6 ms·

Unlike most open source projects, Blender started its life as a professional piece of software by and for artists who had to deliver commercial projects on a de

by dagw 3y ago

Unlike most open source projects, Blender started its life as a professional piece of software by and for artists who had to deliver commercial projects on a dead line. First as an in house tool for an animation studio, and later as a commercial software product. It only became open source later in its life cycle after the commercial Blender company went bankrupt. It has also been headed by the same lead developer from its earliest days as an in-house tool right up until today. All these things really shine through and make it quiet unique among open source applications.

agumonkey 3y ago
Interesting but the special part about blender is the open source part, they managed to keep it alive, revamp the UI fully and bring on new and hard features on a regular basis. I don't remember another foss project of that kind.
- joshfee 3y ago
  krita.org is a similarly polished and professional open source project, but I agree these types of projects are the exception rather than the rule (though I suppose the same could be said about most products).
  - jansan 3y ago
    Krita is nowhere near the level of Blender. It is more on the level of Inkscape.
    - hellcow 3y ago
      As a hobbyist digital painter and 3D artist who has used both Blender and Krita, I strongly disagree with this. Krita is simply best-in-class at digital painting on a PC. Both are exceptional pieces of software. It did take me a week or two to adjust to Krita’s hotkeys coming from a lifetime of Photoshop, but the software in the end is leagues better for painting.
    - danShumway 3y ago
      Krita isn't on the same level as Blender, but it's not on the same level as other Open Source projects either. I'd put it in the same camp as projects like Ardour, and arguably above projects like Godot (although I'm sure some people would debate me on that). It's an exceptional drawing tool that is well-suited for professional work and is headed in a really promising direction. It's got a couple of weaknesses (vector layers) but it really shouldn't be discounted as a professional-level tool. Unless you're claiming that Inkscape is in the same position and has gotten a lot better than the last time I used it, which :shrug: could be true, I don't know. I'm not trying to bash Inkscape here.
      - syntheweave 3y ago
        I can vouch for Krita and Inkscape pairing well as vector art programs and even doing things that CSP, whose vector layers are pretty well liked, can't match. The issue is that drawing in Inkscape is a little bit broken(it can be done, but the current UX is death by papercut) - thus I approach it through the other program, which is basic but consistent. So the workflow I end up using for digital inks is: Open both programs, sketch in Krita, copy-paste the vector data into Inkscape, stroke->path, then use tweak tool to sculpt the lines. This adds line weight in seconds-to-minutes. Alternately, I can apply path effects instead of stroke->path, if I want a more programmatic design. If I want to paint, I can copy-paste the shape back into Krita.
        danShumway 3y ago
        I'll have to give that a try some time, thanks for the tip!
    - agumonkey 3y ago
      no but he's right, there's a level of polish that was higher than average there too, i actually thought about it but couldn't remember the name when i wrote.
- crote 3y ago
  KiCad is slowly inching in this direction! It used to be an absolute pain to use 3-4 years ago, but it has significantly matured since then. With the death of EAGLE it's rapidly becoming the obvious choice for all hobbyist use, and it is powerful enough for quite a bit of commercial work too. It definitely isn't at Altium's level yet, but unless you're designing something like a motherboard you probably won't be missing much.
- dmazzoni 3y ago
  Would you consider Audacity another example? (I'm one of the original authors.) It survived for ~20 years as a community-supported project - it was popular but clunky and limited in many ways. Since being acquired by Muse the UI has been cleaned up and modernized quite a bit, while still keeping it open-source.
  - agumonkey 3y ago
    Audacity is pretty cool, but it's not the magnitude of blender in terms of complexity.
yterdy 3y ago
This is an innocent misrepresentation of Blender's development history. While nothing you've said is false, Blender has been open source for more than 2 decades. It began that portion of its life as something of a mess, with a truly Byzantine user interface that made even trivial tasks troublesome, and lacking a vast majority of the features it's now known for. Getting it to its contemporary state was a long and painful process, including at least two major front-end overhauls and who-knows-how-many under-the-hood, and a commendable (though not unimpugnable) humility from developers who (finally) found the wherewithal to put the user experience before FOSS dogma or their own ambitions. That's what separates it from most open source projects: not that it started as a commercial product, but because its designers and developers stowed their egos and worked diligently on creating a solid piece of software (and documentation and community and support) for a long, long time. In this way, it surpasses even many of its commercial contemporaries, which are driven by a profit motive to become increasingly paywalled and enshitified.
- jasode 3y ago
