[NO CCU LIMIT] Forge Networking now OPEN SOURCE

Yes this does seem like its related to the WriteRaw bytes send issue I mentioned as it was working in 17.11 for me to.

 

You could just set up a system that pings players every 10 seconds or so. If they don't respond 3 times in a row consider them disconnected.

 

NetworkPlayer has a function where you can set this timeout if no messages were sent in a time frame. But i'm not sure if an internal ping is done.

 

Not sure how this ping system works and how to use it with forge. Do you know any code samples to use it?

 

Thank you for the feedback. I got it working using the server cache system (hopefully this sounds like an appropriate use of the cache system). It seems like the problem causing other players to not be able to join was the latency simulation. When I turned that back to zero, others could join again.

Keep up the great work! Thanks again!

 

Sorry no, I've just done it with socket.io a few times - not forge.

 

I have reverted back to 17.11 for the time being and I'm trying to get these events to work:
After having the server load "Map01" I call this block of code:
-----------------
void Start()
{
NetworkingManager.Instance.allClientsLoaded += AllClientsLoaded;
NetworkingManager.Instance.clientLoadedLevel += ClientLoaded;
Networking.ChangeClientScene(GameConnection.socket,"Map01");
}

void AllClientsLoaded () {
Debug.Log ("All clients loaded!");
}

void ClientLoaded (NetworkingPlayer player)
{
Debug.Log ("player:" + player);
}
-----------------

ChangeClientScene is called on the client and loads the scene for all connected players. But for some reason the methods AllClientsLoaded and ClientLoaded are not called.

 

Is this feature implemented yet?
If so is it possible to define serveral "player" objects like in RTS games?

 

does this api can send custom bytes to server ??

 

Just a question regarding support/questions... What is the best avenue to discuss issues with Forge Networking?
I've posted on the http://forums.forgepowered.com/, but that site seems to be mostly dead. I've emailed to the support email address, but have received limited response. I'd like to find out the best way to get communicate.

Thanks.

 

Forge seems to have lost support traction after Christmas. It was thriving before... I hope it does not go the same way the bolt support ended up. I appreciate there are lots of tutorials and stuff to look at but the actual feedback support has kinda died.

 

It's a shame. I received one email with something to try (it didn't work), then nothing.
I have a project at work that I'm trying to convert over to Forge Networking, but the issue is a big roadblock.

 

Does the Slack chat still have a community? I never got an invite, is this still the best place for questions/support?

 

Did you send them your invoice for the asset?

 

Ive been asking questions on slack now for the last month without any reply. It was great before Christmas getting decent answers to some important questions. Is there something about forge I'm missing regarding support?

 

they do this often,
nothing new, last time was i think last year, they abandoned the product for a few months, one of the guys was getting married or something like that, at least that was the excuse, lol
probably this time he is running into marital issues? lol

kidding, but i am pretty sure they will come back eventually. so nothing new to forge developing.

 

He was on January 17th, so only 9 days, not exactly abandoned yet. I do hope they check in soon though. I have seen too many networking systems disappear so very concerned.

 

Hello guys!

Did someone manage connect an Android device running Android 6 on to a server?
I use Forge Networking FREE to test it's capabilities before buying it and I have encountered a problem. I use the scenes provided with the package (ForgeQuickStartMenu and ForgeHelloCubeResources) to make a Windows build and an Android build. On windows everything works (even with multiple computers connected) but when I try to connect my phone (LG G4 running MM) nothing happens.

I have attached the log that I've got after debugging. Hopefully it will help someone to fix this.

If someone knows how to fix this please leave a reply.

Thanks you!

 

http://forum.unity3d.com/threads/download-now-forge-networking-free.365528/
It seems the FREE version can only host servers within Unity Editor for testing.

BTW, how did you get the FREE version? I can't find it in the AssetStore.

 

I know that. The server is in the editor and I try to connect from Windows and Android. Windows works but Android doesn't.

To get Forge Networking FREE go here: http://developers.forgearcade.com/ and register. Then go to profile: http://developers.forgepowered.com/Profile and scrool down untill you see "Download Forge Networking FREE".

 

Still no answers. So sad.

 

Very sad. We have moved on. This is the third one that has disappeared on us.

 

Might just have to suck it up and use Unity's UNET, since that's going to be supported.
Like you, I've gone through 3 different networking frameworks, only to have them dry up and disappear.

 

they'll be back, they do this before, fade out for a while,

like i said last time was because he was getting married, fade out for like a couple of months, and maybe this time is because a brother of him is getting married or getting divorce maybe?, so give it another couple of months and they'll come back, lol

 

Hard to have confidence in a product where support disappears for months at a time, especially for those actually working on commercial games. If you are just doing this as a hobby it may be a little more acceptable. A week or two is a little more tolerable, but months is kind of unacceptable. The fact that it happened before and is now happening again has helped me avoid purchasing another network asset.. Thanks for saving me the waste of money!

