Hacker News Reader

Comment by craigsmansion | Hacker News Reader

It's nice a high profile project such as Blender speaks about freedom to push back against the "source available" movement.

I've always wondered why niche specific software such as Blender doesn't have a ton of industry backing. Any medium sized graphics shop could have a full developer on the payroll for a fraction of the repeated licensing costs of proprietary solutions, that is, someone who works full time on Blender and whom you can directly approach, in-house, for features and fixes.

And it wouldn't even interfere with their competitive edge, since the software isn't their business. They don't have to care about the GPL as long as the software does what they want.

I think it has something to do with appealing to an employee's vanity, where getting a very expensive software package "for free" to do your job makes one feel appreciated.

View on HN

Tangent Animation is trying to change this! They use almost 100% Blender, have some developers and contribute back to the community. See "Next-Gen" on Netflix for their latest movie.

Disclaimer: Worked for them as a Blender developer on Next-Gen

Are you able to talk about anything you did there? I only get to work on open source in my spare time (and nothing as complex as Blender). What did a normal day look like for you?

BlenderConf 2018 had a talk from the co-founder of Tangent (Jeff Bell) about how Blender was used for Next Gen: https://www.youtube.com/watch?v=iZn3kCsw5D8

Thank you, too.

Our repository is public. You can look at every single commit we did: https://github.com/tangent-animation/blender278/commits/mast...

"Normal days" changed during the course of the project. In the beginning, it was work on new features for months if necessary, towards the end it were overnight patches for "frame x in shot y looks wrong/crashes/takes forever".

Thank you!

The whole movie is made by Blender? What was the pitfall? Is there any dev blog I can read?

Here’s some info I found: https://www.blendernation.com/2018/08/20/next-gen-blender-pr...

Looks like a good movie. And Netflix dubbed it to Swedish! Will probably watch it with my kids some time.

My kid definitely liked it. I love anything CG so I'm extremely interested want to watch again knowing blender was 100% used.

My kids went wild for Next-Gen, I had no idea it was done in Blender! The graphics are first-class, really impressive stuff. I'm an amateur, but I couldn't tell it apart from other studios such as Pixar.

Blender’s built in tenderer is known to be quite good, but I’m not sure they used it on the movie.

Cycles (Blender's built in path tracer) with some modifications. (Cryptomattes are now in 2.8, and the Embree patch will be integrated somewhere after.)

"Yes, cycles was used for everything, though our version of cycles was modified (with stefan's embree core, and crypto-mattes which were beyond valuable for compositing). The version of blender we used was the studios own dev version (which I believe was using blender 2.78 as it's base?)"

https://www.blendernation.com/2018/08/20/next-gen-blender-pr...

They had a presentation at the blender conference about it

https://www.youtube.com/watch?v=iZn3kCsw5D8

Yes, almost everything is made in blender except for some textures which were painted in Substance Painter and volumetric effects like explosions which were generated in FumeFX and Houdini before being imported, rendered and composited in blender.

That was pretty cool. Thank you! I wish they talked a little more about the difficulties they faced and how they solved them. If I remember correctly Ton Roosendaal once said one of the purposes of the Blender Open Movie project is to showcase the Blender is capable of handling complex large projects. After 4 open movie project it must have matured enough that Tangent Animation was able to solve most of their problems themselves! Awesome!

A senior lighting artist on the film mentions a few difficulties in the comments below.

- motion blur render times. (They patched Cycles to use Embree as its ray engine to solve this)

- memory consumption

- lack of UDIM

- clunky compositor

https://www.blendernation.com/2018/08/20/next-gen-blender-pr...

I watched this with my kid just the other day, I had no idea it was done with Blender.

It has also been used for many of the VFX shots in the TV show "Man in the High Castle".

https://www.blender.org/user-stories/visual-effects-for-the-...

