The important is that only when it will be usable it will replace edit mesh(which it’s stable,and it’s fast),the community is giving a lot of pressure about this stuff and it’s a wrong attitude.
Personally I am not concerned with how fast it is at the moment or whether it will make it into 2.50 or not. Eventually it will be released, we all just have to be patient. What I am happy about is that it is been actively worked on there was a period there when things sort of languished with very few commits. But now development has picked up. Thanks to all the devs
Thanks for the tour
please for those who haven’t tried ngon or have not used it. refrain from commenting that ngon is bad, ngon is slow, ngon is this or that.
cause you can’t make an objective comment if you have no idea what you are talking about.
seems to me that those people who are complaining are the usual type of people who fear change.
another thing, if you find that you cannot model just because ngon is introduced that means you haven’t really learned how to model yet.
you should not have any mental problem whether you have ngon or not.
It’s a really offensive post,maybe you should use your time better.
I think that´s fairly obvious, yes? The wrong attitude is bitching about something that isn´t finished yet. If you read SVN logs, you would know that Joeedh has started working on optimization, so I wouldn´t worry…
http://lists.blender.org/pipermail/bf-blender-cvs/2009-September/022211.html
http://lists.blender.org/pipermail/bf-blender-cvs/2009-September/022220.html
http://lists.blender.org/pipermail/bf-blender-cvs/2009-September/022221.html
I don’t agree much.
I try to explain,because is really important that people understand 2 things;
1)I don’t have NOTHING against bmesh
2)I don’t have NOTHING against ngons(I use other programs in every day work with ngons)
But I like also the current blender modelling tools,for performance reason.
I follow blender development since 1998/99,not always new things are tested and bugfixed in depth before replacing old code.
If the code is an option it’s really doesn’t interest me if it isn’t finished,but if is a replacements for something that works(with limitations of course) ,well,It should be in theory as good as the old one,with more/better stuff.
When I say"wrong attitude" I mean simply that people here have always a fan boy attitude,but few people used really the stuff coded,so often there are bugs and problems (Without a right feedback coders cannot improve or fix if something goes wrong,ant that is the weak point.)
I have simply given my feedback(and I do that when I have the time),based on my meshes(something a bit more complex than suzanne).
And my point is that something as integral as the mesh code is going to be well tested before it gets into Blender, it´s not up to any fanboys (or even thickheaded old blender zealots), and it´s quite arrogant to put it down already considering the state it is in now, when Joe has specifically stated that it´s not ready for testing. Until it is, you and Endi can go play the bitching game somewhere else…
So THATS why I couldnt copy certain topology Id seen in other apps.
I have the right to test when I want/have time(joeedh do coding when it has time,no?),why there are the builds on graphicall.org if someone can’t test?
It’s a joke?
I’ll stop to help with testing so you can be happy.
it is not testing or suggesting improvements that is the issue.
the issue is some people saying that we should not move forward with this because it is going to throw off their favorite workflow.
anyway. once this is implemented i can see that you can really let your imagination fly.
a lot of people are going to produce really awesome meshes and they will be able to do it a whole lot faster and easier.
Stop twisting words. Of course you can test the branch, but it´s not ready for serious testing, because some functions aren´t added yet, and there are still plenty of bugfixing and optimization to do. You complained about speed - well there hadnt been any optimizations yet, because it´s to early - go figure.
I really don’t think renderdemon is one of those people … Bmesh branch is very useful for sure, but it’s a delicate project: it introduces new tools for modeling, but such basic tools have to be totally bug free and optimized because they have to replace the fast and solid editmesh.
So, I totally agree with rendermon: I hope bmesh build will remain a separate testing branch for a while, letting users to give important feedback to devs.
It seems like many important and useful threads like this are going to become the bitchin’ court of some unpolite fanboy… such a shame.
How about hearing from this fanboy:
The branch still isn’t testable, mostly because some editmesh tools remain. The code that bridges editmesh and BMesh continuously breaks; and while I’m going to try to get it to at least somewhat work, I don’t have much faith in it at this point.
That´s from Joeedh´s latest post on the bmesh blog. Is that enough for you? If I come across as impolite it´s because of people start jumping to conclusions before they have their facts straight.
if you want a software that is totally bug free before release, man, good luck.
cause i say you are going to wait forever.
name me a single released software in any industry that is totally bug free.
Regarding performance issues with B-Mesh:
As Mfoxdogg pointed out, first things need to be working before they can be optimized (its not possible to profile code that doesn’t exist/work yet!) There are all kinds of things that can be done to make things faster. B-Mesh may not even be the culprit per-se, especially when it comes to modifiers since right now there is a lot of redundant conversions going on in the modifier stack that can be removed if they prove to be a bottleneck.
That being said there are always trade-offs to be made. For instance already B-Mesh has a bigger memory footprint than Edit-Mesh, but this was considered nessecary in order to provide the features that Edit-Mesh was lacking. At the end of the day it requires feedback from users in order to determine what is most important.
Renderdemon:
If you could provide examples of models that are peforming poorly under the bmesh branch along with instructions detailing which operations are an issue that would help greatly in Joeedh’s profiling efforts probably.
Cheers,
Briggs
@Vainquer: maybe you misanderstood my post…
I’m pointing out posts like this… probably I am wrong, but I think it’s totally OT, don’t you? :rolleyes:
Now, let’s help joeedh with constructive and useful feedbacks on his awesome work.
Cheers.
It’s hard to tell who the fanboys here are. On one hand are these dudes who are not pushing for Bmesh and would still prefer Editmesh. On the other are just folks who say Bmesh is still very much a WIP. Go figure.
well. it is not my problem mr sideshow bob if you are onion skinned.
Im just being candid and Im trying to point out what is obvious.
now am i a fanboy.
let me just say that i happen to use another 3d modeler that uses Ngon and I am just wishing that blender has it too.
because i just saw how easy it is if blender has it.
is that being a fanboy?
Well, no - but it doesn´t matter. Anyway I´m been totally off my game today, and I´m sorry if I took it out on you guys…
I’ve always thought any hint of resisting change/s is considered fanboyism. Which means if you’re against pushing Bmesh in one way or the other you’re the fanboy, not the other way around.
Personally, I won’t even use 2.5 for serious work other than for testing. Complaining about half-baked features at this point seems immature, if not ridiculous and stupid. I’d use it when the Durian project is finally completed. That would still be about a year or so. Obviously, that’s exactly what the BF envisioned it to be – use the Durian project in conjunction with the development of Blender 2.5. No Durian, no new pro Blender to brag about.
This is entirely different from the previous projects in which features were just being added. This time it’s about UI enhancements as well as jampacking features to make it production-ready.
I don’t see anything wrong with using the current version (2.49b) in the meantime. There is so much to learn. Books about it are still very relevant. Also, the concern about not being able to translate all that knowledge into 2.5 is baseless. The new Blender still maintains the same basic workflow. It’s just a matter of finding where the original buttons went. (Just watching Pablo Vasquez use Blender in his Venom’s Lab videos makes my little Blender knowledge seems even smaller.) :yes::RocknRoll: