A new fork of the engine


(Akira_San) #21

Ok, my question is answered.


(Lostscience) #22

All i was saying is that if he was going to compete with other engines he is going to have to make the bge unique.I was presenting some ideas.He could look at A.L.I.C.E artificial intelligence foudation websight.He could look at pygtalkrobot and learn from it.Here is something he could read.http://www.academia.edu/2841712/Designing_an_intelligent_dialogue_system_for_serious_games


(elmeunick9) #23

Do you know what’s worse than only having one option when you talk with an NPC? Having infinite ones an not finding the one that gives a proper answer.

When you talk with an NPC in a game, it responds in an intelligent way, in the same way any human would. That’s because it’s response was actually writted by a human. However if the computer has to generate the responses, it will be clear that you’re talking with a machine, and the immersion will break.

I agree with you that BGE should have some unique features to compete with others, but right now there a lot of stuff to fix first.


(Lostscience) #24

Why would that break the immersion.When you talk to a person you can talk about anything.But you are not going to ask a storeclerk anything.What i was thinking about like in a openworld game.Talking to the people.


(elmeunick9) #25

The inmersion breaks the moment you realize it’s a dialog generated by machine. Even if you limit what you can say, even if you limit what the machine can say, if it’s generated you will know. You can ask to the storeclerk why something is better than something else you’ve seen on another store, and the moment he sais “Becouse A thing has a resistence to fire of 14 and B thing a resistence to water of 20” the inmersion braks, he should have said something more according to the sitiuation like “B sucks, do you want it or not?” or “That was probably a scam, you won’t find anything better in all C town.” But a machine can’t say that unless it has personalty, knows about the market and is self aware of were it is on the world.


(Ace Dragon) #26

I would think the more pressing thing to note here is his level of experience, keep in mind that we don’t exactly know his history with the BGE since this thread was literally his first post on BA.

He will need a lot of experience to be able to pull off a BGE fork, as the code is not exactly easy to change and the project itself will not go anywhere without sufficient organization and leadership.

Talking about a fork is easy in the case of the BGE, but it’s a lot harder to actually make it happen.


(TwisterGE) #27

I would think the more pressing thing to note here is his level of experience, keep in mind that we don’t exactly know his history with the BGE since this thread was literally his first post on BA.

Now how that (post count) would define his level of experience? Haven’t you seen the github repo? It’s already being worked :stuck_out_tongue:


(BluePrintRandom) #28

Yeah, the person coding is Tristan , (panzergame)
If I am not mistaken, one of the more active bge engine developers.

Many of the added features on the new engine are things that never made it past review
(He would post a update, they would review and take so long that the patches went out of date from what I gather)


(Ace Dragon) #29

Ok, my mistake, but it’s interesting then that he started a fork since I thought quite a few of his patches were already getting into master (if he has complaints, then he could’ve approached Campbell or Moguri to see if they can do something).

It will be interesting to see if he can get the codebase to something modern and clean, since it’s not only pretty messy but also pretty big.

As for the name, why not just ‘GameBlender’, a name like that will make even more sense if he ends up removing everything not needed by BGE users (that is assuming that none of his commits wind up back in Blender master).


(BluePrintRandom) #30

Our goalsWith this fork, we purpose an other development cycle which is spread on one month (one year for the master branch !):

  • 2 weeks to add features
  • 1 week to correct bugs and write the documentation

To reach this speed, we want to make reviews more efficiently with these rules :

  • Each review must not take more than 1 week
  • If a developer reject a patch and if he don’t answer about it after the limited time of rewiew, an other can take his seat, remove the reject and allow the patch
  • If the author of the patch is not longer active after a first reject, the patch will be considerated like inexistant until reopening
  • Patchs will be reviews on a website (developer.upbge.fr) which use Phabricator and Arcanist

this makes me think they were sick of the weeks/months patches took in the past

I personally hope they set it up to automatically avoid gpl on a single key press (build game- minus player)

and installing a bge game ads it’s name to a file somewhere running blender player brings up a list of installed games
(avoiding binding)

so the blender player can remain GPL and the games will be easily accessed.

almost like a bge steam…


(Ace Dragon) #31

But the bigger the patch, the harder it will be to maintain quality control without someone devoting a large amount of time to it.