There’s lots of small shops transitioning to Blender or using Blender. The studio that makes A Man in A high Castle VFX used Blender and there’s some animation studios in Spain using Blender.

The reason why it’s not used more: as good as it is Blender is still not Maya or Cinema 4D. Each 3D application has its own strengths and Maya has a LOT of tools that are application specific and that other applications might struggle to copy because they are so intertwined into its UX/UI and of course lots of proprietary R&D.

There’s also the question of pipeline - .fbx file format is the standard in the industry and for good or for bad it’s ruled by Autodesk afaik. Other applications have to adversarially replicate it (i think?) a bit like .doc files going onto iWork

And on your comment of having employees working on Blender - I believe Disney has a few people who contribute to Blender. But lots of the other applications like Maya and Katana are also extensible and studios like ILM have extensively modified Maya and others so to answer your question - they do extend other applications like Maya, Katana and NUKe and they have people on staff working on those extensions.

The thing those extensions or plugins turn out to often be excellent competitive advantage against other studios so it’s actually helpful to keep them in house. All the big VFX studios have multiple proprietary technologies that they don’t share them with anyone at all - unless it becomes advantageous to sell it to others, in which try often spin it off or sell it to The Foundry.

> I think it has something to do with appealing to an employee's vanity, where getting a very expensive software package "for free" to do your job makes one feel appreciated.

It learned 3D modelling with one of those "very expensive software packages" and it made me feel appreciated at the time. The knowledge however is now completely useless for me because the software, while still existing, is totally niche now. Blender would have been a lifetime investment...

Remember doing bits and bobs with Alias studio (more than 15 years ago) when I was a kid. I think it used to cost about $100K, which to me,at the time, was something out of this world. The principles behind these packages are almost always the same so they are relatively transferable.Of course, understanding new interface and software itself is a step one would need to take when transitioning.

This! If people think that getting to use expensive proprietary software is somehow appreciation, that seems like a massive failure of education in basics of economics?

Getting to learn skills that you can't use unless you continue to pay for the expensive right to use those skills should be valued negatively by a rational economic agent.

It's more likely that the SFX industry has their training done on proprietary software, and shops that don't use the industry standard ones like Maya would have trouble hiring. Plus the after market plugins tend to be commercial in nature and won't want to sell with a GPL licence.

> the after market plugins tend to be commercial in nature and won't want to sell with a GPL licence.

That’s exactly it, developing a high-end plug-ins is very expensive; to give that away for free is completely illogical. It’s a restaurant giving away sandwiches and selling napkins. Selling “service” and “support” or “hosting,” is ridiculous for the plugin industry. Most people in the business know how to use most of the important plugins already, so what kind of support would be needed at a level that users would pay? In all of my 15+ years of owning Waves audio plugins, I have not once contacted their support. If I need a reverb, I pay someone for a reverb. I am not going to pay someone to provide “support” for my reverb. I’m a pro, I don’t need to pay for help with a tool that has been almost industry standard for years.

I love open source, Postgres, Rails, Sidekiq for example. But the GPL-religious side of open source? Not so much. It seems too culty-dogmatic-restricted. Kind of the opposite of “free.” “You are free to do anything you want, except anything we don’t want you to do. The MIT license is the one that is actually free.

> except anything we don’t want you to do. The MIT license is the one that is actually free.

GPL's ideals are good. It's just that the ideals are hard to achieve when there's a need to also make money. I would describe MIT style to be 'free to exploit me [the software] as you please, even if you don't intend to contribute anything back'.

I wonder if a GPL that said you cannot redistribute this for a price equal to or less than the amount you paid for it.

So if you got it free, it is fair to redistribute free or charge for it or whatever. But if the original author sold it to you for $40, you must sell it for at least $41. There's no royalties; all that money is yours, it just means you can't compete with the original author on price alone.

idk tho i gotta think about it

Then AutoDesk would drop the money on the table, fork your code and develop new features in the same time, then bundle free Arnold, and every AutoDesk format. And then youare ruled out. This is the same as it's happening with AWS. They may not be able to compete on prize, but as they got a much bigger budget, you are already out, unless they are very incompetent and you are not.

