Personally, I would rather not want to go into the (rather sticky) discussion of the creation of a fork, as it’s not like seeing positive changes to Blender policy in terms of user interaction is completely unfixable.
So far, it seems the community as a whole is leaning towards simply revising the BDFL model to include more checks on the dictator’s decision making and veto power while bringing in a greater interaction with the community (with the BDFL himself still having what’s arguably the most important role in terms of development planning, overall development, finances, and production), sounds good to me and avoids the pitfalls of a true, community driven model.
Well, talk of a fork is just that. The key developers are hired by the Blender Foundation (i.e. Ton is their boss) and none of them are talking about secession. Forks require developers and without developers any talk of a fork is just so much hot air.
Also, and I know this sucks, I don’t think anyone here can claims the “community as a whole” is leaning in any particular direction. Unlike the UI Controversy, there is next to no-one even talking about this. Three pages in a single thread is nowhere near what you’d need to determine what the community “as a whole” feels.
@BTolputt, re: secession, not sure anyone wants to be BDFL? - currently it seems to be a mix of management, paperwork, occasionally conflict-resolution, applying for subsidies and replying to a steady stream of email.
Of course the job could be split in a few ways - community manager / administrative admin / dev management etc…
From this thread a few things could be ironed out regarding the role of BDFL, (hypothetically)…
What happens if hit by a buss?
What happens if they go mad and start doing stupid stuff?
Worth noting is we pretty much the same problem for module owners for some of the more complex areas of code.
We have a low bus-factor (of one) - http://en.wikipedia.org/wiki/Bus_factor in many cases, and at the moment I’m afraid theres no easy-solutions here, getting new developers helps a lot (which we’re doing), but its not an instant solution.
@saint,
all the open movies so far have lead to huge leaps in terms of blender developments across all cg aspects… modeling, rigging, particles, rendering, etc. I don’t see how anyone would think that that’s some sort of personal pet project when everyone benefits from them.
name something that’s currently being worked on that’s a Ton-only feature…
everything currently being developed is very necessary for blender’s growth. Even cycles that’s loved by so many wouldn’t be anywhere near its current form without open movies.
Well, I understand you don’t and I can’t imagine anyone else is itching for the position… but I really don’t think Ton is ready to give up the position. However, I’m willing to be incorrect on the matter. Let me know when he says he’s going to step down.
As for the “bus factor”, one of the biggest issues is that no-one (to my knowledge) has knowledge of the Blender Foundation / Institute finances whilst simultaneously having community recognition. It’s not like the finances are (or ever look to be) published to the public and, last I heard, not even you were/are aware of how/where the funds are allocated. Having the knowledge of certain code spread over the team is important, but being unable to continue paying the developers will kill the Blender Foundation/Instititue far quicker.
I don’t think that’s the right word to use, SaintHaven is talking about the non-priority given to long requested features by CG professionals (the type that actually do VFX work for blockbuster movies and the like).
A Ton-only feature, in the most literal sense, would be something that even his own Open Movie project team wouldn’t find useful, it would be a dumb move at the least.
and is that more vital to blender than any of the work that’s being done now? i don’t think it’s fair to suggest to any degree that the decisions are based on selfish reasons. The people working on the open movies are professionals too, why should their requests be less important than those of “professionals” that don’t even use blender beyond storyboarding?
Besides, VFX isn’t the only area that has seemingly been ignored by blender. if the BF were to abandon their current projects for the sake of pleasing the VFX guys, wouldn’t the depsgraph (or particles, or hair sim) guys call them selfish too? Trying to please the “professionals” shouldn’t come at the expense of existing users. and i’m glad Ton thinks that way
It’s not so simple as to say they’re just selfish, but their requests are those that they want to see in Blender for the reason that it would encourage them to start using Blender for larger and more complex tasks than storyboarding and other trivial things.
It doesn’t necessarily mean that Ton and co. must remove the existing userbase from the equation when priorities are established, as there’s room for both groups with enough resources and professional-grade tools will no doubt help hobbyists and single artists as well.
Also attracting the professional crowd (and their studios), can become a big deal for development because they have the means to donate large sums of money to finance it. That means a bigger core team, which means a faster development rate.
I took the time to read SaintHaven’s wall of text* and have a couple of comments to make.
