An argument for re-introducing Blender Internal in version 2.81 (or, 2.80 final)


(Ace Dragon) #41

Also to throw back in.

  • Interlaced video output
  • Ye-olde editmesh as a second mesh type
  • Subsurf global to the mesh and not via the modifiers
  • Multires that is permanent and locks the mesh into its current form (if you want to keep your sculpt)
  • The OOPS schematic
  • The audio window
  • Radiosity
  • Horizontal panels, complete with uneven lengths and the other quirks

It seems to make sense, Blender can’t be like the big apps. if it’s not bloated and heavy like the big apps. :blush:


(Mozalbete) #42

Yes, makes complete sense. the BF should spend a large amount of money to reimplement a dated, bad renderer so that you can remake a couple of shots for a legacy project while having some new features. It’s totally not insane, because blender is made for you. Why make useful changes? Why implement useful features and fix bugs if we can spend a gorillion dollars so that your legacy proejcts made with a crappy renderer do not feel “split”?


(Osares) #43

I think this is the more realistic approach, reintroducing BI in Blender is very unlikely even if someone actually wrote the code to support it (maintainance cost also need consideration!). While it would be great to support legacy features i doubt it’s good for the future development of Blender. Forcing people to find new ways to solve a problem is unfortunate but, as mentioned, old versions will still be usable when no other option remain.


(Lsscpp) #44

You politely asked something that - if you follow a bit the new from Blender developement - you should know is not doable, has drawbacks, and for the effort it implies has no logic to spend time on. It is also worse for any workflow nowadays, except for those old projects people still have opened here and there (indeed not a “nowadays” workflow). And it’s all a kindness from Blender Foundation to provide all Blender official builds way back to 20 years old ones, for people in need.
So again, this thread makes little sense to me, I take can’t take it seriously, but more like a whining in loud voice


(mavek) #45

Can we keep this cordial? I think the request comes from a reasonable place but the reality is that it is not feasible.

It is great to see blender being used in the commercial space and i think it amazing to see it been used on TV shows and even movies. As it goes forward the dynamic nature of an open tool like blender is going to butt up against the practical realities of a commercial work flow (thankfully this is not automotive you got to keep that working solutions going 10 years, Aerospace 20+)

To foster further adoption I think a less combative attitude is required. My preferred approach, is if the reintegration of BI is not forthcoming, due to time and cost, is there an alternate path to satisfy some of the concerns of the OP and other people in the same boat. Automating the migration of textures, light groups, edge pass etc. Are some areas, is there work there that can be done? What other areas?

As to the OP, if lack of access to BI is said to cost you money, would you be willing to put some or all of the difference of costs towards a solution? I think any one who is using blender to make things that are earning them money should consider plowing some small portion of what they are saving on licences to fostering the tool you are using, to improve it, and then maybe you might be better placed to suggest and drive desired improvements.


(sundialsvc4) #46

Thank you, Mavek – we all insist on “civility” in what I brought forth as strictly a business discussion. Let’s all keep it this way. We’re discussing technology, that is all. Technology, and the very-real business of a not-for-profit tool that doesn’t cost anyone anything.

As I’ve said: I think that it is an incorrect business decision for Blender to completely abandon a rendering technology, especially one that it has had since its inception. I tried to offer a pure-business case for it.

Yes, I know that BI and Blender were once Siamese twins – joined at the hip – but the effort to cleanly remove it also has the benefit of isolating its various pieces, so that it can now be “re-imagined” and return to its rightful place as one of the several rendering options that Blender offers. Better(!) than ever it was. (Which is certainly something that all of its users want.) I argue that it is still valuable; still used; still needed. Not merely “as it was,” but, “better.”

I’m not dreaming of the past or “pinin’ for the fjords” when I politely make this argument. I propose it as a business case that is of impact to me, and to many other people like me. I’ve expressed what that impact is.

I don’t have the money to help such a very-large effort, nor do I have the expertise in the innards of this massively complex software system. Instead, I’m stating my case to the managers of the people who do.

And, having stated it (several times now), “I’m done.” Thanks for listening.


(Lsscpp) #47

Reproaching myself for sounding too rude here. Anyway what I wrote is not to be taken personal, it’s definitely not an attack to you and your ideas.
What I failed to communicate is that having old builds always available publicy is already a great value for people who rely on old technologies (including harware btw), and since resources / coding complexity needs a tradeoff, I think that you can accept the effort of splitting in two: old blender for old jobs, newer for new.

Noneless all of the above can be read as Eevee :wink:
We still miss stuff like light/layers linking, but hey, it’s beta


(Matt) #48

Like all tech and software, eventually support will run out to make way for advances. Supporting software indefinitely doesn’t make any sense and it shouldn’t be expected. Especially from FREE software with limited funding. In business you must foresee these risks ahead of time and adjust or invest accordingly.

It is silly to expect the Blender devs to pull resources away from Eevee and Cycles to produce effectively a new BI renderer with full compatibility. Also for free… If that is a requirement for you, you must invest in producing your own renderer. (Which can potentially become a source of income) Or switch to Eevee, Cycles, or even an external renderer. For now stick to 2.79 until you find a path forward and begin the migration process. You may need to build a new team to handle the migration and training.


(Zinogg) #49

i don’t think they have the man power to bring an old tech back in ,especially if it used an old version of opengl…etc, it’s best for them to work on the 3 main engines…the only possibilty is if someone is willing to volunteer and do it as an addon, and yet it will not be easy or as polished.