Also, this is assuming your initial software is worth "stealing", it has to be good enough first, which Blender is nowadays.

>And it wouldn't even interfere with their competitive edge, since the software isn't their business. They don't have to care about the GPL as long as the software does what they want.

Most developers know this, but company managers that forbid open source contribution don't understand this. That's why I'm personally not too much against source available licensing. It's the only language old companies understand.

“Source available” licensing is useless. Many companies and many developers have tried to misappropriate the term “Open Source”, to make it be understood as “source available”, because Open Source / Free Software have sex appeal for developers, so they are trying to have their cake and eat it too.

Don’t fall into that trap. The freedoms provided by both Open Source and Free Software licensing are well defined and mean (1) the right to use the program for whatever purpose and (2) the right to fork.

Anything else is unacceptable.

As for what language companies understand, many of us don’t give a fuck. If they are good citizens and distribute software under FOSS compatible licenses all is good, otherwise we’re not interested, because the availability of the source code is not the point, the right to use that source code is ;-)

The recent reaction of those developers how non-copyleft licenses are being "abused" kind of speaks otherwise.

Which reaction?

Being enranged that companies just made what MIT style licenses allow them to do to start with, and creating derived licenses to work around it.

Yeah but those enraged people don't count- those are just whiners who didn't understand the licenses they were licensing their own work under. Or they don't understand the point of FLOSS to begin with- they thought that releasing their stuff as FLOSS was some sort of business or career move that would somehow benefit them later, which is rarely ever going to happen, and will almost certainly never happen to someone who's only contributing because they think it will personally benefit them someday. To those people, a PSA- just because some CS undergrad with self-diagnosed Aspergers and a trust fund who's never had a job in their lives told you on Slashdot that putting stuff on Github will someday make you rich or famous, doesn't mean it'll happen, no matter how much everyone involved wants that to be the world we live in. There are tons of great reasons to put stuff on Github, but none of them are related to money or fame or helping one's career.

Anyway, so then, when someone else somehow makes money using their code and they don't get a cut, of course that seems like "abuse" to them. If they had any idea what they were doing they would have realized that that's not "abuse" at all, because other people using their code for any purpose whatsoever is literally the point of the license.

It's just as bad if not worse when anyone reflexively GPL's their work without understanding that, because there's a freaking holy hell of a lot more legal consequences to understand about the GPL than there are about say, 2-clause BSDL, which is basically just "this is copyrighted by me and you can't remove my copyright" and "if you use this and something goes wrong, you can't sue me". If they didn't understand such a simple license, what hope do they have of understanding the GPL? None at all whatsoever. People who've been arguing about the GPL for 20+ years rarely understand the GPL, at least not in its entirety.

There's nothing wrong with someone making their code proprietary or GPL'd or whatever- it's their code, and it's their choice to make. But it's also their responsibility to understand all the consequences of that choice, and it's their responsibility to understand why they want to release their code in the first place.

Source available is useful for debugging at least. I sure wish I had 'source available' when making iOS apps!

> Source available is useful for debugging at least

how so?

I'm not the parent author, but when I developed Android applications having the source to the widget libraries available to me me was incredibly valuable.

It's not.

Looking at source code without the license to copy from it ... is a legal minefield ;-)

It's being buzzworded pretty badly. I hate hearing about non-software things referred to as "open source."

Speculation on my part, but I think one of the reasons the big graphics companies have their own software (think Pixar) at least in part is because it gives their results a unique visual look that becomes part of the brand.

Why anyone would use all third-party proprietary software, I have no idea.

Pixar has a house style, with procedural textures, written in Renderman. Everybody else mostly takes pictures of real world stuff and uses those as textures. Pixar's approach has the advantage that the camera can get very close to a surface without the texture going blocky. "A Bug's Life" and "Toy Story" use this heavily. Wouldn't help in making an Avengers movie.

