Hacker News Reader

Comment by elaus | Hacker News Reader

As a hobby videographer using Ubuntu I find myself switching between Kdenlive and Davinci Resolve, depending on the size and complexity of the project. For a small 30-second clip I like the more approachable, simple workflow of Kdenlive but holiday videos with a hundred or more cuts work much better with Davinci for me.

Both are free which is really great, but from an ideological standpoint I much prefer Kdenlive (OSS). Also it's a bit annoying that Davinci doesn't support MP4 with the free version on linux. It's always a lot of re-encoding with ffmpeg to "prepare" clips for use in Davinci (as almost-lossless MOV containers).

View on HN

Resolve is built by a fairly small team in s/Australia/Singapore. If you want the features in the Studio version, you just need to buy once, you get a lifetime of updates, no subscription.

Ideologically I love to use FOSS products too, but I also want to support commercial businesses that make software under customer-friendly and non-subscription licensing terms.

If we look at Blender's history, one could make the argument that they would get more support if they fully opened their product.

I for one would contribute to get a Libre Resolve, even if I am not a videographer.

> If we look at Blender's history, one could make the argument that they would get more support if they fully opened their product.

I think it's important to know that Blender probably became big because they went open source with a nice tool, not that they were big before and then released the source as open source. If Blender didn't go open source, I'm unsure if Blender would even be around today, they were in big trouble before buying the source code back and open sourcing it, if I remember the history correct.

DaVinci is not in trouble in any meaning of the word, the opposite. The editor is taking the media sector by storm and gaining popularity like crazy, so they don't have the same need of doing something drastic to save themselves.

Don't get me wrong, I'd love for DaVinci to end up open source, but I don't think Blenders history is relevant in this particular case.

being open doesn't mean it becomes free to use, you can still sell your licenses

https://www.aseprite.org/

successful product, and is open source: https://github.com/aseprite/aseprite

there are many examples like that

being open source allows your users to participate in the development, fixing things, and adding new features the company wouldn't have done themselves because: lack of budget, or "too risky"

this is why i dropped Sublime Text for a different editor, i was waiting years for them to fix a feature that annoy me so bad (file preview when you search), and it still not fixed, closed source mean i can't do shit

i no longer use proprietary software for that reason, whenever possible i decide to use one that is open source

also the industry adopting Blender is a clear sign that a product being open source plays a big role for adoption

same for game engines, people preferred open source solutions for decades, the closed proprietary ones all died (minus Unity but they provide source in exchange for a license, wich i believe will be made fully open in few years, they will have no choice)

for existing business it is understandable if they started closed, and use themselves licensed code/product

but for new businesses, if you don't consider open source, you start with a massive disadvantage, worse, it makes you impossible to grow

I’m curious why you say Unity will have no choice but to go open source?

it already is happening (kinda), they are modularizing their engine, and all the new parts are made open source

biggest part what their graphics stack https://github.com/Unity-Technologies/Graphics (only the scripting part)

competition is growing, unreal is fully open source, and more and more people have been asking for the source, there was even some drama few years ago when unity DMCA'd (if i remember correctly) a reference source code of unity engine (decompiled) on github, then unity made it officially available on their github: https://github.com/Unity-Technologies/UnityCsReference after users complained about it

wich already proved effective and caused them to fix a huge performance issue shortly after: https://github.com/Unity-Technologies/UnityCsReference/pull/...

unreal is not "open source". "open source" has a specific meaning and Unreal doesn't fit. Some people call unreal "source available"

https://www.google.com/search?q=open+source+definition

aseprite isn't under an "Open Source" license, but a EULA that even prohibits redistribution. I'd call it "shared source" or "public source" rather than "open source"

https://github.com/aseprite/aseprite/blob/main/EULA.txt

> also the industry adopting Blender is a clear sign that a product being open source plays a big role for adoption

I strongly disagree. Blender is a wonderful story and a wonderful team but it’s truly the unicorn in the FOSS software space.

Industry adopted Blender because it has evolved into being a really good tool for the job, is frequently updated (thanks in part to those corporate sponsorships that most pieces of software don’t get, unfortunately), and because it is good software.

Thinking the license played any role in that, I think is foolish. Industry, particularly the video and vfx industry, is not moved by ideology. It’s about what works, what is fastest and what can do things the best way. The editing world has relied on proprietary software since time immemorial and the two cases where we’ve seen more adoption of FOSS tools (Blender, OBS) are both recent (and again, I would argue are largely exceptions b/c the software is good and usable which is not the case for most video packages that are open source) and in tandem with existing tools, not a replacement for those tools.

