Well what spurred this thread on was Madcow’s report on Jaguarandi’s raytracer being slower than vanilla 2.5, but now I tried it out and it was ultra fast for me.
Oh, that’s right…blame me!
I seem to have a knack for stirring up CD’s paranoia nest.
Thanks Ideasman - I was beginning to think I would be alone in my viewpoint 
Seriously, we can trust the developers to do what they want to do. That is, a version of Blender that is stable, efficient, and has the features they want to use. I am simply grateful that they were sick enough of us complaining about the inability to change shortcut keys, the inflexibility of mouse usage, and the cluttering of the UI that they decided an interface overhaul was necessary.
Which goes to show that those telling us (for years) not complain about the interface because others had already and the devs weren’t going to change it need to realise that persistence pays off.
Don’t thank fanboys, thank those who persist in clamoring for change for the better to the risk of being flamed. Fanboys are a plague for change. All they can say is yes, yes, yes, everything is perfect, nothing could go wrong. Fanboys should realize that developers are big and adult enough to accept criticisms.
Interesting thought.
Any change involves risk, and 2.5 is a HUGE change, so naturally CD is right about some risk potential. Caution can definitely be warranted.
But compare that to the risk of standing still, or of being afraid of changing too much, or trying something to hard. That’s death in 3D software, or most anything else.
w00t f0r bl3nd3r d3vs
I don’t think is due to lot of people demanding stuff. That can be a part of it… Maybe…But I would not bet. Also, you can rarely do a project adjusting it to every other idea popping in forums… I guess is perhaps their own study, plus maybe some blender aces in close connection with them, and open movies are for sure one of the best things for that…Nothing like real cases, imo. I learnt that the hard way… However they do, which I don’t know, it works. (and man, there are so few things in this world that actually work… ) And so many years been working. I’m no open source fanboy by any mean, and still trust that core devers group more than any 3d software company I can think of. (while had always a tendency to learn and use (3ds max sort of geek here) comercial software (job requirements) .Imo, that says quite… ) PD: As I see it, a 3d tool only survival way is go forward at crazy pace, given how the projects requirements and other tools evolve so quickly… The thing is blender evolves super fast, but I see not the weird errors, and bad management produced flaws I clearly see in other 3d apps, or, in general, in software companies (I work in one) , usually driven for pressures and some dudes not very good at what they should do…
Any change involves risk, and 2.5 is a HUGE change, so naturally CD is right about some risk potential. Caution can definitely be warranted.
But compare that to the risk of standing still, or of being afraid of changing too much, or trying something to hard. That’s death in 3D software, or most anything else.
Wise words!
Being a betting man, I would put money that the reason for a large portion of the interface changes were due to the large number of people who continually complained about the existing interface.
Look through the history of the forums & mailing lists and the consistency of the complaints in reviews by respected artists outside the Blender community. The inflexibility & counter-intuitiveness of the interface would have to rate as the issues that have dogged Blender throughout the years.
There are many other changes planned that are probably not to do with the level of complaints, but given the history and Ton’s “jokes” in regards to the interface changes - I would put the reason down to pressure from complaints for this set of functionality.
That wasn’t the primary reason.
Martin
Something I think is not taken into account is one of the reasons for 2.5 was to make the code less messy and have clear separation between data, tool api, interactive tools etc.
Blender2.5 isn’t one big update, Its multiple updates in different areas + code cleanup.
Everyone talks about the interface because they are so shallow ;), or maybe because its what you first see, but most of the updates in 2.5 don’t rely on interface code.
Could break down changes into…
- GUI (python interface & new GUI API)
- Event System, Multiple Windows (GHOST api)
- Operator api/design
- RNA data API
- Python API
- Other Changes/Improvements smoke, volumetrics, particles etc.
So lets say one of these areas end up really sucking and the developer totally fails to make an improvement (or screws up the code - unlikely).
What happens? - its re-written/fixed in under a month I’d guess… and since we still have the 2.4x code maintained it can be used again and modified from that point if the 2.5 code is that bad.
