FreeGameDev

7/27/2019 - FGD was down for a day due to an IP blacklisting error. That gave me time to set up a sister site: mobilegamedev.org

This is not intended to fragment the community, but rather to provide a place to discuss open source mobile development and hopefully grow the discussion beyond the piecemeal hearsay the subject currently enjoys.

-- Original post --

Maybe this is a half-baked idea, but I'm thinking a few things here:

* The issue of the GPLv3 disallowing distribution on the iOS app store was mentioned on the SuperTux board, and the guy sounded pissed about his game finding its way on there. Maybe that platform is lacking in FLOSS.
* The F-Droid forum is relatively low-traffic, powered by Discourse, and not specifically related to games.
* The Debian Android tools are hardly ever used AFAIK, which leads to Google's Android SDK; bad for people of our local mindset developing for Android.

If there are people around here who have an interest in mobile game development, maybe those discussions could have a home. What do you folks think?

Would be a bit ironic to have such given that this outdated phpBB is not mobile friendly.

I am not against the idea, but I think they can be discussed like all other games in the normal forums, no? Why seperate them out?

Regarding the forum software, I still don't understand why you can't update to a newer version of phpBB while keeping the posts and user accounts.

As for categorization, you may be right. Trouble is, everything here seems to just be for Linux, with a few recent exceptions. There are other platforms in the world. On TIGSource, people will work on various projects for various platforms and talk about them in the general categories, but around here I don't see much platform diversity. I don't know why that is.

fluffrabbit {l Wrote}:GPLv3 disallowing distribution on the iOS app store

How does GPLv3 prevent that? Just curious, couldn't understand it from the linked post.

mteufel {l Wrote}:Due to the GPL it is illegal to distribute SuperTux via the App Store.

Must have something to do with App Store DRM violating Stallman's religion. However, that doesn't stop the distribution of more permissively licensed software on iOS, nor on Android.

If I recall correctly Apple requires some non GPL compatible piece of SDK code to be used for compiling software for their appstore, bit I might remember wrongly.

The forum is so outdated by now that upgrading would likely require manual intervention at the SQL database level, and both the database and the forum itself are hosted on a server I don't have access to. The people who do, don't have time (and motivation?) to go through the likely hassle of upgrading, so things stay as they are.

Splitting forums makes sense if the volume of posts is high enough to justify it, here we don't have much mobile game dev discussion, so no need. But it is a bit of a chicken / egg problem

Ugh... License compatibility. GCC is license-compatible with Windows because it grants an exception for the standard library. In the case of independent developers using the GPL, often I think, such as with SuperTux, disallowing distribution on iOS or similar app stores is intentional on the part of the developer.

Okay, screw forum upgrading. SQL is messy and while it might be worth investigating, you're more of a devop than most folks and I can understand their apprehension.

Chicken and egg, yes. I propose a new movement: Open source software development discussion that is inclusive of mobile platforms! Fight for freedom!

Another forum I'm on about RPGs adds new forums when a game reaches a certain level of discussion. I believe it is 5 threads with 10 different people contributing. My suggestion is to simply start talking about mobile development if that interests you. I played with the Android SDK at one point but never did anything. At the time I wanted to make Color Monsters for it. However, the Android API didn't seem very condusive to that idea. Seems like games run native code on the devices in a lot of cases. Varying screen size seemed like a hassle as well. Have you done any mobile development? Are you using native code on Android or sticking to java?

Don't steal my trade secrets mannnnnnn.

Seems like games run native code on the devices in a lot of cases.

That is definitely my plan. Kotlin is an interesting idea, but I'm too deeply in love with C to switch to running everything on the JVM. Getting native code to work with Debian's Android SDK is tricky though.

Varying screen size seemed like a hassle as well.

The modern best practice is to not limit your games to a specific aspect ratio. I made that mistake for an Android puzzle game, and let me tell you, it's really jarring to look at its screenshots next to other apps that better follow best practices and don't force everything into a 4:3 window on your phone.