 

I too work for a company that has deadlines to meet. When a networking asset does not operate in ways expected/advertised, that causes issues, throws off schedules and really REALLY puts me in a bind. When I have to wait over a month for a response, that's holding me up.
Yes, I understand the authors may have day jobs. That's fine. But at least respond to email, especially with something more than "see this video".

 

I'm very disappointed as well. At least I'm at the beginning of my project but still...

I was evaluating Forge for couple months, tried free version and decided to buy full version on December. After I made my purchase all support went downhill. This is second time that it happened to me.(First Bolt).

Anyway I'm moving on, only question is to where as UNET is still unstable/buggy.

 

Did you try asking on the slack channel? I've had pretty good luck on there.

 

We used to get answers there, but after a while even that was useless. You would see them come online then go back offline a few seconds later. If you have to chase the developers for support, then it isn't a solid solution.

If anyone is interested in taking part in a closed beta of a coming open source networking solution that will get released soon (2-3 months), contact me. You will need to understand networking basic as we don't have time to teach it and the system.

I don't want to hijack the thread, but wanted to say that we ditched Forge a while back for this solution in our game and haven't looked back since.

 

you mean open source as a free beer, or open source as paid asset with non .dll code?

also what features does it have so far? why dont you create a new thread for your networking asset?
thanks

 

Full open source, none of the BS dll crap. We aren't releasing any major public info (i.e. starting threads) until it is ready for release and has documentation ready for it that isn't in the code itself. We also haven't decided on an end license yet, though we are leaning toward GPLv3.

 

There really isn't any mention about Slack Chat in the docs as a means to communicate with the developers. From all that I've read, there was these forums (strike 1), their official forums (strike 2), and Contact Us/Email (strike 3). Really rather upsetting, as that's another $75 wasted on a DOA networking solution.

 

I tried requesting a refund via their "Support" email, because at this point it was like putting money in the blender.

But I haven't even received a response saying yes, no, or lick my boots.

Very, disappointing, I could have used that money to pay my artist, or for a Greenlight pass.

 

Well you could try using their 'EpicJoin' solution to try and get an answer..... hahahahahahahaha! I think the last response there was back in November. About the time the WakeNet system was born.

 

I looked at their EpicJoin site. Good lord, that thing is a mess and is in no way conducive to holding a conversation.

 

That site took up loads of their teams time when it should have been focused on the core of Forge. Instead, what it turned into was time sucked into EpicJoin, then more sunk into a master server that was a nightmare to work with, then their 'Bare Metal' system that never fully materialized, then a http web control solution that also was left half completed, then finally ForgeCloud that I have still not seen anything use. If they had focused on the core of Forge and just left the rest as 'eventually' projects they could have had a good solution that everyone needs, instead of trying to feature/scope creep to the product that everyone says they want but never uses. This team suffers from severe project ADHD and can't focus on one item for more than a month, it seems.

 

I can understand a lot of stuff, but not coming here and letting us all know that they are abandoning the project is unforgivable. I think folks just get embarrassed when they can't deliver and don't want to deal with the backlash. Same thing happened with others here on the asset store. It is understandable that people get in over their heads, but how they handle it is the problem. I remember being told that we should trust them. Well...that was a mistake. Trust also means being honest, especially when there are problems.

 

But is there really any alternative to what Forge Networking does out of the box?

Reading one of the popular comparison threads that comes up on Google mentions DarkRift. Anyone have experience with Dark Rift?

 

um, so yeah.. was just getting back into unity after a hiatus and was going to pull the trigger on buying this.. it looked fantastic... until this page.. so it's dead then? doesn't work? they just not supporting it? what's the story?

 

Your questions are pretty much what's on all our minds. I sent an email and forum post back in December for some assistance as to why the asset was not able to communicate across certain scenes, but fine with others. But I've yet to get a response beyond "go look at this video" (which I already had).

My suggestion would be to hold onto your money until we hear something from the developers. I mean, if this is abandon-ware, why invest in it.

 

the source is still included though, correct?

 

Yes, it is. But unless you fully understand networking, sockets, and how that traffic works, I would suggest finding another solution.

 

i do, was just wondering what works and what doesn't in a general sense.. if the code works for the most part, then it's just a matter of architecture/fixes etc. if there is a hard roadblock, then obviously another solution would be required. just didn't know what to expect

 

Like several people here have said, if you need support, this is not a product you want to pursue. (I have a better success rate with Kickstarter projects than Unity networking assets.) If you're a savvy (network) programmer, perhaps the source code is a good way to jump start your own network implementation.

What boggles my mind about this, and the other networking packages that I have purchased and watched disappear, is that no one starts with the basics and builds from there.

I.e. We're going to make a server-hosted level that multiple people can connect to. Done, now we need to remember those people. Done, now we need for players to be able to do something, etc. All of which is really just repeating what CaptainMurphy said earlier about process and focus.

Now for a slightly different angle to the rant :

I started working in video games in 1997, and basic multiplayer networking tech was already solid. Actually, even before that direct-connect dial up had head-to-head gaming solidly sorted out. Mainstream MMOs and large scale shooters came into their own shortly after that, not to mention lots of small-group (A)RPGs - and networking has just kept growing in the years since.

