Quoting: Doc Angelo

Quoting: TheSHEEEPOnce you recognize you have this problem, the solutions really aren't too hard to pull off. So... yeah, I somewhat doubt this is the (only or even main) problem here.

Is it really that easy to translate the game logic from floating point to fixed without changing the behavior of the AI and other things? Even the slightest change that only occurs in extremely rare cases would render it non-compatible.

Not easy, no, but not half-a-year difficult, either.

On the other hand, I've witnessed people take multiple months to implement a prebuilt 3D sound library, so... *shrugs*

View this comment - View article - View full comments

Quoting: Doc Angelo

Quoting: PatolaIs there any programming literature on this problem or is it always case-by-case?

As far as I remember, floating point operations don't always get 100% exactly the same results on each platform/compiler/processor. Sometimes, they are just a teeny tiny little bit off, which may or may not throw your game logic up. If you have deterministic game logic and never check if the results are the same on each platform, you have a hard time fixing that afterwards when the game is finished.

As far as I understand, Civ 6 is deterministic. In multiplayer rounds, only the player input is synced, and every system calculates the game state for itself. If one system comes to different results, the round is out of sync. That's why there was no cross platform MP until now.

That's pretty much the problem with floating points being transferred in multiplayer cross-platform. And that is actually very much common knowledge for any programmer worth their salt - but since nowadays most coders "learn" their craft in everything-is-done-for-you-boxes like Unity, of course this knowledge has actually gotten rare and whoever implemented MP in Civ6 just naively went with floats because "what could go wrong"...

But when you actually dig around and look for some guidelines about multiplayer programming, this problem is mentioned a lot.

What makes me sceptical about this actually being the problem here is that this would actually be somewhat simple to solve without taking months of time.
Replace all important float usages with fixed-point arithmetics (very slight performance and possibly memory impact), regularly sync the host state with the client states (especially doable in a turn-based game as those are rather slow by nature), ...
Once you recognize you have this problem, the solutions really aren't too hard to pull off. So... yeah, I somewhat doubt this is the (only or even main) problem here.

View this comment - View article - View full comments

I always die a little inside when someone says "PC" to distinguish it from linux and/or mac.
It's just so... wrong. PC describes the hardware, not the software.

View this comment - View article - View full comments

Quoting: BrisseWhat I'm curious to know is if you buy it now and play using Proton, does Feral still get their cut? Someone should ask on their Twitter, but I don't have a Twitter-account :)

In that GIANT Proton news discussion, I read a quote from one of the devs that how you play it within the first two weeks determines what platform it gets counted for. And Proton does count as linux - no idea about non-Proton Wine, though.

Don't ask me to find that quote, though.

View this comment - View article - View full comments

A Battle Royale clone that isn't good.
That's... not exactly something new.

Too bad this one would also support linux. But we don't really need bad games with linux support, so... nothing of value was lost?

View this comment - View article - View full comments

This is like that thing I was talking about to make Wine actually usable by most people - seamless integration and no need to install any stuff via winetricks.

Yeah, there is Lutris, but let's not fool ourselves: it just doesn't work.
Tried installing Wine Steam with it - doesn't even start, instead getting stuck at patching Steam. Even installing a vanilla Wine and running the Steam installer with that works better :S:
And the net is full with "just doesn't work" stories of Lutris.

View this comment - View article - View full comments

I have this game on my wishlist since a very long time.
It looks really interesting - and yeah, satisfying.

View this comment - View article - View full comments

Quoting: Doc Angelo

Quoting: TheSHEEEPAnd what the devs of Banner Saga did was just ill-informed, as Adobe announced dropping linux support very early, yet they choose to use AIR for a cross-platform effort.

If the Banner Saga devs would have started development before Adobe announced that they drop Linux support, would it have been a good idea, because Adobe is a big company with big money? What good is that if they drop support anyway?

Actually, it would have been a terrible idea anyway, because AIR is a terrible product.
If it was a good product, then it would have been a good idea, yes. Middleware dropping platform support for platforms that are actually alive (even if not really for games, back then) was almost unheard of - and still is.
And as I said multiple times by now - even if it would have been open source, the developers couldn't have maintained a +linux branch due to lack of resources, and if someone else would have done that (and of what quality) is in the stars.

