about an animated character I will be interested to see how it reacts. So its not really similar to zbrush is it. All and all I don't really see the need for micro poly displacement or multi resolution sculpting at this moment in time, more importantly its good to hear that there is patch coming for Normal mapping.

A large number of Blender users can't afford Zbrush either.


lol, if you want to nitpick maybe I should of said they are more likely to afford zbrush than Max or Maya. :D

KayosIII
05-30-2006, 11:39 PM
With regards to the SoC Projects there seems to be some nice developments for the future but to be honest they're not really needed right now, they are not necessities. I have used Sharpconstruct and its a nice tool but to get the detail you are looking for in your models it requires your mesh to have a high face count, it comes up with nice results but the high mesh takes away your systems resources.

[QUOTE=RedSquirrel] I think a more logical path is for the blender team to look into "Micro poly displacement" (think Zbrush) which would required them to rewrite Blender Internal renderer or export to a renderer that can handle micro poly displacement.

Micro Poly displacement is a method for subdividing a mesh until each poly is less than a pixel in size thus guaranteeing a perfectly smooth subdivision surface based model at render time. It is most commonly found in Reyes based renderers. As far as I know it is currently used only as a rendering technique.

Not many Blender users can afford programs such as Maya or Max, but they could be able to afford Zbursh and to be able to import a zbrush model into blender and rendering it with all its detail would bring a nice smile to many beginners and Professionals alike.

ZBrush is still about $800 local currency (about the same as XSI) and as far as I know it is not available for all platforms that Blender runs under. (What are we Linux folks supposed to do for instance - appart from wait until we can get zBrush running under Wine, or spend a couple more hundred on a windows license (shudder) )...

There are other things that the blender community is always screaming about and requesting which I think are more import such as proper Normal mapping (blender only allows Normal mapping on flat surfaces, although there has been some rumours that this issue will soon be resolved), updating the Metaballs, adding Ngons, Voxels, caustics for blenders internal renderer and the list goes on and on.

Such is the nature of SoC and Open Source coding. People pick things they want to work on the SoC people choose the things they like most (I think choosing the projects which were most likely to get finished is what is going on). The normal map thing should be sorted by 2.42, Ngons are in the pipeline - via a new mesh type - not sure how this will be intergrated into blender.


What is being proposed though is more than just sharpconstruct being bolted into blender. The Author is proposing to use what they have learnt with sharpconstruct to build a similar tool inside of blender (the current tools are python scripts, this would be c code and potentially run much faster). The proposal would also include Multi Resolution Meshes - which I am personally very exited about.

Zarf
05-30-2006, 11:46 PM
I don't consider it a separate issue due to the fact that the end result will be rendered on screen.

I will admit its a related issue, but its still not the *same issue*, therefore in my mind its a seperate issue.

The fact of the matter is that micropoly displacment during rendertime will allow Blender users to render models with high quality displacment maps but will do *nothing* to help them create them in the *first place*. Multi-resolution sculpting tools will allow them create very detailed and high quality models that they can then extract displacment+normal maps from (there are several rather good free tools for this step already).


Micropoly displacement mapping excels with complex scenes the amount of detail that can be added with this method is astonishing.


Micropoly 'dicing' also allows 'perfect' tesselation of higher order surfaces such as subdivision surfaces and NURBS as well. Theres other benefits as well, but its all very basic stuff thats been covered in detail in texts about the subject for a long time now.

Multi Resolution meshes is similar in a way to how the decimate option works in blender (of course its a more advanced version of the decimate function)


Blender's decimate function and multi-resolution subdivision surfaces are in no way 'the same' from a theoretical standpoint, a coding standpoint or even from an end users standpoint.


I can see the advantages of multi resolution meshes on static object but what about an animated character I will be interested to see how it reacts. So its not really similar to zbrush is it.


The multi-resolution sculpting will be 'similar to z-brush' in that it will behave roughly in the same manner as Zbrush's multiresoltion subdivision surfaces do. This is all outlined in the SOC proposal.


All and all I don't really see the need for micro poly displacement or multi resolution sculpting at this moment in time, more importantly its good to hear that there is patch coming for Normal mapping.


I can't say that I agree with you, but it dosn't really matter what either of us say since it will get worked on regardless because thats what people are interested in doing.


Cheers,
Xarf

Nichod
05-31-2006, 02:32 AM
Its sad this got SO off topic.

Nichod
05-31-2006, 02:37 AM
By the way. Along with the new sky. It would be pretty nice to see an intergrated:

http://innerworld.sourceforge.net/index.html

JA-forreal
05-31-2006, 03:14 AM
I never thought too much about sky generation inside Blender, but I could use it. SharpConstruct has been a great app for applying detail to meshes as it is. Now that it's going to be integrated inside Blender that prospect opens up many new options for detailing meshes.

Blenders Modifier Stack is already a very handy asset. I hope that in the future Blenders modifiers offer more parametric functionality between each level. I could see how Blenders new modeling features could be enhanced with improved shape guiding modifiers, etc. Soft-selection modifiers would really be useful when applied to vertex weight groups. Advanced armature modifiers will further the power of character animations.

The array modifier is a favorite of mine now. I have used it for spiral stairs and plant model generation so far. It's going to be nice to see any updates to the array modifier.

I wonder how far some of these SOC features will be developed. I think that the important thing is to get the code working at some usable level first. Once a features in it's in. I'm still just beginning to better explore last years SOC feature for water simulation.