So why is networking so difficult in middleware engines today? Why are the basics not already complete, integrated into the editor tool set, and ready for people to build on? I get that there are a multitude of different directions/features to go with any given project, but the foundation is similar enough between genres that a feature set could and should already be there to work with.

Having worked with Unity since the beginning, and watched this asset store scenario play out repeatedly, I'm stunned it hasn't been properly addressed yet.

(Edited for grammar.)

 

I saw on Slack where you took the plunge, I'd be interested in your take on the overall state and usefulness. My career has been in art & management, but I can slog my way through code implementation as long as the code's largely stable to begin with.

 

This is exactly what the developer of BlackWake and I have been working on. The Unity LLAPI base with a simple RPC handler built onto the top of it. The LLAPI has been a fantastically simple system to build off of, and is remarkably stable as well.

 

An FYI for anyone thinking of getting Forge, the slack channel had one dev say that the other dev is taking some 'personal time'. I will never fault someone for taking time for themself as I know this field can be rather demanding, but seriously, own up to your responsibilities.

 

My intro to Unity networking was via Forge, but after looking through Unity's own networking docs, it seems like you get quite a bit with vanilla Unity (RPCs, reliable UDP, network instantiation, etc.) Yet there are many networking libraries like Forge out there.

Is there something I'm missing? Why not just use Unity's HLAPI & LLAPI? Does Forge have some killer features I haven't come across yet? I'm very curious - need to make a decision about whether to continue along the Forge path or just scrap it and build on Unity's core features.

Edit: I should mention I'm looking to build authoritative servers running headless on Linux. But it seems like that shouldn't be an issue with vanilla Unity either.

 

LOL, i see you and i are gonna get along famously
(i'm the original wall 'o text guy)

The issue as i see it, is largely the same on a general scale as it is on a smaller scale like middleware specific stuff. Mostly, it boils down to: Features. Platforms. Extensibility. Security etc. As we learn, things change. The code base that you are referring to is still widely in use at it's core, as it was 10-20 years ago. Look deep down under the covers of things and you'll find it. Think .net doesn't have COM deep down? nope, it does. How old is that tech? What was the death of ActiveX? Security flaws and scalability among others. Has it gone completely? nope. How do you think .Net came about in the first place? As a C++/ATL/COM developer you learn 1 very important thing up front. If you create a pointer you are responsible for deleting it, if you *obtain* a pointer, you are responsible for *releasing* it. Simple right? Evidently not for a lot of folks. So MS in their infinite wisdom dumbed it down and wrapped the reference counting (that's all it was), added a dash of this n that and call it a GC.. Or you could just use one of a hundred different smart pointer implementations that would do basically the same thing and release/delete that pointer from the inside, instead of relying on a 'GC' to manage it for you from the outside. Shrug.

Overall, what's changed are the interfaces and functionality fixes across time, which also drive new implementations to support the fixes etc, but a packet is still a packet, a byte is still a byte. Even if i was to re-write a networking core library, i'd still use the tools that were available years ago. Granted, my implementation would obviously differ since i would take what i know of the accumulated knowledge (the code base, the successes, gotcha's and failings) and architect from that, BUT, and this is the caveat, *i don't know what i don't know*. Multiply that with the ever changing technology demands of consumers, understand that most companies have a budget and what you end up with is someone's 'take' on what and how a thing should be implemented. in this case, networking; And unfortunately, the clock is always ticking which also drives implementation and architecture.

I can tell you from my experience that it's difficult enough to keep up with the issues caused by what I've laid out above, without adding 'artificial' (sometimes self-imposed) constraints due to politics (unity vs zamarin), inability (indie finances), or greed (company monopolies) multiplied by our (sometimes) McDonalds mentality (gotta have it now! dammit!) and we end up with the internet as a graveyard of half-assed and unfinished works.

Shrug. I just call it as i see it, and take a shot of jack once in a while lol.

Just turned 50 last august. Sigh. Seen a lot of crap in that time. Been developing since i was 15, so what, 35 years now, off n on? And in all of that time, only *1* thing has been constant: Hackers > Corporations. Why? Think about this for a sec: Back in my day, we'd throw up a full screen (can you say ASCII?) image and mess with folks, but we weren't destructive. We'd phreak ma bell. We'd hack BBS's and puff our chests at each other, but that was about it. Well 90% of us weren't anyway. Nowadays with the trillions and trillions circulating on the internet, it literally drives technology.More than it would ever drive any single company. A company whose wares suck, die out. A company's wares that don't suck, get hacked, thus the security part. Whatever can be made, can be unmade. Once you start to understand the history of things it tends to give context to where and more importantly, WHY we are where we're at today.

So, in summary, did i know buying this might not pan out? sure. But i have the source code. I'll do what i've always done in these cases, fix it, make it better (lol, i'm a legend in my own mind) or lament the loss of 75$.

In any case, it is what it is.

 

Sure, i'm still slogging through code, ok well, i *did* take a quick nap, but now i'm back on it lol