Google is releasing a new platform for HTML5 games. Could this work with Defold?
https://blog.google/technology/area-120/gamesnacks-brings-quick-casual-games-any-device/
1 Like
It’s HTML5, so I would guess so?
Facebook instant games probably have very similar restrictions on speed/size.
What are your concerns?
1 Like
The blog post doesn’t reveal much of the technical requirements. Do you know more?
No concerns, what I’m experiencing is EXCITEMENT! 
I do not regrettably. I was hoping the good people at King might!
I saw the same blog post as well a couple of days ago. I’ve reached out to Google to ask for more details. I’ll share more if and when I can.
4 Likes
Looking at the text they mention game loading time being just a few seconds on a 1Mbit network. With 1MB Defold game that already would take more than 10 seconds, so I guess Defold won’t be a perfect fit.
One thing that is a bit interesting is that their showcase games aren’t that small:
https://storage.googleapis.com/gamesnacks-showcase/jewelishblitz/gameCode/jewelishblitz/index.html
The Chrome dev tools report that this game downloads about 10Mb of data.
This game is a lot smaller though (1.1Mb )and I guess more in line with what they have planned:
https://storage.googleapis.com/showcase_games/tower_v2/index.html
Most of the launch titles are created with Phaser it seems like. Like @britlz says the games on their site are over 1MB generally. The Phaser JS is nearly 1MB on its own.
Defold could be absurdly ideal for a project like this if there was official caching of the HTML5 runtimes like I used to try to provide until it became apparent it was futile to try to do. If you wanted to put Defold based games on a site like this ideally you would want them all running on the same exact engine runtime file so it is only ever downloaded once. That means freezing at a specific engine version for extended periods of time, and rebuilding every game when updating all with the same NEs, but the benefit is the runtime is downloaded once when visiting any page on the site, and then loading any game could actually nearly be instant. This is an advantage tools like Phaser have over Defold.
5 Likes
8 months later and I’m doing a review on GameSnacks because it being pushed by Google is still interesting. It being part of https://area120.google.com/ and it being in its current state gives me the impression it has been abandoned.
Video from March. I don’t see much noise that is not from February / March.
The main site has scrolling bars in strange places.
When going directly to games some of them have scrollbars instead of having the game fit the size of the view, which is also strange. This is when viewing the site on desktop Chrome. On mobile Chrome this is not a visual issue. On mobile the presentation of the site and games is generally good and shows its potential but I feel the site is still missing things that would help it do better, such as categorizing games into genres / tags. It would also most likely benefit by tracking gameplay data and uplifting the best performing games to the top, instead at the moment the list is alphabetical.
At least one game, billiardsclassic, appears to be broken.
It’s strange that going directly to the main domain redirects to https://gamesnacks.com/embed/ it used to be https://gamesnacks.com/games but that doesn’t work anymore, and also shows their 404 isn’t setup nicely / there are probably live links for the old link that are not being redirected.
If it did not show a notice that it was owned by Google I think anyone’s honest first impression of this site in its current form as being a low budget HTML5 reskin site that are a dime a dozen. The issue is the generic look of everything even if it does have some polish qualities on mobile. They need stronger brands. This is tough though because the new golden age of webgames isn’t here yet. Flash is finally dying this year. Most talent is focusing making native mobile games.
Checking over some of the games, most of them are larger than 1MB to load. Many of them take a while to load since their assets are all loaded separately on different HTTP calls instead of being bundled into a few larger files. Some games are as big as 8MB to initially load. I didn’t check all of them but average size is probably ~3MB. IMO they should pick a standard and stick with it, they should use tech that bundles games into a single file if doing so makes games load as fast as possible. Technically multiple files can be better now, but not when multiple files are all tiny and there are tons of calls to access them.
I don’t know if they ever included any Defold games but based on these standards there is no reason why they couldn’t have. Even with the lack of runtime caching, there are zero advantages any of these other games have over something made with Defold while a Defold game would have distribution advantages over many of them. It’s head to head with Defold being better in some cases most likely.
One of the games Google highlighted is Tower which is 1.2MB, it is graphically minimalistic, and a Defold version of it could easily reach that if not load even faster / have an even smaller footprint.
Here is a general domain review, this is good but not amazing. For comparison, Defold.com has a much better rating at 64.