Have you done any mobile development?

With the Android SDK and Android Studio from 2012 to 2013. In the years since, I've poked at it, but it hasn't really gotten easier.

My personal experience is stick to native code. So far I'm always using SDL2 (and the Android specific function in it). The difficult part is not the API/SDK level, but the optimization of game play and UI for touch-screen control.

acme_pjz {l Wrote}:My personal experience is stick to native code. So far I'm always using SDL2 (and the Android specific function in it). The difficult part is not the API/SDK level, but the optimization of game play and UI for touch-screen control.

I'm still stuck with compiling a "hello world", but I definitely get what you're saying. Back in the day, I played around with Unity, and the first feedback I got on a simple Doom-like FPS was "these touch controls suck".

But the Android API is weird also. For instance, what the hell am I running when I call super.onCreate( savedInstanceState )? If it's so important, why am I even given the option? Is an app developer expected to override parts of the operating system? Coming from less object-oriented programming with simple callback functions, I find Android's way of doing a lot of things strange. Java is a difficult language, but at least it's well structured. It's not as screwed up as C++. As a language. But the Android API kind of ruins the elegance of it IMHO.

I had a look at Puzzle Boy, and it looks like a functional repo, but very complicated. It's also pre-Gradle and locked into Google's SDK. I set up a toy repo to figure out the Debian droid SDK here:

https://github.com/fluffrabbit/debian-android-simple

I'm glad the Google repos aren't blocked in China, but I stand by my statement that the SDK Manager paradigm is a dumpster fire. I'm trying to use the alternative.

Yeah, mobile games. Maybe I can finally get that kick the buddy replacement and a better free software spaceship shooter. I just want a few nice mobile games. I tried to play battle for wesnoth on mobile and it wasn't great. I think with the very different controls, and the fact I want simpler, more casual games for mobile. (things my son could play without wasting money on my credit card, although I would be willing to support development, I don't want a bunch of micro transactions, or crazy amounts of ad videos.

Also, I think for the forum, much as I love this forum, I think it would be better to start a new updated forum, direct new topics, and post over there. Give it a few weeks to a month to start filling up with posts/topics, and then start by disabling new topics here and then archive this forum to read-only.

I'll host the forum. Will update when everything is set up.

fluffrabbit {l Wrote}:I had a look at Puzzle Boy, and it looks like a functional repo, but very complicated. It's also pre-Gradle and locked into Google's SDK. I set up a toy repo to figure out the Debian droid SDK here:

https://github.com/fluffrabbit/debian-android-simple

I'm glad the Google repos aren't blocked in China, but I stand by my statement that the SDK Manager paradigm is a dumpster fire. I'm trying to use the alternative.

In fact I don't really know how to setup a build script for Android properly just by trial and error. Basically I just copied the Android template provided by SDL2 package.

I hope one day the build script for Android can also be generated by CMake, which is my preferred build script generator for desktop platforms (Windows/Linux/Mac).

Jastiv {l Wrote}:Also, I think for the forum, much as I love this forum, I think it would be better to start a new updated forum, direct new topics, and post over there. Give it a few weeks to a month to start filling up with posts/topics, and then start by disabling new topics here and then archive this forum to read-only.

We were at that point all ready to go with a nice platform (Flarum), but then people complained a lot about its Javascript use so I honestly lost motivation and removed it again. The problem is that every currently available FOSS forum software that would be a real upgrade to this one requires JavaScript to function reasonably well.

If/when there is some alternative that runs better without javascript I will definilty consider it, but right now forcing a new system with mandatory javascript would split this already small community.

Julius {l Wrote}:then people complained a lot about its Javascript use so I honestly lost motivation and removed it again. The problem is that every currently available FOSS forum software that would be a real upgrade to this one requires JavaScript to function reasonably well.

Discourse is particularly awful in that regard, almost anything is better.

