Quoting: omer666

Quoting: MaelraneSorry, but if I was about planning a port, why would I start an engine with DX9?

Quoting: Avehicle7887"why not start with OpenGL in the first place?"

Because it is much easier to do so.

It's no wonder Kronos started Vulkan from the ground up, as developers all around have been complaining about the numerous flaws that affect OpenGL.

Each time you see a port being criticized because of poor performance, remember that there are some things OpenGL handles very differently from DX9/10 and are a real mess to port or to code.

That's one of the reasons why having all the major engines already ported to Linux is a priority if we want Steam's venture to be a success -- something that's already in a very good shape.

True, OpenGl has several flaws, starting at functions (several functions for the same usecase but to use differently with different performance outcomes for different use cases - hey, we know that from alsa already :D). The other bad thing is ressource optimization, it only uses a realy low part of the available ressources (CPU/GPU whise) and that's a design-based flaw (thread contexts / querying primarily), compared to DirectX/D3D < 12 is pretty much the same, if you know which functions to use in OpenGL. But there are not too many OpenGL experts out there, compared to D3D. I certainly know a bit, but I'm far from being good enough to port/write a whole 3D engine (as DrMacCoy is, kudus for that... my best effort for an engine yet was a iso/2d based engine years ago for a proof of concept).

View this comment - View article - View full comments

It looks so great... I'm still not sure if it's my kind of game or not and didn't buy it. I think I need an in-depth video of that game to make my mind up.

View this comment - View article - View full comments

Quoting: Guest

Quoting: MaelraneSorry, but if I was about planning a port, why would I start an engine with DX9?

Yeah, they should have used OpenGL or use an engine that they can update to Vulkan support, and use OpenGL/Vulkan for every platform. Pretty dumb to do anything besides that.

"updating" OpenGL with Vulkan is pretty much an own render path as well, that does not make a difference to DX9/D3D to OpenGL porting [at all]. It's still a fully own rendering path. Using an engine which will be capable of would have been possible, but I get the issue of why people often rather rely on their own engines than taking engines on the market.

An example would be Shroud of the Avatar: Going for the Unity engine being very limited in terms of Open World won't do the game any good, and is one of the reasons the game is really really annoying, despite their marketing you have a lot of zoning and a lot of waiting time.

If OpenGL was a viable choice on Windows, well, they could have started with OpenGL. There is a reason why most game devs don't and do take up the workload of going for a seperate OpenGL rendering path for Mac/Linux.

Vulkan isn't finished yet, and didn't have proper dev tools when they started, so it wasn't really a viable choice for an engine to start with. Now you probably could start, but still Vulkan is not done (yet).

As I see it, they got their priorities right. They didn't have the funding for a port, and concentrate on their Windows game. Maybe they'll have the money to do a port later, but at this time, they made the only right choice business whise and for their kickstarter badgers: Bring the game out for Windows before you can't afford the Office any longer, and don't dabble with side projects not funded.

It's not my kind of game, so I won't jump to support it - though, I still hope they can do it some day for all the users interested in the game.

View this comment - View article - View full comments

Ohhh :( ... that game was my plan for my christmas holidays ...

Though, as a supporter (and not actively taking part in the development feedback), I'm rather glad if they push out a good product later than a rushed and flawed one earlier. As long as it's not like Duke Nukem Forever :D. But that's a hard mark to beat.

View this comment - View article - View full comments

Proprietary Linux-Driver of NVidia is fine performance whise. That's really the ports having the issues ;-).

Even though, I have hopes of Vulkan will change it a big deal, since even the difference between Mantle / Metal / Vulkan / DX12 is not very much, it's basically not as hard to port a DX12 to Vulkan as it is to port D3D to OpenGL. Though, GNM (PS4) according to an engine-dev I talked seems to be a bit of a different story.

A good example of a well performing game (after the patch, before it had it's issues) is Shadow of Mordor. I was pretty impressed that a title like this would have that good performance on my 3 year old rig. The Metro-Series is a good example either, performance is pretty good.

We've had our fails, we've had our great ones. Let's see what's up for the future.

Anyway, for the guy telling he had a significant performance increase ditching KDE for Xfce. That's interesting, I've not had any significant performance increase ever switching a desktop environment. Just more RAM available ;-). Though, I'm using a .. very basic KDE setup here with a lot of stuff not being installed (the good part about the "minimum depends" depends in Arch philosophy).

View this comment - View article - View full comments

THAT looks like a game for me. Bought :D.

View this comment - View article - View full comments

It's a long-term attempt. The real attempt is not SteamOS / SteamBox but the SteamLink. They're trying to get into the living room market attrachting the Windows & XBox users. By streaming, not by SteamOS. SteamOS is just the playground for a bigger picture of them providing their own console. It's the playground to show off the market, before they go on their own. SteamBox has the same issue as PCs: Different specs. And that gap is gonna close some day if you ask me. They will allow others to produce, but sooner or later, we'll see a Valve SteamBox if the concept of SteamLink somehow shows off.

I guess this to happen a year after Vulkan. Let's see what Valve has in shelve.

View this comment - View article - View full comments

Bought it, and have to say, sadly not my kind of game...

View this comment - View article - View full comments

And if I bought it on Steam? I'd guess Steam knows I bought it :D.

View this comment - View article - View full comments

Quoting: Guest

Quoting: STiAT

Quoting: GuestBut Qt looks ugly or did it change?

That's bullshit, it's all about the theming. This one looks as if designed with QML. Anyway, if you look at standard themed GTK+ applications, they're ugly as hell ;-).
Anyway, Qt is at the moment the more professional and better (in sense of features and cross platform compatibility) toolkit compared to GTK+. Even long living GTK+ projects switched to Qt (Surface, LXDE, Unity8, Dropbox, Wireshark, VLC - to name a few prominent ones). Mostly they named the reason that cross platform in GTK+ was mostly a stepchild, especially the native look-and-feel on other platforms was just bad (and still is). Some even because they couldn't implement certain things in GTK+ without ugly workarounds. Sadly, QML Desktop elements do not yet have feature parity with Qt Widgets (as icons in dropdown menus).

The largest issues of Qt were (for a long time) the licensing issue not being LGPL, and that the libraries were pretty big, so you loaded pretty much a junk of things into memory. They changed that in Qt4 (licensing, Nokia) and Qt5 (splitting Qt into more libraries), which were really positive and long awaited moves.

Toolkit for me still is a developers choice. As long as the distributions and community manages the theming part, I'm fine with the developers using their preferred toolkit.

Doesn't GTK just adapt your GTK Theme.

If there is no proper GTK Theme for my KDE Theme that's rather hard. Same goes for Qt in GTK+ Environments, if there is no proper Qt-Skin it does not integrate nicely.

View this comment - View article - View full comments