  >This is an innocent misrepresentation of Blender's development history. While nothing you've said is false, Blender has been open source for more than 2 decades. It began that portion of its life as something of a mess, with a truly Byzantine user interface that made even trivial tasks troublesome, and lacking a vast majority of the features it's now known for. [...] That's what separates it from most open source projects: not that it started as a commercial product, I think your minimization of its original commercial nature with additional facts about the UI is also an innocent misrepresentation. Even though the 2003 Blender didn't have the optimal UI, what the commercial investment did for Blender was put enough value into software such that it had a headstart and momentum for subsequent investment from corporate donors/sponsors and volunteers to create the later UI overhauls. The substantial €4.5M business investment set the stage for the later developments. That it happened 20 years ago isn't the key. What's key is the financial investment to help motivate 20 additional years of work. Compare that to the open-source development of Octave (a MATLAB alternative) where the developer is lacking money and is looking for employment: https://news.ycombinator.com/item?id=13603575 https://news.ycombinator.com/item?id=13603575 Observers can say Octave is not as polished as MATLAB. But a similar hypothetical investment of €4.5M of 20 years ago might have helped give it the "Matthew Effect" like Blender got. That the hypothetical Octave v2023 has a revamped UI compared to Octave v2003 would not remove all the causal effects of that hypothetical investment.
croes 3y ago
Wasn't the user interface fundamentally changed after the open source release because it was so cumbersome to use?
- sbuk 3y ago
  Not for a while. The first revamp was 2.5, then another big one with 2.8, which was the starting point for the current UI, in 2019.
  - taneq 3y ago
    (That's a "yes". Blender 2.4 was an in-house gamedev asset authoring tool that was open sourced, Blender 2.8+ is very different and apparently pretty pro-level.
- thomastjeffery 3y ago
  Not fundamentally. Some very carefully thought out tweaks were made to make the interface more approachable. The overall UX structure has not diverged much from the old 2.x days.
crote 3y ago
Blender only became successful well after it was made open source. As a commercial project it was a failure. It only became successful due to two decades of open-source development, and the willingness of its users to invest in it - even if only to stimulate the development of a competitor to expensive proprietary software. Blender only became an even remotely viable option in 2011, after being open-source for 9 years. Its popularity only really started in 2019 after a massive UI rework made it actually nice to use. This and related changes led to Blender receiving a $1.2M grant in 2019, leading to other companies re-evaluating it and awarding even more grants. If anything, compared to today's successes its initial proprietary development should be seen as nothing more than a historical curiosity.
- grobbyy 3y ago
  It's not a historical curiosity. The decisions needed to make an architecture viable on the long term happen early on and need to be sustained as code is being developed. A lot of that comes down to days structures used, system modules, abstractions, If that's done right, it's possible to make a good UX later. That's very hard in modern software processes which emphasize short term spirits.
- PaulDavisThe1st 3y ago
  Blender received huge infusions of cash before 2019. Back in the mid 2000s, I was on the "FOSS talk circuit" in Europe talking about Ardour. Blender stuff was often happening in the room next door (way more people ... graphics ... sigh :) Blender got huge grants from both the EU and also, I think, Apple back in the mid-to-late 2000s that in retrospect look critical in boosting them out of their somewhat "stuck" development status.
  - fb03 3y ago
    Thank you for writing Ardour. Lovely piece of software
- neovive 3y ago
  Blender has such a rich and amazing history! I used it for some hobby projects many years ago, but the UI was very complex and the learning curve was high. Recent versions have greatly simplified the UI and the features feel on par with many of the proprietary industry tools. The future definitely seems bright.
orbital-decay 3y ago
Most of this is technically true but not really relevant, and no, this is not what makes it stand out. Very little of what constitutes modern Blender came from NeoGeo/NaN. The original software was uncompetitive and unremarkable. In fact, Blender had to change most of its original UI conventions to feel less alien for CGI professionals. (because the original paradigm was from the ancient times before the commonplace UI conventions) What made it stand out in the open-source community is devs using it for actually creating something (best investment!), positioning it at the intersection of interests of non-competing companies that need to get shit done, and also attracting a massive army of game modders.
- somat 3y ago
  Not really, I find the blender workflow is still very similar to the old 1.7, fits on a floppy disk, days. It's the same modal editor with a million hotkeys and dense packed stacked dialogs, the most intrusive workflow change was when they swapped the default mouse keys. Not to say the blender project has not done amazing work making things more discoverable. But as a occasional blender user from the sgi 1.X days. I suspect "the complete 3.0 overhaul making blender more standard" was more marketing than anything else. Blender had got a reputation has being hard to learn(all 3d programs are hard to learn). So while they did do a lot of work on the UI and it is much better, mainly it was loudly saying "we made the UI easier" and everybody sort of went along with it.
sproketboy 3y ago
[dead]

by dagw 3y ago

agumonkey 3y ago

Interesting but the special part about blender is the open source part, they managed to keep it alive, revamp the UI fully and bring on new and hard features on a regular basis. I don't remember another foss project of that kind.