Quoting: Doc AngeloAnyway... if everybody would make business decisions like you suggest, we wouldn't have games on Linux - because we wouldn't even have Linux. Linux was that purely "hobbyist" product you speak of. Companies have started to use it, and they succeeded. Is the whole history of Linux a "risky business decision" for you? Would you have advised to use Windows, because they have paid employees?

I'm sorry, but your idealistic viewpoint completely blends out reality and common sense.

It started out as an hobbist product, sure. But its advantages were so obvious that people started using it and then an industry started forming around it. Industry = money, and that money is what kept and keeps it kicking. Take that (partially closed source, btw.) industry away and we're back at square 1.

And in those crucial years of 1995-2000, when that industry formed, there honestly were no good server platforms available that came close. Running a Windows server was utter trash, and that was known.
Linux wasn't successful just because it was open source (though that may have helped), it was successful primarily because it did something better than everything else. In other words, there was no equal competition - and it was free, too. Whoa!
Yes, even then, picking linux was a risk. But a risk offset by a serious performance advantage.

You can blare about FOSS and idealism all you want, but if you have a choice between a small FOSS project that does what you want, but has like a handful of people developing it in their free time - and a closed source alternative that does the same thing and doesn't cost a lot, then you pick the closed source one if you can afford it and the support is known to be good.
Because that gives you access to people who are paid to aid you in your requests - while FOSS gives you the chance that nobody does anything about your issues or the whole project gets dropped.

I'm not against FOSS, don't get me wrong. There are many FOSS projects that are just outright better than possible closed source alternatives - SDL comes to mind. What matters in the end is the quality, though, not if it is open source or not.

The best case scenario is something like Godot which combines financial stability with open source. It doesn't have the plugin ecosystem of Unity yet, but that is something I grew very cautios about as those plugins are often just shoddy and end up causing more harm than good.

View this comment - View article - View full comments

Quoting: Doc AngeloSituation B: You are using FOSS. The devs of this software are stopping support for it or are removing features you depend on. You are now allowed to take the source code and maintain your own fork of it that doesn't remove what you need.

That a FOSS project removes functionality people are using is already less likely to happen. But even if it happens, you are not without options. That is a really, if not the biggest, argument for FOSS in businesses.

I marked the hopelessly optimistic part.
That is just what I wrote. Sure you could do that in theory, but most devs, especially small ones, simply do not have the resources to do that.
An option is worthless if it is purely theoretical. For all practical intents and purposes, there is no difference between both situations.

Quoting: Doc Angelo

Quoting: TheSHEEEPdo you truly want to place your products in the hand of some hobbyist who might just drop the whole thing again due to lack of interest or free time?

If people wouldn't do exactly that, we wouldn't have gaming on Linux.

And which game that runs on linux actually makes use of free open source engines that are not being maintained by someone who is actually paid?
Compare that to the amount of games made on engines by people who maintain said engine for a living.

It's just not a wise business decision to place your own products in the hands of pure hobbyists.
You also cannot compare that situation to software like Ubuntu, which is open source, but its developers (at least partly) do earn money with the ecosystem around Ubuntu.
Godot is similar in that the main devs do earn their living with the engine.
I consider both Ubuntu and Godot very safe to use, as they combine the open source benefits with the stability of people actually being paid.

But placing your software in the hands of projects that do not have a monetary backup means letting lady luck decide, and that is risky at best.
And what the devs of Banner Saga did was just ill-informed, as Adobe announced dropping linux support very early, yet they choose to use AIR for a cross-platform effort.

View this comment - View article - View full comments

Quoting: ArehandoroOn the other hand, about the 1/4 of gains from the cost... well, it's ta vicious circle. If there isn't a proper supported port, there aren't sells. If they don't sell, the port isn't profiterol* and so on.

A lot of linux gamers likely also didn't buy the game because part 2 and 3 are not on linux.
What would be the point, really?

View this comment - View article - View full comments