As I said in another comment, I think in many cases, the open development model is best. But I also recognize that it isn’t always the right business move. And in this case, when you’re talking about an industry that doesn’t have any ideological ties towards doing things the open way, I don’t think it’s an idea that makes a lot of sense. Even if it would be nice to see.

Exactly. What's more I prefer for the software I use to be proprietary, because then I know that me paying its price is enough to support its ongoing maintenance and development. When it comes to open source projects seeking donations, I'd need to pay incomparably more, knowing that most users don't pay. With the proprietary model everyone pays their fair share to keep the project afloat. With donations an army of freeloaders bears the benefits of the funding provided by few who pull all the weight. So if a project is donation-funded I look instead for a proprietary one, because it will be more likely to stay afloat in the future and return my investment of learning it and organizing my workflow around it instead of becoming abandonware.

May I suggest an alternative strategy/view? Don't think of it as "freeloaders" vs "few who pull all the weight", and don't think you should be paying to "keep things afloat".

Instead, keep whatever tool is best for you at the moment (even if proprietary) but don't stop looking/fomenting the development of a viable free alternative. If you find any project attempting to build an alternative for the commercial ones, follow up with (a) small financial contributions, (b) periodic best-effort attempts/evaluations of the tool as the alternative and (c) feedback on what is missing for it to become your main alternative.

Unlike businesses, free software projects don't need to have a short-term ROI and they don't need to give up if the venture is not profitable. But your small contributions will at least show as a signal that the developers have a potential audience, and your feedback can also provide them with guidance/help them iterate faster on what is relevant.

Aside from marketing materials and mission statements, companies rarely care about their customers and they will not hold you in high regard just because you are giving them money. If they find themselves in a position where they can make more profit by screwing you over, they will.

Conversely, if the free alternative does become a hit, you and everyone else gets to benefit. If it doesn't, it can be picked up later by other people. And even if it fails completely, you didn't lose much.

It's the type of strategy that gives you very little downside, a huge upside and also gives you bragging rights and the feeling of a righteous soul. :)

For every Blender we have countless other projects getting almost zero donations or support. Why take that risk if the current model works for the Resolve team?

Blender made their opening conditional on getting support. They basically ran a crowdfunding campaign that said they would open source it when they reached X amount of money. (I think it was 500k€ at the time) that was good not just financially but also because it worked as a marketing/awareness tool.

BMD could do is something similar. They could make every (yearly?) release closed but GPL it once their revenue hits a mark.

They had to buy rights from investors and it was 110k eur.

https://www.blender.org/about/history/

Yeah, thanks for the correction. I totally misremembered the order of events (I thought they were still open as a business when they created the foundation) and amounts.

From my perspective, if a product is open source, it means I'm not at the whims of a company which might or might not go under or change models. I have huge amounts of content from decades ago locked up under no-longer-existing proprietary formats. The companies long went out of business, activation servers are gone, and in most cases, there's no way to find the .exe files.

My LaTeX files still work.

Open source doesn't mean I won't pay for it. Indeed, I'm more likely to pay for it.

I think in the case of Resolv, if the bulk of the money is on software, open source is a huge risk. If it's on hardware, it's a nominal risk and a huge potential upside. To the extent there are cloud-based collaboration services, the risk also goes down, since that's what most businesses pay for. Even if the server is open-source, it's cheaper and more robust to pay BM than to hire a sys-admin to run it myself.

I have an old Olympus AIR camera which is great hardware, now useless since Olympus discontinued the associated software, then got sold, and it's basically now dead.

If Blackmagic were 100% open source, I'd know their hardware is forever. Perhaps the software won't continue to improve -- there are plenty of dead projects -- but SOME nerd will at least port the existing functionality to Ubuntu 32.04 or whatever comes out. If push comes to shove, I might even be that nerd.

OK, but that doesn’t answer the question. That shows why you personally would feel better about it, but it says nothing about why a company should change a model that is working for them to a model that would require more work (it’s likely that Resolve contains components and libraries that would be very difficult to open source and would require rewrites — to say nothing of the general work required to make a project open source and accessible) for frankly very little gain.

Again, given the various codec, driver, encoding, and other patent and license encumbered issues related to video stuff, I would seriously be surprised if this was something BMD could open source even if they wanted to. And for what? The 0.0001% of video professionals who insist on using an open source piece of software to use for their work?

