Epic's own hosting framework (Single - and Multiplayer)

“The missing puzzle piece in Unreal Engine development and for deployment, success.”

Why not add hosting with a similar business model like Gamesparks.com does, just dediated for Unreal Engine? Currently GS does not offer multiplayer server (working on it since around Nov 2015), and implementation via Steam seems not very straight forward and lacks in many regards.

Pros
Another reason to pick UE
Another possible revenue system for Epic
Would help developers who are not experienced with network infrastructure, or lack resources
Would result in more UE games
Would result in more multiplayer games
Optional for existing and future UE devs.

Cons
Requires robust scalable network infrastructure
Requires additional service staff
Requires dedicated development team

Possible features
Distribution platform, via UE game client, think Steam
Backend tailored to offer BP and code setup via server backend for dedicated tasks and game functionality
Backend editor for game admin tasks
Integrated setup for easy integration of game features, eg. achievement system, login, highscores / ladder
Hosting of game client (See also Google Buckets)

Integrated extensions
Extensions could be tailored framework marketplace additions (UE backend compatible), which once bought, would become available in the Unreal Engine Backend, eg. chat system, quest system, auction house etc. This would tremendously help to unify code setups, and streamline development. Thus, offer easily accessible logs, debug tools, database editing for quests and such.

Development
How long would it take to setup the integral parts of such a system? With scalability in mind, and possible partners like Google Cloud, Amazon, Azure etc (Epic could pick their partner or do it alone based on existing infrastructure, but i would pick Google), a reasonable time frame to begin testing it seems.

Dev Team
To assemble the required team should be also not a big issue, looking at the skills around here, and the company’s brand and resources.

Vision
Imagine the new UE experience layer, start the “UE game client”, begin browsing all the new games from games actively developed, or switch to the shop tap. Steam would be no longer a requirement.

Business model
While initial costs and directed resources would require some commitment, once the infrastructure and software is in place, you basically are down to administration tasks, depending on scalability. If for instance a specific title becomes successful you could put a revenue system in place to compensate and for your revenue share. Or you get a share from each title (Steam is about 30ish % i read somewhere). This would also bring all the existing Epic games under one hood.

Related
Sweeney thinks Microsoft will make Steam “progressively worse” How would Microsoft destroy win32 platform? - Community & Industry Discussion - Epic Developer Community Forums

Think big!

Hey Unit, thanks for the suggestion! I’ll run it by some folks and see what our devs here think. If anyone else is interested, has ideas or wants to voice concerns, please let us know here!

I can see why it could be a good thing - we need an alternative distribution platform.

Exactly, think a few years ahead, UE could become a real alternative, when extending into distribution and tailored network framework territory, to separate UE from the competition. This would make things so much easier when you have one resource to learn / deal with instead of third parties. Epic could build upon their existing experience in those regards based on titles like UT and Paragon.

And many successful company titles do not rely on Steam (LOL, Blizzard games, WOT etc), showcasing that alternatives are feasible and rewarding.

Sorry guys, but epic launcher would not become alternative for steam, it could be a nice little addition, but its to late to fight the steam without huge investment, you need huge hits, available only here, so people start really using launcher over steam, as blizz, riot, wargaming do.

Everything else, i really like, especially server stuff, it would be amazing, even if unreal just got some sort of master server part, so people could do their own session infrastructure out of box, im not even mention server hostings.

Ofc, there should be an extra UE client for browsing, buying and downloading game contents. This would also expose users to other content, help bring your title to an audience, and all the UE titles in one place, and no need for an extra launcher. With this you have the existing dev launcher, and the one for game distribution.

Epic is more than capable of putting out content worth coming for, Already UT, Paragon, A place holder for Fort Night, And a tab for modding tools (for Ark) already exist on the once solo purpose Epic Launcher.
Also just like many of the other companies out there you can support multiple launchers. No reason a person cant allow customers to come into the Epic ecosystem through their favorite launcher (most often steam).
Numerous titles on steam install and launch through their own launcher.

I personally would love to see something like what is being described above, In fact its what i always thought should be the case from back before i even knew how it actually works. Trying to deal with things such a either building your own backend or using someone else. Such as. Playfab, or Gamesparks can be highly frustrating. As those services often dont maintain their API for UE C++ much less UE Blueprint. Having a engine specific use case i would think it could highly beneficial for us 3rd party guys, As well as EPIC own internal use.