If I were GameSnacks I would be trying to do for webgames what Stadia is doing for AAA streamed games. Try to get that level of polish, and try to get bigger titles (either paying to get them made, buying exclusive rights, or just aggressively looking for the titles and making deals). Both while not breaking the bank on data - it can be done. If I were Google, I would be trying to make the Newgrounds in its Flash glory days of webgames. It’s going to happen, but most likely independently, and one of the big tech companies will buy it up.
With the current selection of games, they are not bad, but they all look generic, and the lack of a persistent feeling seems like wasted potential to me. Like if the site had even a basic achievement system, auto generated a guest account and encouraged people to graduate toward a registered account to save their progress, that would be compelling to keep gamers there and playing more games.
I did not see any attempt at monetization on the site. At least it is nicely clean, but a lack of monetization means lack of incentive to improve it. It means those managing it don’t have a reason to really make it great outside of their personal interest to make it what they want it to be.
Here is the developer site https://developer.gamesnacks.com/
I remember filling out a form back in February and I did get a response but no reply afterward. To be fair the only web games I had up were premium ones ~100MB to initially download. Even premium games could work on GameSnacks following their premise of fast loading thanks to LiveUpdate and selective loading of resources.
The site does have a builtin high score tracker / share UI but even that I feel can be improved as it wants people to copy and paste share links instead of having the share UI right there for services. This could be done with navigator.share if it’s detected to be supported in current browser https://web.dev/web-share/ and general share buttons like Facebook or Twitter / picking others appropriate to geo-location would be better.
While doing more research I found https://gamesnacks.com/play/ which seems more interesting than the main site. Here you can setup lobbies and invite friends to join, then you have so long to play certain games, and can vote on which next games to play together. It’s a good use of what they already had. Strangely this is also on https://play.gamesnacks.com/play/ with a different session cookie.
Trademark status https://app.trademarknow.com/mark/us/88819616?authkey=v018sXcEWfzgAcvSLMuIkKJG5WME2hujPC9RkbNW8AGcm_Blk0
Verdict:
- Maybe it’s there but I didn’t see monetization of any type. Google could leverage its ads services to provide very good options to gamedevs on the service. They must have offered some incentive to the games they already have on their site. Maybe if the price is right it would be worth it, but without monetization happening on the site it’s never going to last, and the devs who already got on the site may be the only ones to ever see money.
- The domain has some authority in ranking but nowhere where it needs to be. It’s at 40~ it needs to be closer to 60-90~ to be competitive. It also needs genuine interest from gamers which I don’t see atm, I don’t see any of the share links showing up organically on social sites where you would expect them to be if the site was doing well.
- I don’t see users being sticky to the site and that’s a problem. Most of the traffic and links to the site are most likely news articles talking about Google making a webgame site “for slow mobile internet connections” which even that premise is questionable to me when all games are over 1MB and still have generic looks. I would expect super strict optimization at least. If you’re going to allow a 8MB game you may as well go higher, as high as app stores allow OTA which is currently 250MB I think?
- If Google puts its weight into promoting it it could be worth trying to get on more, but at the moment it’s evidently not. Unless they pay you a good lump sum putting your games exclusively on your own site is a much better idea if you want to build a brand long term. All of the brands on the site currently seem like throwaway generics/clones with no interest in long term development, which is a problem.
- It’s clear this service is meant to be Google’s answer to instant games / games that are meant to be embedded into chat apps. I don’t know if that ever happened. There’s also an implied emphasis on hyper casual games but without an understanding for why hyper casual games are working on mobile.
- If Google / those managing it follow through with it and be aggressive with making it great and putting great games on it I still feel it has potential but right now it’s a miss.
I said this before in previous post, if I were mega serious about making a service of games that ran instantly over bad internet connections I would pick a single game dev tool like Defold and mandate all games be made with it. I’d also require source access to them all. That way I could build automation tools for rebundling them, and making them all target the same engine runtime. This would mean the runtime that would be about 1MB could download once and be cached/precached in chat apps, and then games could be loaded as stand alone bundles that could make the Flash dream of games being as tiny as 25KB possible. That would be instant! And tools like Defold with selective texture compression could make making those super tiny sizes really possible. I don’t expect Google or others to do this, but if there was a genuine commercial market advantage in this kind of concept of low bandwidth / actually instant games that would be the way to do it.
9 Likes