Here's where that procedural texture approach started: "Road to Point Reyes".[1] First "photorealistic" render, 1983. Today, that's not even an acceptable game asset.

[1] https://lucasmuseum.org/works/detail/asset_id/1292

It’s risk management - it’s better to pay a bit more for a packaged product and accept it’s drawbacks than hire a developer who may or may never deliver.

I would think the obvious way a corporation could contribute would be financially.

That's what they do to contribute financially to software development. It's called buying a license.

Sure, sometimes they buy closed sourcr licenses, but as a corporation providing services to a third party you don't care about the license - you care very much the software adds the value to the output of the employee who uses it.

The GP was talking about hiring a developer to contribute. I pointed out that financial support is an option. Yes, licensing is one model, but a check is also valid.

But it's voluntary. When a company (or most individuals) get the product for free, there is little incentive to pay anything.

I think the notion that people (and companies) pay for proprietary software because there is support or some other reasons are complete BS. They pay for software because they have to in order to use it. It's just a cost of doing business.

As free software has become viable (high quality) in some areas, business folks are more willing to allow it. The primary concern is "can it do the job". The second is how much will it cost me?

We were discussing “hire a developer to contribute” vs “financial contribution”. Of course it’s voluntary and the incentive is low, but that’s off topic.

> ...than hire a developer who may or may never deliver.

I don't really see the difference between this and "accept its drawbacks", except that if your developer turns out to be productive, you have something a lot more valuable on your hands, and if the developer fails, you can replace them (whereas for a given proprietary product, you typically can not replace the vendor while keeping the licensing agreement and investment in training [this last bit being a big part of why in-house developers make sense for Blender, as they always have for 3D animation studios]).

>I don't really see the difference between this and "accept its drawbacks"

The difference is one is known, and the other is unknown. When you're evaluating existing software you can evaluate multiple competitors based on their finished product. You can't do that when you're hiring a developer.

>and if the developer fails, you can replace them (whereas for a given proprietary product, you typically can not replace the vendor while keeping the licensing agreement and investment in training [this last bit being a big part of why in-house developers make sense for Blender, as they always have for 3D animation studios]).

I'd say this argument supports using only existing established successful products, to avoid being burnt.

> one is known

Only if the software you're buying meets all of your current and future requirements, and is maintained indefinitely. If new requirements come up, they're at least not any more helpful than in-house resources (unless a lot of people suddenly have the same requirements at the same time); and if they stop maintaining it, there's a good chance you literally can't do anything about that.

You just have a different set of unknowns, one of which seems more manageable (which is probably why it is more popular), and probably is more manageable in some subset of use cases.

>Only if the software you're buying meets all of your current and future requirements, and is maintained indefinitely.

Nobody expects 3d design software to be maintained indefinitely or expects 3d design software to satisfy future requirements magically. You buy a 3d design software to make you money by creating content. Either it makes you money or it doesn't. The easiest thing to do is to simply buy what everyone else is using because it makes it easier to hire talent from a pre-existing pool of experts, or if need be, train them on a software for which training programs already exist because its popular.

>If new requirements come up, they're at least not any more helpful than in-house resources (unless a lot of people suddenly have the same requirements at the same time);

Well, wait a minute. Before you even get there, you not only have to be good at producing 3D art/content (or atleast good enough to make a decent chunk of change), you have to now hire dev/test/pm folks and successfully run a software development team in-house. I don't find this to be a realistic proposition.

>and if they stop maintaining it, there's a good chance you literally can't do anything about that.

A vendor for any software that you rely on can go out of business. Do you plan to run independent software development teams for your OS, accounting software, browser, IDE, etc ? The answer obviously is no. And the same reasoning applies to your 3d design software.