See the risk is divided so that any failure is fairly local to each module… It turns out that really this isn’t so bad even if we make mistakes.
I started using Blender a long time ago like in late 2002 soon after it became open source and gave it up for a commercial 3D application. I didn’t like Blender, and getting things to render the way I wanted them took too much work and knowledge. But in 2005 I tried it again and I loved it… Blender had not changed much (well it changed a lot but the feel of it was still the same), but I changed over those years. The user community grew and with that growth came an amazing source of knowledge, techniques and even support that I have never found anywhere I didn’t have to pay for.
The current Blender code looks like too many parts rely on too many other parts which makes development take longer, produce more errors and results in slower software. It looks like they are putting in a huge amount of work to get make the code more modular, but I think it will result in the developers being able to develop new features and updates much faster than before.
Blender has become much more than just a 3D modeling, animation, game engine and rendering application, thanks to the users and their willingness to help others and share their knowledge, Blender is also a community of extremely talented and skilled 3D artists. Which I think is a feature that far surpasses any other product. The point I am trying to make here is that even if Blender changes so drastically that most current users will be left trying to relearn the software, there is an overwhelming amount of support available to make the transition almost unnoticeable.
I think I ended up blabbing too much so I will just say that: as long as the developers remain as amazing and focused as they have always been, I really don’t think there is any risk.
yeah, lately I’ve been thinking a lot along the same lines,
in the sense that these day’s blender isn’t so much a 3D specialist application anymore,
as rather a general all around graphics & effects ‘workstation’ (+ game engine).
I’m willing to be wrong, mate 
Can you tell us why it is being changed now then when it was outright rejected as unnecessary and ill-advised before? Serious question as said position was made quite clear to me and others whenever interface change along these lines was suggested.
The primary reason is that some people were both unsatisfied with the current situation and willing to do something about it (write complete and legible proposals, evaluate different designs, code them or convince other people to code them).
Complaining by itself doesn’t achieve anything, people willing to work it out do.
I wish people would also take that as a call for participation. There’s a lot that can be done, even if you can’t code. (and not all of it is writing documentations )
Another good reason is that core modifications in the Blender code made it easier to do that kind of UI changes now than they would have been before.
Martin
Of course, people will not be willing to admit to anything. Giving credit to complainers doesn’t seem right. And if some think that complaining about the UI is shallow then they don’t belong in the software business. Ease of use is a feature. That’s the bottom line. Imagine if 3ds Max or Maya are free. And why does Newtek is working on Core now?
complaining ONLY about the interface is shallow
dont forget to bitch about format support, communicating context to the user, editing settings for multiple objects at once, operators crashing when out of context, macros not being usable, python-UI-scripts running an insane number of times and the python api missing loads of stuff.
Theres no need to fixate on one thing!
I can only speculate but here it goes:
They didn’t want to spend all that time just to change the UI from how a lot of the user community liked into something that people that are used to other software packages liked it. Also changing the old code to make the UI customizable would have been a very time consuming error prone feat. They had changed several small things about the UI, but making it almost fully configurable all at once would have been too much work for too little advancement.
However changing the underlying code to make development for everything faster and less error prone is something that the whole application can benefit from. So the UI was not the only reason, but because of the other reasons it is now easier and safer to upgrade the UI.
Edit: I also think that cleaning up the code and making it more modular will lessen the learning curve for new developers to come in and help out because they won’t need to have a lot of extra knowledge about how the part they are working on affects other parts of the application.
you all realise 2.5 has been in development for nearly 2 years and has been designed for a year before, the main reason it wasn’t done earlier is because it was impossible without the rewrite required and all the developers had bigger things on their mind
thank your lucky stars that it is here now
The Blender devs are my lucky stars! Thankyou each and every one.
Ha ha ha! Ideasman for the win!
Yeah, and Blendix, Mfoxdogg, JesterKing, Broken, Theeth, …
There’s no need to fixate on one coder! 