(Matt) #50

Agree, its ridiculous to ask a company to continue to support 25 year old software which has had a full replacement out in the wild for years. Heeey Microsoft, can you bring windows 98 support back? My business DEPENDS on win98 and Ive been to lazy to update all these years. Please support it for free. kthxbyyyeee

Many game engines can produce better results than BI in real time, just like Eevee. Time to embrace the new and stop waiting hoping the inevitable never arrives, which now it has. Maybe some of the transition time can be made back in reduced render times and a more productive workflow…


(apoclypse) #51

This is wrong. It does cost someone something. It costs the BF money to maintain and implement code, it costs them resources they could have better spent elsewhere. It costs the larger community features that they want and progress for those features. There is always a cost related to everything. It may not cost you anything but if you are truly running a business you should know that nothing is free.

Your proposal shows this. You want the BI back in Blender because it costs you effort and time to import your project back to 2.79 from 2.8. Now here is the issue for you. You are making it out to be a business case, that’s all well and good but you have no data to back up your claims that there are a significant amount of people who need this. If you provide evidence and data showing that the BI is something that a large amount of users want then the case can be made to put it back in or add it in as an option. Even so the BF has to weight their resources, time and money to see if it’s even worth the effort.


(stargeizer) #52

There was once a time an external renderer based on Blender internal. It was called Microrender and was based on Blender internal. Probably you will have better luck asking the original coder for a copy of it (was commercial) or asking somebody if they have a copy and the python script to export to it. (http://www.msbsoftware.it/microrender/)

Is FAR (by years light) more realistic to find a python coder to adapt the script to 2.8x than to reintroduce the legacy blender internal back into 2.8x.


(Harley) #53

workflow


(motorsep) #54

The only things I’d need from BI is texture baking modes. Probably make sense to add them to Cycles than re-introduce BI to Blender 2.8x


(Unreal3DFX) #55

Stay in 2.79 that has ancient BI renderer, 2.8 is future and meant for next generation who can adapt and embrace much better render engines.


(Matt) #56

If you are standing on train tracks and you hear a train 3 miles away, do you get off the tracks asap or wait until the last second risking death?

OP chose the later with his company and has put his entire business at risk, companies across different industries often must invest a lot of money in constant software upgrades, hardware upgrades, workflow changes…etc. It’s to bad OP had no foresight in this case with software his company has saved an incredible amount of money on already. Don’t have someone knowledgeable to build a 2.8 version of BI? Hire someone who does and pay them well. Or be satisfied in being “left behind” in 2.79 (which is OK, 2.79 is still a great free piece of software) for the time being until you have a solution going forward. If OPs customers expect tech advances maybe they should find another studio willing to adapt newer technologies at a faster pace.

But to ask the BF to undertake a huge project at the communities cost while inferring a large % of the community wants BI back without hard evidence seems extremely selfish. This is why this thread seems a little heated, because it should be. I want BF to focus on Cycles and Eevee, and making 2.8 the best it can be, I’m sure the super majority of the community agrees with that stance. Feel free to make a pole, if I am wrong I’ll eat my words! Blender hasn’t had a huge growth lately because of freaking BI, most of us haven’t touched BI in years…


(Craig Jones) #57

I thought that the missing baking functions were added to 2.8 already for Cycles…


(tyrant monkey) #58

I am at a loss to understand this thread if you are not planning on making use of one of the bigger new features in Blender 2.8 why even migrate to 2.8 why not stick with Blender 2.79 and have BI. It’s a solid release of Blender and I am sure you could easily use it for the next four to five years or who ever long you want to take to migrate.


(Dorro) #59

Hire a skilled coder with the money you save using free software and integrate 2.79 into your new 2.8 workflow. You could run 2.79 as external render engine on a server and never even have to open 2.79 again. Your full time coder would get a job, be able to add features you want to the 2.8 codebase, give these contributions back to the community, which would improve your standing within the community and improve your influence on the direction of 3.0…

Win, win, win, win etc…


(Yassir) #60

So you’re just hoping that they support BI temporarily to support you and people that currently depend on BI and then once you don’t need BI, you are going to fully support Blender’s future once again?
Then what if the next time you’re done with BI, there are people that have similar business case and need the dev to support BI more? Isn’t this going to be an endless path?

I am not a developer thus I believed the devs when they said they need to deprecate BI in order to focus the resources for Blender to be more expandable, efficient, powerful, etc. They need 3+ years of work to make Eevee to what it is now, and it also needed them to rework, recode, redesign the core foundation of how Blender works to make it happen: the collection system, copy-on-write, etc. And I believe even if the dev would agree to integrate BI in 2.8+, it wouldn’t be just a month of work for them, maybe a year or more.
Now based on your business, do you still need BI in 2020? If not, then just let it go. 2.80 stable is scheduled to be delivered in April 2019, but It certainly needs bug fixes as more people would likely to use it by then. And the usual 2.80a and 2.80b with the bug fixes, I assume, will need more months to be delivered. So let’s say the bug-free 2.80 will be delivered in the Q4 of 2019. It means you won’t even get the perfect 2.80 after at least 9 months, and then let’s say they try to integrate BI for more months, and we still need more bug fixes, and that means more months, and so on…

That means, by the time they finish integrating BI in 2.80, you supposed to deliver many works done in 2.79 to your client, and then BI is integrated and you no longer need them as Cycles or Eevee can do what you needed in BI, as they are optimized more in those months.