If they start getting patches that add or replace entire systems, then they will need to have someone doing patch review full time (providing you don’t want a bunch of bugs and regressions as a result).

Also, having just a week of bugfixing may be a bit tight if they’re doing the project in their spare time. They must also be careful to not allow deadlines to rule the development process (otherwise you will have the same Russian roulette in terms of hoping your projects work as today’s BGE). The development of Unity 5.x is also putting emphasis on deadlines and people in their community are wondering if the wheels of their favorite app. are coming off.

Also, I hate to break it to you, but the issue of the GPL is mostly unfixable as far as the BGE fork goes (you can work around it, but you will have to break the connections your game would have to Blender’s code to break free of its restrictions).


(BluePrintRandom) #32

Exactly, downloading a bge game from a market, that distributes the games seperate from the player, and the player has a system to know about any installed bpender games, avoids GPL completely.

gpl is about distribution,

So

  1. Player downloads and installs game

  2. Player downloads and installs blender player

  3. Player runs blender player, blender player shows a list of any compatible installed bge games

  4. Blender Player opens seperately packaged game avoiding any gpl in the game.


(Ace Dragon) #33

And if the market goes down or someone doesn’t have the time for it anymore, bye bye games.

That’s the thing about those online markets, if it goes dark you can’t play your games anymore.


(BluePrintRandom) #34

The market does not have to be the exclusive distributer, that is just a idea,

you can host the player and the game on a web site, they just can’t be distributed bound together into a exe.

so just have 2 links

download player

paywall
(download game)


(Ace Dragon) #35

You would still have to make everything cloud-based if you don’t want people to legally redistribute your game for free (ie. the game file is never on your computer to begin with).

The trick then would be that all games would require an internet connection to play, which generally does not sit well with most gamers.


(BluePrintRandom) #36

you are talking about DRM,

My work being licensed commercially (game files etc)

has nothing to do with DRM

I think if you sell your product through say the blender market, and it is fun and well recieved, more people will buy it, then say a android app,

in the end it’s all a gamble, the secret is make a good game, and sell it, learn and repeat I assume.

also, I kinda always wanted wrectified to be mit 3.0 assets so it could be dissected,
but I have found others don’t really understand how to use it even when it’s given away 95% of the time because apparently I have a strange style…


(Akira_San) #37

And now the BGE can use mipmaps and lod_bias. :slight_smile:


(Ace Dragon) #38

You would still have to count on people to not only buy your game, but also not redistribute your game for free. The GPL forbids all forms of obfuscation and encryption and the source must be available in a place where it can be freely obtained.

You cite the Blender Market, but any selling is based on following an honor system, anyone right now can go to Github and download a copy of all of the paid addons for free (now someone who does that may have a major issue with ethics, but he’s not doing anything illegal and would not be considered a pirate). In fact there’s already a number of people who have done that regardless of the total quality of the addon in question so you will have a noticeable impact on your bottom line.

Personally though, I’m leaning towards favoring development that would allow for an increasingly tighter and fully-featured bridge between Blender and Godot (which could even include Godot reading .blend files directly or a better translation of materials that could even create the shader in Godot’s visual node system).


(BluePrintRandom) #39

you don’t understand,

the game is not gpl
the blender player is,

I could pursue people redistributing the game with legal action,
just like any other game that is not GPL.

anything distributed separately is not bound to GPL.

This is why there is so much confusion,

Even seasoned BGE veterans don’t understand.

Blender player with a menu system that looks for blender games (not statically linked) = GPL

Game loaded by player *(not distributed with player) is not GPL.


(elmeunick9) #40

And why is selecting the game from the blenderplayer better than having a launcher for the game (made in C++) that looks for blenderplayer in the computer and then runs the game?