First and foremost, I think Ton is a pretty damned good developer. Not the absolute best I’ve seen (I’ve worked with some incredible code-cutters in my time), but amongst the top 10% for certain. I cannot, & will not, support the proposition that problems I/we/the community have in regards to the BDFL situation is due to his development skills. Also, for that matter, his project management skills are pretty good. Again, not the best, but not what I’d consider an issue. I think the main issue is, & this is going to be a little controversial, that he suffers from tunnel vision when it comes to setting priorities.
I’ll use my current work situation as a comparison. I work day-to-day with two & a half other guys in a company where we develop & license custom & retail software. We have a data guy (only “half” cos he is semi-retired) who looks after all the data stuff. We have two developers (myself and another full-time guy) who are in charge of the back-end server development (the other guy) and desktop/mobile client/standalone application development (myself). Then we have a guy who is not skilled at all in development or data management. He is in charge of sales, client interfacing, and (most importantly) making sure that our priorities remain focused on getting the best possible outcome for the people paying us money.
It is this sales/client aspect that I think is missing from the decision-making process. Yes, Blender is an open-source project and the Blender Foundation does not (& cannot) rely on “sales” of it to remain viable. However, the sales of merchandise, tutorials, etc not to mention the continued success of Institute open film projects means that the BFI cannot alienate their community either. In other words, the focus of Blender development still needs to on how the community paying for the BDFL’s wages (& those he employs) or that money stream will dry up. There was already a hint of that with the last open film and, I suggest/predict, it is a point that will become somewhat more obvious when they go to the community AGAIN to fund the project that didn’t get the funding last time.
I liken the current situation in BFI (with their BDFL) to what my company would be like without a person focused on the client (which, in this case, is the community). The data guy is always wanting more time to perfect the data and us developers are always wanting to spend time fixing code/features that the client isn’t that interested in. Our sales guy adds next to nothing to the development process… except fixing the focus on what makes the client happy. It’s that focus that keeps us successful. I honestly think that, provided someone to keep focus on the community & how they use Blender, the BDFL model would work just fine - even with one as focused on his internal vision as Ton
–
[SUB]* Not an indictment, just describing it’s length and the imposing nature that length entails. Just look at mine![/SUB]
I see what you mean… and i agree that certain aspects of blender deserve a lot more attention, but isn’t that the whole point of open movies? if we have enough open movies, eventually, every aspect will have its turn. just like the huge leaps we saw in blender’s VFX tools during Tears of Steel
I’d really like a toon-style open movie, but it doesn’t give me any justification for suggesting that the BF is selfish…
Actually, it will only give the aspects of Blender needed for the movies the BFI wants to make attention. If your focus is not on those types of movies or, as is the case with many Blender users, not even on movies at all - there are going to be aspects ignored completely by these projects.
Ah, but let’s say that one of the Blender core developers put together a feature that would better enable toon movies made in Blender. Let’s also say that, instead of letting that feature through, it gets held up due to an idea the BDFL dreamed up ten years ago, and has managed to do without in all the open films to date, but he refuses to approve of the feature being committed unless/until it is reworked to fit the idea he desires but didn’t develop himself. Most people would consider that selfish, especially as committing the feature in question brings Blender closer to the idea he dreamed up a decade ago and doesn’t stop him doing anything with Blender he already does.
Yeah unfortunately… but one can’t really expect to see monumental progress on NPR for example when the current open movie is centered on VFX. I’m not saying that an open movie justifies ignoring other aspects, and believe me i’m frustrated too about not having the tools i need in blender’s todo list. But in all honesty, i cannot say that the NPR department or character creation tools should take more priority than the depsgraph or viewport overhaul
I’d be furious if that’s the case… but the info in the example is insufficient. it would be nice to hear his reason for declining. i can’t imagine Ton actually said that. Maybe the feature really isn’t polished enough to make it to blender. Maybe the devs aren’t prepared to handle the additional responsibilities required to maintain that feature. there could be a million good reasons for declining the feature.
I do get furious about these kinds of things. so furious in fact that i moved a lot of my projects to unity. for a long time, i barely used blender for anything besides modeling… but i never once blamed Ton for not giving me a proper toon renderer… or a good alternative to freestyle… or a decent opengl renderer (which is as basic as you can get).