joshfee 3y ago
krita.org is a similarly polished and professional open source project, but I agree these types of projects are the exception rather than the rule (though I suppose the same could be said about most products).
- jansan 3y ago
  Krita is nowhere near the level of Blender. It is more on the level of Inkscape.
  - hellcow 3y ago
    As a hobbyist digital painter and 3D artist who has used both Blender and Krita, I strongly disagree with this. Krita is simply best-in-class at digital painting on a PC. Both are exceptional pieces of software. It did take me a week or two to adjust to Krita’s hotkeys coming from a lifetime of Photoshop, but the software in the end is leagues better for painting.
  - danShumway 3y ago
    Krita isn't on the same level as Blender, but it's not on the same level as other Open Source projects either. I'd put it in the same camp as projects like Ardour, and arguably above projects like Godot (although I'm sure some people would debate me on that). It's an exceptional drawing tool that is well-suited for professional work and is headed in a really promising direction. It's got a couple of weaknesses (vector layers) but it really shouldn't be discounted as a professional-level tool. Unless you're claiming that Inkscape is in the same position and has gotten a lot better than the last time I used it, which :shrug: could be true, I don't know. I'm not trying to bash Inkscape here.
    - syntheweave 3y ago
      I can vouch for Krita and Inkscape pairing well as vector art programs and even doing things that CSP, whose vector layers are pretty well liked, can't match. The issue is that drawing in Inkscape is a little bit broken(it can be done, but the current UX is death by papercut) - thus I approach it through the other program, which is basic but consistent. So the workflow I end up using for digital inks is: Open both programs, sketch in Krita, copy-paste the vector data into Inkscape, stroke->path, then use tweak tool to sculpt the lines. This adds line weight in seconds-to-minutes. Alternately, I can apply path effects instead of stroke->path, if I want a more programmatic design. If I want to paint, I can copy-paste the shape back into Krita.
      - danShumway 3y ago
        I'll have to give that a try some time, thanks for the tip!
  - agumonkey 3y ago
    no but he's right, there's a level of polish that was higher than average there too, i actually thought about it but couldn't remember the name when i wrote.
crote 3y ago
KiCad is slowly inching in this direction! It used to be an absolute pain to use 3-4 years ago, but it has significantly matured since then. With the death of EAGLE it's rapidly becoming the obvious choice for all hobbyist use, and it is powerful enough for quite a bit of commercial work too. It definitely isn't at Altium's level yet, but unless you're designing something like a motherboard you probably won't be missing much.
dmazzoni 3y ago
Would you consider Audacity another example? (I'm one of the original authors.) It survived for ~20 years as a community-supported project - it was popular but clunky and limited in many ways. Since being acquired by Muse the UI has been cleaned up and modernized quite a bit, while still keeping it open-source.
- agumonkey 3y ago
  Audacity is pretty cool, but it's not the magnitude of blender in terms of complexity.

yterdy 3y ago

This is an innocent misrepresentation of Blender's development history. While nothing you've said is false, Blender has been open source for more than 2 decades. It began that portion of its life as something of a mess, with a truly Byzantine user interface that made even trivial tasks troublesome, and lacking a vast majority of the features it's now known for. Getting it to its contemporary state was a long and painful process, including at least two major front-end overhauls and who-knows-how-many under-the-hood, and a commendable (though not unimpugnable) humility from developers who (finally) found the wherewithal to put the user experience before FOSS dogma or their own ambitions. That's what separates it from most open source projects: not that it started as a commercial product, but because its designers and developers stowed their egos and worked diligently on creating a solid piece of software (and documentation and community and support) for a long, long time. In this way, it surpasses even many of its commercial contemporaries, which are driven by a profit motive to become increasingly paywalled and enshitified.