+1 it would be good for Epic to have some backend services

Interesting idea, though I still think it’s a good idea to support Steam in your game anyway (it’s not too tough really) - since it’s just the dominant platform and is going to be for quite some time, UWP aside. Microsoft would destroy themselves if they started killing Steam, consumers would riot.

I guess this is kind of what the Launcher is to Epic already mind you. Long term plans might be to turn it into a distribution platform too.

This is a good point, indeed a second client is not really needed.

Related

Facebook is building its own Steam-style desktop gaming platform with Unity Facebook is building its own Steam-style desktop gaming platform with Unity – TechCrunch

OOOOOHH… this is awesome idea. Will make multiplayer development so much easier and straight forward. I really wish Epic will listen to this and take it seriously :smiley:

Errr… For those of you who aren’t aware of Epic Launcher client(especially you, who has fast internet), I can tell you that Epic Launcher client won’t download for installation including downloading new updates. With internet speed of 1mb/2mb are bound to suffer. You can Google search for this issues, many complained and we know overseas users or Aussies mostly are the ones that get this issue to download it.
This is if you guys are suggesting to host it in that client. If not mistaken, Epic has no CDN.

This is good idea afterall.

I like how OP listed his statements. Very well said, should throw more to grab Epic’s attentions further.

I’m both for and against the idea.

I’m against it because I don’t want to see “MY FIRST GAME” all over my launcher. There would need to be a QA process before a game appears on it - Similar to how PS4/XBone do it. That’s my pet peeve with Steam - The low quality of games now flooding in.

To be a bit pessimistic about this.

Yes, we have an emerging marketplace, which was something Epic didnt do before. And there is the problem. They dont have the experience with that.
When you look at the missing features the marketplace still has, i am naturally pessimistic about how fast hosting and multiplayer servers could evolve. Especially when you want to have some autoscaling, this is not something easy if you are not generally in that business like Amazon with AWS, and this would be a HUGE pre-investment from Epic (1M for Loadbalancers/switches/routers/firewalls, around 800k for a minimal server setup with an ESX cluster, 500k custom development for scaling, ordering and provisioning of VMs, 24/7 support personnel, etc, etc) which can easily go up to a few millions for a BASIC equipment for like 20 medium sized multiplayer games.
Not counting in rent for a datacenter (building your own would be too costly then).

Amazon took quite a big investment and some time for building up AWS, and they already HAD the servers and most of the infrastructure already.

So, sorry for being “that” guy, but i felt i need to get that out.

Cheers,

I believe the original suggestion was with a pre-existing hosting platform in mind, i.e. AWS, Google.

Never going to happen IMO
Also better learn some distributed computing.

A thing I have noticed among game devs is that not enough know how to set up a server like enterprise developers do.

This sounds like a great step in the right direction for real-time mobile multiplayer support! Although maybe that’s just my ignorance getting ahead of itself. I love this idea, and while I’m sure it’s a TON of work, it would make networking so much easier for many of us.

AAA/published developers will use the Microsoft and Sony and Nintendo APIs.
A/funded indie developers will use Steam.
Mobile developers use Apple and Google services.
The Unreal Engine can be very CPU intensive for server physics, so if you want hosted game servers that are authoritative, you’ll have to pay a fair bit in fixed cost.
And if all you want is matchmaking and leaderboards, then you’re likely to go with Steam, or perhaps Xbox Live for Windows if you go the Universal Windows route.

How, exactly, will EPIC make money on building this? Who are the developer/customers that will actually pay enough for it to cover both hosting, operations, and the significant development effort?
Maybe they exist – but I have yet to see a reasonable market sizing for this.

+1

Unlikely, but would be a great idea. There are probably a whole lot of legit technical reasons not to, but are they insurmountable? Disregarding the significant initial outlay that is . . . But as for the “why bother” reasons, well vertical diversification and a little aspiration. If there was a distribution platform implemented, and there was enough incentive to use it (with proper quality control), why not? I mean if you are making games, and a game engine, and you have the ear of so many developers, and the probability is that there will be a few dozen decent games in the next year or two . . . bears thinking about.