>You just have a different set of unknowns, one of which seems more manageable (which is probably why it is more popular), and probably is more manageable in some subset of use cases.

I don't agree at all. I think it actually is more manageable.

> you have to now hire ...test/pm folks

Do you? Depends on what you need done.

> I don't agree at all. I think it actually is more manageable.

Well, at that point it is a matter of opinion.

I think what you're referring to is paying for -support- There is no blender version of "Red Hat" that companies can rely on to deliver fixes or support.

I like the idea but I don’t think it scales. How many developers are qualified to work on Blender? It would be hard to hire for. Also you need someone that both works in your organisation and that fits in well with the blender developers so their patches get accepted etc.

I don’t think the popularity of proprietary solutions is vanity, it’s rather that switching between 3d modellers is really hard. It might be months before you’re even close to productive again. Also interoperability between packages is a huge problem. Blender only really started to get a good UI this decade, and I think you’re more likely to see it get picked up in new shops rather than old shops migrating.

"Switching between 3d modellers" undersells the issue.

COMPLETELY ALTERING YOUR PRODUCTION PIPELINE TO ACCOMMODATE ONE PACKAGE gets the point across a little bit better.

There are studios out there that have decades of in-house scripts and programs built around a given set of software. If you do a lot of architectural rendering you may have your own outdoor lighting systems. If you focus primarily on title sequences you may have certain effects that you developed. Having to completely rewrite all of that in-house code would be a MASSIVE undertaking.

> Any medium sized graphics shop could have a full developer on the payroll for a fraction of the repeated licensing costs of proprietary solutions, that is, someone who works full time on Blender and whom you can directly approach, in-house, for features and fixes.

DIY is time-consuming, difficult, and doesn't scale. Small companies often start out with as cheap a solution as they can, but as they grow they'll find they're doing way more work just to support the DIY solution, and features are difficult and time-consuming to produce. If you have the money and you need a feature now, licensing makes sense.

> I've always wondered why niche specific software such as Blender doesn't have a ton of industry backing.

Because prior to Blender 2.8 it was really really bad software compared to commercial alternatives.

Blender is becoming quite good. Which impresses the hell out of me. But it’s taken a long, long time to get there.

What changed to 2.8 to make it compare favorably to commercial alternatives? Is it just the UI?

User interface, discoverability(That whole UX thing), and a new reasonable sane default control scheme.(Which can be toggled back to the old one.)

What's great along with Blender, is the availability of very very inexpensive rendering using https://golemgrid.com/. It's name your own price and can be had for 1/10th the cost of using a render farm.

Maybe they don't have time to wait for features to be developed? Only the biggest projects have people saying "we'll need new fur/hair/dragon breath/water effects" early on and have a long or flexible timeline. Whereas your average project just need to work with what they already have. Even GoT which had a huge budget must have tight delivery deadlines that will make a project manager think twice before putting software development on their critical path.

Statistically speaking software development is extremely risky from the point of view that n units of work will create x units of deliverables.

If you are not a software shop you really would like to outsource those risks.

Hence, they pay the capital cost to get a fixed, known deliverable.

Developing new software is hard and risky. Sure, if you can get the right team for the job it can work out wonderfully but this just creates another level of risk - you need to find the right developers.

If someone wants fixed, known deliverables, the last thing they should be looking at is proprietary software, where a seemingly stable product suddenly updates and everything is different and the license changes retroactively, e.g Photoshop. Version controlled, open source means you can check out any version you might need, at any time, now and forever. Seems worth throwing some money at such a promise.

Spoken like someone that doesn't own a business. You do not throw money at promises. You throw money at solutions.

Possibly a better phrasing would be "An open source ecosystem is worth investing in to hedge against the risk of license and source changes by the vendor"

If you are short sighted enough to only care about day to day operations and don't see competence in software usage as an investment that will bring competitive advantages in the long run.