jasode 3y ago
>This is an innocent misrepresentation of Blender's development history. While nothing you've said is false, Blender has been open source for more than 2 decades. It began that portion of its life as something of a mess, with a truly Byzantine user interface that made even trivial tasks troublesome, and lacking a vast majority of the features it's now known for. [...] That's what separates it from most open source projects: not that it started as a commercial product, I think your minimization of its original commercial nature with additional facts about the UI is also an innocent misrepresentation. Even though the 2003 Blender didn't have the optimal UI, what the commercial investment did for Blender was put enough value into software such that it had a headstart and momentum for subsequent investment from corporate donors/sponsors and volunteers to create the later UI overhauls. The substantial €4.5M business investment set the stage for the later developments. That it happened 20 years ago isn't the key. What's key is the financial investment to help motivate 20 additional years of work. Compare that to the open-source development of Octave (a MATLAB alternative) where the developer is lacking money and is looking for employment: https://news.ycombinator.com/item?id=13603575 https://news.ycombinator.com/item?id=13603575 Observers can say Octave is not as polished as MATLAB. But a similar hypothetical investment of €4.5M of 20 years ago might have helped give it the "Matthew Effect" like Blender got. That the hypothetical Octave v2023 has a revamped UI compared to Octave v2003 would not remove all the causal effects of that hypothetical investment.

croes 3y ago

Wasn't the user interface fundamentally changed after the open source release because it was so cumbersome to use?

sbuk 3y ago
Not for a while. The first revamp was 2.5, then another big one with 2.8, which was the starting point for the current UI, in 2019.
- taneq 3y ago
  (That's a "yes". Blender 2.4 was an in-house gamedev asset authoring tool that was open sourced, Blender 2.8+ is very different and apparently pretty pro-level.
thomastjeffery 3y ago
Not fundamentally. Some very carefully thought out tweaks were made to make the interface more approachable. The overall UX structure has not diverged much from the old 2.x days.

crote 3y ago

Blender only became successful well after it was made open source. As a commercial project it was a failure. It only became successful due to two decades of open-source development, and the willingness of its users to invest in it - even if only to stimulate the development of a competitor to expensive proprietary software. Blender only became an even remotely viable option in 2011, after being open-source for 9 years. Its popularity only really started in 2019 after a massive UI rework made it actually nice to use. This and related changes led to Blender receiving a $1.2M grant in 2019, leading to other companies re-evaluating it and awarding even more grants. If anything, compared to today's successes its initial proprietary development should be seen as nothing more than a historical curiosity.

grobbyy 3y ago
It's not a historical curiosity. The decisions needed to make an architecture viable on the long term happen early on and need to be sustained as code is being developed. A lot of that comes down to days structures used, system modules, abstractions, If that's done right, it's possible to make a good UX later. That's very hard in modern software processes which emphasize short term spirits.
PaulDavisThe1st 3y ago
Blender received huge infusions of cash before 2019. Back in the mid 2000s, I was on the "FOSS talk circuit" in Europe talking about Ardour. Blender stuff was often happening in the room next door (way more people ... graphics ... sigh :) Blender got huge grants from both the EU and also, I think, Apple back in the mid-to-late 2000s that in retrospect look critical in boosting them out of their somewhat "stuck" development status.
- fb03 3y ago
  Thank you for writing Ardour. Lovely piece of software
neovive 3y ago
Blender has such a rich and amazing history! I used it for some hobby projects many years ago, but the UI was very complex and the learning curve was high. Recent versions have greatly simplified the UI and the features feel on par with many of the proprietary industry tools. The future definitely seems bright.

orbital-decay 3y ago

Most of this is technically true but not really relevant, and no, this is not what makes it stand out. Very little of what constitutes modern Blender came from NeoGeo/NaN. The original software was uncompetitive and unremarkable. In fact, Blender had to change most of its original UI conventions to feel less alien for CGI professionals. (because the original paradigm was from the ancient times before the commonplace UI conventions) What made it stand out in the open-source community is devs using it for actually creating something (best investment!), positioning it at the intersection of interests of non-competing companies that need to get shit done, and also attracting a massive army of game modders.

somat 3y ago
Not really, I find the blender workflow is still very similar to the old 1.7, fits on a floppy disk, days. It's the same modal editor with a million hotkeys and dense packed stacked dialogs, the most intrusive workflow change was when they swapped the default mouse keys. Not to say the blender project has not done amazing work making things more discoverable. But as a occasional blender user from the sgi 1.X days. I suspect "the complete 3.0 overhaul making blender more standard" was more marketing than anything else. Blender had got a reputation has being hard to learn(all 3d programs are hard to learn). So while they did do a lot of work on the UI and it is much better, mainly it was loudly saying "we made the UI easier" and everybody sort of went along with it.

sproketboy 3y ago

[dead]