It works on Linux not for the free software crowd but because video professionals who are using specialized Linux servers for high-end editing workflows (using almost entirely proprietary software, I might add) requested it. As a side effect, desktop Linux users also have access to a pretty great editing suite.

I like open source too. I think that it is generally a better development model.

I also understand why specialty tools often aren’t open source and think that history has proven to us time and time again that the financial upside to OSS is not what people wish it would be.

I’m not saying it wouldn’t be nice. But for Black Magic, I honestly see absolutely no compelling reason why it would be better for them to invest a bunch of money and resources into changing a model away from something that works to pacify an extremely small number of users who aren’t even their target audience.

That does very directly answer the question of "why."

You answered the opposite question: "why not."

Neither of us know enough about BM's business model, plans, codebase, etc. to know whether open sourcing makes sense, and if so, how. I trust BM is doing something reasonable here. The only point is that the decision isn't obvious.

In terms of 0.00001%, my experience is that bigger businesses care a lot about business continuity. I'm at a large organization, and it's cheaper to buy than build. At the same time, it's lower-risk to buy open-source. That's where things are trending.

Their hardware isn't "forever" though. The industry they cater to and the consumers that industry serves are constantly demanding higher resolutions, higher frame rates, more quality, and at the industry level, the ability to process the vast amount of data that generates in a workable way. BMD aren't in the business of supporting cameras or video recorders that are 20+ years behind the times, they're in the business of delivering higher-performance equipment that improves on a "Moore's Law" type trajectory. Resolve is a tool they use to help manage this hardware process. Open-Sourcing their Intellectual Property and investment where Resolve is concerned wouldn't confer much benefit, other than catering to a microscopic niche of people such as yourself who have the time and patience to make changes in a highly complex codebase. Businesses don't owe it to you to make their hardware "forever". You could equally be trying to make a case for demanding the VHDL that runs their FPGAs open source, which would no doubt be lovely while being commercially damaging, and harming their ability to pay the high wages of competent and focussed engineers that bring you such complex products at a price you can afford.

P.S. Your text files will work better than your LaTex files in 50 years.

There isn't a "the industry." There are multiple industries.

Places where I've used AV equipment costing >$5k where lifespans were in decades:

- Permanently mounted systems in lecture halls designed to record talks.

- Machine vision systems in manufacturing processes.

- Scientific instrumentation.

My local CCTV station also has ancient, high-end equipment.

What make BM cameras great is that they can embed in a lot of places like these.

If "the industry" is limited to professional TV production, I agree with your point. However, the market for cameras is much greater than that, and those sorts of oddball segments are growing exponentially as costs fall, machine vision improves, etc.

I don't demand anything from BM. I also have no intention of diving into their codebase. I do have a handful of C-mount and MFT cameras in different capacities, permanently mounted, for various needs. If BM is a better solution for me, I'll buy from them. If it's worse, I won't.

>Resolve is built by a fairly small team in Australia.

I used to work at BMD in the BRAW team. With a few exceptions here and there, Resolve is pretty much exclusively maintained by a team up in Singapore. They are quite a small group, though!

s/Australia/Singapore …

> As a hobby videographer using Ubuntu I find myself switching between Kdenlive and Davinci Resolve, depending on the size and complexity of the project. For a small 30-second clip I like the more approachable, simple workflow of Kdenlive but holiday videos with a hundred or more cuts work much better with Davinci for me.

What about FLOSS Olive video editor?[0,1]

As for me, for a small 30-second clip Olive is much better choice than Kdenlive.

[0] https://github.com/olive-editor

[1] https://olivevideoeditor.org

It doesn't even support mp4 on Linux if you pay. If you pay you just get the ability to decode H.264. You can't encode H.264, and AAC is completely unsupported.

Knowing both software packages this comment feels a little bit like it can not be serious.

If it is: there is no way you can compare Kdenlive with Resolve - I guess as a hobby videographer you just can not see the huge differences. Kdenlive might be OK for some occasional hobby editing, but Resolve is a professional tool for people that depend on delivering perfect results quickly.

BTW I was not able to use Kdenlive one single time without crashing in the middle of the project and leaving me with only lots of time wasted. Do not use it for serious (money) work!

As somebody who likes to use Linux privately I can tell you that Resolve works best (and most cost effective) on M1 Macs. I tried to use it on Linux but I was wasting so much time fixing this and that (gfx card drivers horror stories) and still not getting the real thing, it was much cheaper to buy a M1 Mac and enjoy pro software on a system that is mobile, quiet and more powerful than my old-school video editing workstation.

Reply on news.ycombinator.com