Your competitive advantage is your staff, not your software. A company like Pixar only has so many proprietary tools because they've been around so damn long that rolling their own was the only way to go to get the results they wanted.

Even then, Pixar's advantage was the people they hired and allowed to thrive... not Renderman.

I agree; software people tend to think about business in terms of software. But if you're a production company then you should focus on your core competency and outsource software development. You generally don't have the in house expertise to manage software projects.

Conversely, if you're a software company and have media production needs, you generally don't in-house it — you focus on your core competency and outsource media production.

Software people tend not to question outsourcing when it's done by a software company outsourcing things outside their core competency.

The main reason why Blender,GIMP and the likes don't get a lot of backing from the industry is because pricing is just a small puece of the overall picture. If a shop is anything but some random visualisations for real estate, there are budgets, decent salaries and etc., so even spending a few grand on Maya or 3DS Max doesn't quite change anything. Development in such a niche and complex field takes time.I doubt one developer would be able to do much,even full time,apart from fixing some small bugs..

I've used extensively Blender and HAD TO use Gimp. I strongly disagree the correlation. Although it is one you could make on paper, Blender is a fantastic piece of software and highly usable. Gimp is gimp.

"Any medium sized graphics shop could have a full developer on the payroll for a fraction of the repeated licensing costs of proprietary solutions"

If the returns to the particular company that employed the developer was higher than the opportunity cost of expending that money elsewhere in the company, then you'd have a strong case for that. However, most medium sized graphics shops are probably concerned that they'd effectively be subsidising their competitors' production costs, since the 'exclusive' returns to the shop employing the developer are likely minimal compared to any other user of the feature they develop.

The free rider problem / tragedy of the commons strikes again.

> However, most medium sized graphics shops are probably concerned that they'd effectively be subsidising their competitors' production costs, since the 'exclusive' returns to the shop employing the developer are likely minimal compared to any other user of the feature they develop.

What causes most usage of big proprietary software is that the software has 10,000 features and every customer only needs five of them, but for each customer it's a different five. That makes it hard for a free competitor to get users because adding five features only gets one customer; it scales poorly without revenue.

But as the customer yourself, having your developer add the five things you yourself need is completely feasible, and doesn't help your competitors that much because the five things they need are different.

Nobody wants to wait for a developer to implement the things they want. If they have work to do, they need a solution NOW. Pick what's available and pay what it costs. For commercial software that's some $$$ and for Free software that's $0.00. Then once the purchase has been made, the question of paying someone to work on it is a separate issue. Most rational people would say "why would I do that?"

Comparing paying a FLOSS developer to paying for commercial software doesn't make sense. Companies don't see themselves as funding development either way.

> Nobody wants to wait for a developer to implement the things they want.

And yet they still want them, and things take non-zero time to implement regardless of who is doing it. It's not as if proprietary software appears from whole cloth in zero time containing every feature requested by every customer.

> If they have work to do, they need a solution NOW.

If they have work to do they can do it the same way they did it last week. But when the question comes whether to upgrade to the new version of the proprietary software, you should ask whether it makes more sense to pay the money to do that this year, or stay on the existing version for a year while you use the money that you would have spent and instead get the free software equivalent into a state that it satisfies your needs, then never have to pay for the proprietary software again.

> Then once the purchase has been made, the question of paying someone to work on it is a separate issue. Most rational people would say "why would I do that?"

Because you have twenty or more full-time people using it and if you can make them each even a single digit percentage more efficient on a permanent basis, it becomes highly profitable to hire the software developer that allows that to happen.

> Comparing paying a FLOSS developer to paying for commercial software doesn't make sense. Companies don't see themselves as funding development either way.

What did they think the proprietary software license money was funding?

They probably see it as rent. There is awareness that newer versions may be better, but they pay to use the software.

You could make the same argument about commercial software. The money you pay the developer will go into building features and better quality software that helps your competition as much or more as it helps you. Yes they are also paying, what difference does that make? I don't see why you'd waste time worrying about this in either case.

