Hacker News Reader

Comment by droptablemain | Hacker News Reader

Stopped reading after I encountered this ridiculous statement:

"Well, strange, if someone made it in 1 night, they could certainly fix it in 1 night"

I'm guessing the person complaining is not a developer and/or has never worked on a large scale project.

> But it's understandable that there are bugs, particularly with a large codebase and many contributers. The main issue I have is the broken promises and lies. For example, once on Twitter I asked Juan about why mixed mode shadows were not supported (this is when the level is baked but characters have real time lighting, a standard feature in Unity and Unreal). And he tells me that shadow mapping is faster. I don't know if he misunderstood me, but real-time shadows are not going to be faster than baked lighting. There is a reason almost every AAA game in the last 20 years has used some form of pre-computation. Eventually I found someone with the same problem, and they looked in the code and it was actually a 1-line fix (I saw the image, it worked). But the PR was never merged, for whatever reason. I also had a problem where I loaded a DirectX normal map into Godot (which uses OpenGL coordinate space) and it was messed up. I mentioned I had the wrong normals, and he says there is only one normal format, despite like 20 years of history of differences between DirectX and OpenGL. Again, I thought there might have been a misunderstanding, but these little kinds of things were starting to add up.

Doesn't seem like I'd give the Godot developers full benefit of the doubt either.

Yeah, I read the HN comments before the actual post, and it was less unreasonable and unhinged than I was expecting.

How many folks here have spent linear weeks patching up someone else's 1 night efforts?

[/me raises one hand, buries face in the other]

Bonus points if it's your own 1 night effort. :-)

Or the wonderful week+ bugs where the fix turns out to be a single line. (Especially when it comes down to something that corrupts memory.)

We had this in c++ reading rfid tags. Solution was addig on character: ^

Typical off by on

What you did there. I see i

I spent two weeks once tracking down a kernel scheduler bug.

The fix was to swap two instructions. That was a hard one, hard-won.

Hmm, I'm pretty sure it wasn't 1 night, but mostly because there's only so many bugs you can write in a single night.

> Bonus points if it's your own 1 night effort. :-)

Are you really a serious developer/architect/sysadmin/devops/engineer if you haven't muttered "what was this idiot thinking?" about your own code at least once?

It's way better if you leave a place of work, go have a fun career, and come back 15 years later and get assigned some cleanup work. You can't even rant about it.

So much for letting that problem ride for the next sucker...

Brendan Eich created JavaScript in ten days, and people have been trying to fix it for 27 years.

That's not a kind reading of the complaint. Someone added a "feature" (baked light maps) with one day's worth of work. However, the feature was incomplete and broken and it took a whole year to "fix." That definitely seems like project mismanagement. This ties in to the OP's broader complaints that things are promised and never delivered or delivered in an incomplete and unusable state.

I mean, even that interpretation seems like a wildly distorted view of reality. I've been watching the Godot community for years now and they're one of the most productive teams I've ever seen. Simply looking at the changelogs for minor versions and seeing how many bugs they fix and features they add should be proof enough of this. I mean, my god, for 4.0 they rewrote all of GDScript, and that was just a line item. The fact that someone's pet feature slipped through the cracks for longer than they would have like smacks of either a lack of understanding of the software development process, or entitlement.

> Simply looking at the changelogs for minor versions and seeing how many bugs they fix and features they add should be proof enough of this. I mean, my god, for 4.0 they rewrote all of GDScript, and that was just a line item.

This doesn't disagree with OP at all. For instance, I remember the announcement for baked lightmaps. According to OP, despite the fanfare, the feature didn't actually work. I don't use Godot, so I can't really comment on the validity of OP's claims, but just reading the changelogs isn't sufficient to dismiss OP either.

The changelogs don't tell the full story. I've been trying to use the C# since Godot 3.0 when it was announced as the new hot feature and.. it wasn't ready, it was missing big essential integrations (like exported variables). Then 3.1 comes around promising "no for real, it works now" and it STILL wasn't ready. Sadly I've been noticing that this overpromising and underdelivering is the trend in the release notes.

I'll admit that it took time, but C# in 4.0 is amazing - I've used it and it's totally a first class citizen.

> overpromising and underdelivering

Man, this is an open source project done off donations. Doing an integration with an external language is so challenging that the Unity team - a group of fully salaried professional developers - actually abandoned their "JavaScript" and "Python" integrations.

They are still ironing it out. Not missing anything particularly of what you would expect from C# itself, but GDScript still has the edge in some things eg. master/puppet function annotation for multiplayer (unless that changed recently, but I could not find documentation about it).

You're free to demand things if you paid for it.

They overpromised on a free project, which you didn't pay for. It's a non-commercial project - I'd give a lot more leeway for that. Otherwise, go to Unity, and see how fast they fix the bugs in the things they promised to deliver.

Moreover it wasn't that they were lackadaisical; they overdelivered on other aspects of the Godot project. Sometimes some pet feature can wait, compared to improvements for other serious features that affect a wider profile (e.g. GDScript, C# integration, and the tons of other stuff in their 4.1 changelogs). Their resolution and closing of issues on the Godot repo is also really impressive and consistent for an open-source project.

> They overpromised on a free project, which you didn't pay for.

I think part of OP’s issue is that “they” (the main Godot developer) is starting a for-profit business that directly profits from the community’s goodwill. OP themselves is paying hundreds of dollars a month to run the forums not to mention the time investment. I can understand the frustration of seeing the community languish while the main focus turns to a for-profit enterprise.

The OP has only been in control of paying for the server for a few months. They only took over May 15th.

The original owner only paid a few hundred dollars per year for hosting the forums. Whatever the OP is doing that is making it cost hundreds of dollars per month is on them...

Furthermore, why did the OP take over control of the forums only a few months ago if they have been feeling this way about Godot since 2020?

I regularly send a few euros their way every month because I still wanna support the idea of having a nice FOSS engine, but whether I'm paying or not doesn't matter. If you want people taking your project seriously you need to build a credibility, and by always shipping half-baked features and calling them complete you do the opposite.

Instead what we get is the fact that I cannot trust release notes to be accurate while redux complains on twitter that the big companies are focusing on sponsoring O3DE instead.

"The project is mismanaged" is a much different and less incendiary claim than "the project is a scam", though.

Yeah, there's a lot of entitlement in this post. It sounds like this person is mad that they heavily financially contributed to an open-source project and they didn't get a return on their investment.

I don't have enough context from this post alone to understand what is going on with the $8MM situation, but if the rest of the post is anything to go by, sounds like a nothingburger.

If I understood the person is also the maintainer of the forums (dev + admin), so probably more contribution in time than money, and the time being significant part of the person life.

He mentioned that just paying for the forms cost $100s a month, so yes he is investing heavily.

Having that kind of money to throw at it makes it kind of odd that he doesn't understand financing and how it would be fraudulent to raise money for W4 and then spend it on the game engine

Eh, people can have money and not understand finance. Early in my career I had a very lucky and very lucrative contract that gave me a six figure income at 23 years old. I was not experienced enough to understand financing but I paid for a lot of things for other people’s projects and hobbies because I felt like doing so.

yea, or any project really... making something and making something that works right can be wildly different amounts of effort

I was also put off by the use of the term "scam". Seems like the person is just upset funds are not being spent in the way he would like and yea maybe that is a problem, but "scam" is a leap.

…

Well maybe she shouldn't frontload her argument with stupid fluff.

Complainant is consistent in that her main complaint is also... Not very realistic, maybe what you would call stupid fluff.

The statement says a lot actually

Reply on news.ycombinator.com