Minutes of the first CG workshop: what holds back web gaming?

Hi CG participants,



First CG workshop took place a couple of weeks ago. Notes taken by Noël 
during the call are provided below. The transcript is also available 
online in the usual document:

https://docs.google.com/document/d/1CQtWyYPOjLy7np7SW-exbW20SXv61YaStBBVi042oHc/edit#heading=h.eng6u8j06uml



Many thanks to those who attended the workshop. It was very good to 
exchange on ideas, let's see how we can progress them and do it again!



Francois.



-----

Topic of the workshop

The web platform has come a long way since the days of Flash. Megahits 
such as Candy Crush and Clash Royale or indie games such as Among Us are 
completely achievable on the web today. So… why don’t more game 
developers build games for the web? In other words, what holds back web 
gaming?

Points made by participantsKoen Bollen (Poki)
1- Developers don’t know what is possible to achieve on web

2- Real time multiple is hard (WebRTC might help, but it is not 
mainstream yet)

3- The code open for everybody to steal

Fabio (Godot)
4- The API are different from more classic game development (for 
example: loops don’t work the same way) . It makes it hard to port.

5- API are not as stable as they used to be in the past (old games are 
breaking)

Paul (OP Games)
6- Tooling is not at the same level as native game development

7- Loss of flash made it harder to get on the web, and nothing is at the 
same UX level. What could replace it?

Andrzej Mazur (js13k game jam and more)
8- Non-technical side: discoverability and monetization. How to make 
money and reach an audience?

9- The ecosystem needs an equivalent to Steam. There are many publishers 
but no central place.

10- On a positive note: technically, it is much better than years ago.

Q&A and free chat
Question from Francois: how are native games doing multiplayer games? 
Technical answers by Koen and Fabio (note: unfortunately, a bit too fast 
for me to take notes)

Fabio: UDP is not directly accessible from Web. WebRTC helps but it is 
complicated.

Francois: Developers could use Web Transport (client to server) but 
there is a need for peer-to-peer, and WebRTC is better for that.

Fabio: adding another protocol (Web Transport) might bring more 
complexity.

(Fabio also raised concerns about API that could stop working after some 
years, as he experienced it in the past)

(Note we tried to explore points #1, #9 and #7 a bit deeper)


1- Tom: There are not enough success stories of games made in HTML5.

Koen: web games are not “cool”, and there not enough high quality games. 
There is an image problem problem with web games in general.

Andrzej: we have seen impressive tech demos, but no huge hits following 
those. And agreed, not enough success stories.

Koen: cross-play would be helpful/motivating (need for a success story 
here too).

Andrzej: some games were ported to native. Having the same game on Steam 
brought credibility (wow effect).

Fabio: tools are super important. Developers are making games, not 
specifically web games. They just want to be able to distribute 
anywhere.





9- Koen: Poki aims to become the Steam of web. Selling games (like 10$) 
is not common on web. How do web developer make money?


About monetization: Poki looked at Coil.


Andrzej: monetizing with ads is the only thing that works at the moment. 
But we are pushing web monetization hard. It is difficult to make it 
happen (people are not convinced yet). We really need to find 
alternative to ads.

Tom: Youtube monetizes with ads, but also with sponsorships and 
patreons. Can we learn from it?

Andrzej: All the options already exist, but advertising is still 95% of 
the revenue. The problem with patreon: it is served by a unique company, 
everyone needs to go through that website and all. Side note: Coil is 
experimenting with tipping.

Paul: There is no standard for in-app purchase.

Koen: At Poki, we don’t want to have to handle our own wallet.

Andrzej: Coil, with the tipping tech, is working on in-app purchase.

Noel: alternatively, what do you think of bitcoin mining while playing 
as a way to monetize?

Andrzej/Koen: It was tried before but got abused. Maybe it needs 
refinement.

Koen: There might be a problem of transparency. Asking for money is more 
transparent by default.

Andrzej: But it is great to bring it up, we need original ideas to 
challenge the possible ways to monetize.

Koen: We already have examples of working alternatives. Twitch works! 
They make money from people even if nobody must pay. Huge fans are 
paying.

Andrzej: Patreon model for the web also has a huge potential. We need to 
find ways to make it work.

Fabio: The platform itch.io is quite used and monetize well (to pay the 
developer directly is possible, redirection to the native app is 
possible, etc.)

Andrzej: yes, itch.io is experimenting a lot (including with web 
monetization). An asset marketplace also available.

Koen: it feels more targeted toward developers than toward consumers.


7- Paul: Before there were a lot of creators using Flash. They are not 
successful anymore. Why?

Flash was used by artists, that was helping getting more people to make 
games.

Koen: Web is not seen as a gaming platform, it does not attract “real” 
developers.

Fabio: When in reality it is just another platform. Game engines should 
invest on it. People want to make games, but they don’t know the tricks 
of the web.

Koen: It seems Unity/Unreal don’t see it as an opportunity.

Fabio: on a side note, 99% of games on Steam are made with game engines.

Paul: so are game engines the problem?

Fabio: it is true that the resulting applications run much worse than 
native ones.

Koen: In Unity, using“export to web” often results in something not 
workable.

Fabio: From the game engine side, it is hard because there are a lot of 
limitations than come with web.

Andrzej: the unity team explained that there are ongoing efforts to make 
the export better, plus unity tiny is being improved.

Tom: They expected Unity and Tiny Unity tech to converge. Now it is not 
clear where it is going, especially project tiny. The main use case for 
it is playable ads.

It is far from what it needs to be.

Koen: At Poki, we rarely see Unity games perform well unless they are 
big IP.

Fabio: Godot has a 12Mb footprint on web, but it still too big. Then 
loading assets usually adds something like 300Mb, it becomes super 
heavy.

Koen: games on web load assets on demand, which is not the case in 
traditional game development.

Fabio: True. It is a different tech, making it hard to port. Example: 
access to cache, etc.

Tom: Game streaming is a brilliant solution for games. But will it 
succeed?

Koen: The main problem is input delay. It is difficult to avoid it.

Conclusion
Tom: It was a great time. Next time we should tryto get more actionable 
outputs though.

Francois: having a marketing campaign to raise awareness could be a 
great idea (see screenshot shared at the top of this email).