At least, your competitors will pay too. If it was open source, they could use it without contributing back.

The big issue with blender compared to some of the other options on the market is the cluttered ui- it doesn't look friendly to artists/tech artists and the controls are very unintuitive for someone who uses other 3d softwares.

I'm curious as to why the community working on it hasn't taken steps to change this- the software is equally if not more functional than alternatives like Maya and making it a bit more familiar to industry professionals could go a long way

The "source available" movement was already a thing back in old days, before the rise of GPL.

I bought plenty of commercial software, specially developer tools, that had source available on the installation floppies.

If anything we are turning back to those days, after many are starting to realize how hard it is to keep a business living from donations and occasional consulting gigs.

I think it's more about vendor support. There's a big difference between having a person on your payroll who can add a feature and having a help desk and applications engineers who are under contract to make sure the software works for you.

It has nothing to do with "employee vanity" and that's one of the stupidest things I've ever heard.

There's a simple reason Blender doesn't see much use in film production: Support. There is no way the Blender Foundation could provide the level of support that a company like Autodesk does. When you run into problems you call your enterprise support team and get it fixed. This is why you pay them the big bucks. Their support staff does nothing but fix problems, where anyone that you could hire to do the same would necessarily spend most of their time sitting on their ass doing nothing since stuff just doesn't go wrong all that often. That alone will keep Blender out of a LOT of shops, especially those that are just large enough to be working on major projects.

The idea that a company should just hire a programmer they don't otherwise need just so they can make some software work is asinine. That's like saying a home builder should hire a full time auto mechanic instead of taking their work vehicles to the shop when they break down. No, hiring a guy will not offset the cost of purchasing licenses and support. This is especially true when you hit one of those moments where things get so shredded that the cost of enterprise support seems like a bargain.

There are other, more technical, issues that keep Blender from being used much. Chief among them is that it isn't really exceptional at anything. It's not uncommon for things to be modeled in one program, textured in another, then imported into another for animation, and then brought into a final one for lighting and rendering. There are better programs for modeling than Blender. There are better programs for rendering, texturing, animation. Blender is oatmeal. It's there, it works, but it's not really great at anything and as a result has never found a niche.

Because, ultimately, that's what it's all about.

uh, couldn't a (or multiple) blender foundation(s) sell support? the software remaining FOSS, but feature requests / bug fixes being paid services. Feature and bug attention would be driven by the big commercial users, and the community would profit. Since the shops can choose which blender support team they buy support from, they get competitive pricing too.

> The idea that a company should just hire a programmer they don't otherwise need just so they can make some software work is asinine

Weren't you just saying the magic ingredient was "support"? Do you believe a first line help-desk is going to provide better support than a guy whose livelihood depends on it and you can call to your desk?

>That's like saying a home builder should hire a full time auto mechanic

In this specific case, it's more like fairly large transporters having their own mechanics, which they have, because it's cheaper, and better, and faster, and generally very convenient all round.

I worked as one of several in-house Blender developers on "Next Gen". In some cases, were able to get fixed builds into our artists' hands in within 24h of a bug report. I have not had such a quick turnaround with any commercial software vendor yet.

https://www.netflix.com/title/80988892

>Do you believe a first line help-desk is going to provide better support than a guy whose livelihood depends on it and you can call to your desk?

I can guarantee you there are no top-level VFX houses or game developers that ever speak to first-line support.

If you're not top level you have no need to waste money on a programmer that will likely be sitting on his ass 23/7. No, that isn't a typo. The vast majority of the time these systems work as intended. When they don't, who do you think is going to respond better? The multitude of professionals that are paid to work on this code day in and day out... or Steve, who spends most of his time pretending he isn't shopping for waifu pillows? WE KNOW IT'S YOU, STEVE. CLEAR YOUR BROWSER HISTORY.

Reply on news.ycombinator.com