I should maintain a list of forums that don't require awful amounts of JS on my website that is JS-based. (I'm not making fun, traditionally forums do NOT require JS for normal use and I would also hope for it to stay that way.)

Of course there are ones that don't require it (like this one), but what is the point of upgrading if the fuctionality stays more or less the same? I think we are better off waiting a bit longer for an option more people can live with.

Edit: mainly the responsiveness required for nice mobile view is usually depending on javascript.

I set up the site. It can be used without JavaScript and works on mobile. Original post updated.

Needs a section for non Android and iOS mobile platforms like KaiOS, Ubuntu Touch, Plasma Mobile etc.

Julius {l Wrote}:Needs a section for non Android and iOS mobile platforms like KaiOS, Ubuntu Touch, Plasma Mobile etc.

Done. Please forgive my lack of creativity; it's been a long week. Any suggestions on what to call it aside from "Other mobile"?

EDIT: After a little reading, it seems that KaiOS is huge in India, the most populous country in the world, Ubuntu Touch is community-maintained, and Plasma Mobile is hardly worth mention. Maybe KaiOS should get its own board and we should all start learning KaiOS.

KaiOS games are just HTML5. It's a fork of FirefoxOS so all apps are web-apps. It might become more popular globally, but I rather suspect the Chinese manufacturers will drive down prices for Android capable hardware to a point where KaiOS phones lose their current cost advantage.

I love China so much. Cheap cheap cheap CPUs make fabless manufacturing costs almost negligible. There is also a basic version of Android for the Indian slum market which doesn't have Google's strict hardware requirements. (Android Orange? I don't remember what it was called.) HTML5 actually costs a lot of CPU power.

In a more realistic world, we might still be using ARM chips without hard floats, costing $20 each and built into phones that more closely resemble toasters. Even on that kind of hardware, C code runs super fast as evidenced by small3dlib. Maybe in the future the carrier subsidies market will collapse and low-end phones will be sold at cost or at a slight profit. Then Moore's Law will plateau for the mobile market. Funny how people don't understand the current business model. I see it all the time with the $10 Android devices at Dollar General. Poverty is the only thing you can count on in this world aside from death.

Julius {l Wrote}:

Jastiv {l Wrote}:Also, I think for the forum, much as I love this forum, I think it would be better to start a new updated forum, direct new topics, and post over there. Give it a few weeks to a month to start filling up with posts/topics, and then start by disabling new topics here and then archive this forum to read-only.

We were at that point all ready to go with a nice platform (Flarum), but then people complained a lot about its Javascript use so I honestly lost motivation and removed it again. The problem is that every currently available FOSS forum software that would be a real upgrade to this one requires JavaScript to function reasonably well.

If/when there is some alternative that runs better without javascript I will definilty consider it, but right now forcing a new system with mandatory javascript would split this already small community.

I'm still not sure where in the holy book of emacs Javascript is written as an evil that must be abolished at all costs. Sure, its SAAS, but if users can download the code AGPL3 style, how is that worse than anything else? Its not as if the purpose of a forum is to use software offline all by yourself. Its almost like saying, I should be able to send an email and never connect to the internet or any kind of network, and someone else is supposed to receive that? Um no.

Jastiv {l Wrote}:I'm still not sure where in the holy book of emacs Javascript is written as an evil that must be abolished at all costs.

I don't think of it that way at all (and I wrote one of the holy books of emacs. It's not an official one, of course. It's got "Church" and "Emacs" in the title.) Too much JS is a horrible thing. It's bloat. It's stuff the user (FSVO user) doesn't want. On the other hand, sometimes it's totally alright, but a little goes a long way.

I think the main thing is that most people get really tired of having to buy a new machine every 3 years just to visit a website that didn't need a bunch of flashy nonsense added, while gamers are going to upgrade just to play the latest games. Also sometimes JS is used to do things that "take over", when you just want to read a page. But really JS doesn't kill websites-- designers do. JS just gives designers a lot more rope to hang it with and they often prove can't be trusted with it, but poor JS gets blamed instead.