Ok, i just read through most of the thread and here are my observations:
Campbell himself has repeatedly admitted that the feature is incomplete… it’s still being worked on, although he has to spread his time across multiple projects… Ton is the owner of the interface module… the colored wireframes being an interface feature require the module owner (not the BDFL) to give a thumbs up… Ton AND Campbell have met regarding the topic and agreed to polish it some more before including it in blender
i honestly don’t see the problem here… or how it’s even remotely close to the example you gave earlier
I would say the focus is even a bit more narrow then that. Not only is the focus only on certain types of projects (mostly films), but the way the BFI approaches those projects is extremely unique. Everywhere else in the industry having a pipeline that consists almost entirely of a single application is largely unheard of. Not to say that Blender’s goal of being a complete end-to-end solution isn’t laudable, but it does ignore certain realities of the industry. As a result features, that are more useful for people using mixed pipelines, e.g. Alembic IO, will inevitably be given lower priorities. There is also the issue that if all of the people working on the projects are Blender experts, but have little experience with other tools and pipelines, their views will necessarily be limited.
The BFI’s open projects have certainly advanced Blender in many ways, but I think it is a mistake to overlook their limitations or assume that they are the only way to get good user feedback.
Actually, I read that it was implemented to the point where he and Pablo were happy with it… but Ton was not. Would you mind pointing me to where Campbell stated he thought it was incomplete as opposed to where he stated Ton wanted more to be done with it?
Well, frankly, you seem to be splitting some mighty fine hairs there. Not only is there no practical distinction between Ton’s veto as module owner and BDFL (especially as he, in his role as BDFL, gave himself that module ownership), but you completely sidestepped what I stated to make some argument about whether or not Campbell agreed with his boss.
OK, I’ll break it down for you, line-by-line, matching what I said with the situation… even though I get the impression that you’re going to reject anything remotely critical of Ton. Ah, but let’s say that one of the Blender core developers put together a feature that would better enable toon movies made in Blender.
Campbell inarguably implemented a feature that better enabled users to differentiate objects in a scene by their wire colour, a feature that has been requested for ten years and implemented a few times (but rejected by Ton).
Let’s also say that, instead of letting that feature through, it gets held up due to an idea the BDFL dreamed up ten years ago, and has managed to do without in all the open films to date, but he refuses to approve of the feature being committed unless/until it is reworked to fit the idea he desires but didn’t develop himself.
Ton mentioned ten years ago that he would like to see wire colours based on themes that could be controlled by rules. He has stated multiple times since then that he doesn’t like the idea of custom wire colours, the very feature requested, but would like them instead to plug into the theme system and be controlled by rules. He has refused to accept patches in the past that enabled such and multiple times refused to approve of Campbell committing his version to trunk.
Ton sets Campbell’s day-to-day priorities (being his boss and all) and states that the coloured wireframes need to plug into the theme system (you know, closer to his decade old idea) before he will approve it.
Most people would consider that selfish, especially as committing the feature in question brings Blender closer to the idea he dreamed up a decade ago and doesn’t stop him doing anything with Blender he already does.
Committing coloured wireframes to trunk as they are inarguably brings his vision of themed coloured wireframes closer, they in no way stop him using Blender as he already does (they are completely optional), and he is refusing to approve them until they plug into the theme system anyway.
Honestly, I don’t know how you can think the situation isn’t “remotely close” unless (as you’ve done) one starts adding in extra conditions, provisos, etc that were not in the initial statement. No-one said that Campbell didn’t agree with Ton… it’s kind of expected when Ton is his boss :rolleyes:
@jedfrechette,
the open movies arent about user feedback. they’re all about putting blender to the test in real practical applications and improving it in the process. the focus is only as narrow as the demands of the project. This has little to do with blender being an all-in-one package. To be fair to blender, it’s very strong in the IO department. Alembic just happens to be a small exception
Sometimes, having no experience in other packages is a huge advantage. it means you have no bad habits to unlearn. not to mention the fact that it gives you a fresh perspective on things.
Sorry, but I’m not going to keep arguing with spin that could power a small country. Themes integration is just “a kink”? Ton is not the BDFL unless he explicitly states he is before every decision? Unless Campbell argues on a public forum with his boss his statement that he was happy with a feature and Ton was not doesn’t count?
I get it. In your mind, Ton is not responsible for making decisions as BDFL unless he states he is such in the same breath, he’s not in a position of authority over his employees unless he publicly threatens to fire them if they don’t follow his demands, and employees must be in 100% agreement with him unless they start undermining him on BA. I’m not after such a ridiculous level of evidence.
Ah, but let’s say that one of the Blender core developers put together a feature that would better enable toon movies made in Blender.