Jump to content
Josh

Let's talk about Linux support

Recommended Posts

First, Linux users account for less than 1% of sales, so Linux is not making any money for me.

There does not appear to be any way to make the editor run in Ubuntu 18.04 without crashing. Something changed in GTK and the application now crashes when you try opening a file.

Building the editor and engine for Ubuntu 16.04 requires two hard drives with two different installs of Ubuntu (32+64 bit).

Building the editor for Ubuntu 16.04 and the engine for 18.04 requires three different hard drives with three different installs of Ubuntu.

It's clear that Linux support will not be what I hoped it would be going forward with Leadwerks 4.x, due to decisions that were made early on to rely on BlitzMax and GTK. Too many things change in each version of Ubuntu and the rug keeps getting pulled out from under us.

I do not regret using BlitzMax for the Leadwerks 4 editor because it allowed us to try out many different designs before settling on the workflow we have, which most users say is something they really like about our system. The reason the workflow is so nice is because I tried out about three different designs for every single aspect of the editor before choosing the one I thought worked best, based on both my own opinion and customer feedback.

At the same time, the requirements for attention to Linux are more and more, and the money it makes is nothing.

I'm more optimistic about Linux support in the next engine because we can use the lessons we learned here to make better decisions next time. Most importantly, it will not use any middle libraries except the bare minimum, and everything will be pure C++.

Going forward with Leadwerks 4.x, what should I do? I don't have a lot of options here. We can either continue to make Ubuntu 16.04 the official supported distro or we can drop support for Linux entirely in 4.x.

Share this post


Link to post

Sounds a lot like mobile back in the day. If it's not making money and is a pain to support then it should probably have the same fate.

Share this post


Link to post

I'm not a Linux user so I don't think I get a vote but yeah, if it's just work and makes next to no money then I guess there's no point.  I'm surprised you're considering still supporting it in Turbo, unless it's really that much less work.  And I say that as someone who bought Leadwerks for both iPhone and Android (and a Nexus pad for development) and was initially very disappointed by its discontinuation.

Share this post


Link to post

I've given this a lot of thought, and although I love Linux, it's not my platform of choice for game development. I use it for personal stuff and play with SDL2 but that's it.

I've compiled a suggested roadmap which you may consider.

1) Roll back to Ubuntu 16.04 for the complete user experience.

Leadwerks (Standard Edition) works as intended out of the box with Ubuntu 16.04. This is most likely your main user base so it would be kind of a shot in the foot to disrupt them. All you really need to do is to make sure games compile with no issue and replace Code::Blocks projects with ether my CodeLite template or CMake. Speaking for someone that really just wants a Linux executable, I don't use the Linux editor as it's inferior to the editor found on Windows. If my game compiles on 16.04, but runs on 18.04, I'm set. I have a VM set up to do this.

2) Keep Linux support until the next paid update

I recall on our chat two weeks ago that you plan on a Leadwerks 5 (Not Turbo) and how you may consider that a paid upgrade. I suggest keeping the 4.x supporting Ubuntu 16.04 and skipping (OR having a "build only" support for people like me.) with Leadwerks 5. This way a free update doesn't disrupt the sadly few Linux users, they would have to opt-in for a Linux-less engine. If you include only the build libraries, we can still do Linux ports without you loosing your mind about the editor/GTK!

3) Bring it back Turbo.

Cut as much middle ware out as possible and all dependencies are under the app's root folder. This not only makes it easier on you, but for game developers as they don't want to update their 4 year old game because a new version of a distro made it not work. I also suggest going all the way down and build on core Debian instead of Ubuntu. There are lot of concerns in where Ubuntu is going so you might just want to fall back to the root and work your way up.

As much as I love Linux, It would make more sense to support only Windows and MacOS on Turbo's launch. I only suggest MacOS as it would be cutting in on Unity's terf, and my 2017 iMac needs something better to do than manage music collect dust.

 

Linux is still a viable platform and I consider it a bomb shelter in-case anything else happens with the other operating systems and/or their companies that own them. The more applications that run on it, the more I consider using it as a daily driver.

Share this post


Link to post

I think having linux support is good for sales even if just 1% use editor from linux, many want the linux build.And most other important engines support it.

Heres an option:

Keep the editor + engine supported for 16.04. For latest ubuntu lts version have one build without editor.

When 16.04 reach end of life (2021) discontinue support for editor linux.

I dont know how long you intend to support 4.6 , by the time linux editor is discontinued i maybe turbo will be usable.

 

I hope turbo will not have this kind of problems, linux is a really good platform for gaming.

Share this post


Link to post

Okay, thank you for your input.

  • The editor will continue to be built on Ubuntu 16.04.
  • The engine will be built on Ubuntu 18.04. I checked and this builds successfully on 16.04 as well.
  • The Steam store page no longer lists Linux support, although the program can still be launched on Linux.

In the future, I think good Linux support will be much easier because we will not be relying on BlitzMax, GTK, or Scintilla. There was really no way we could have done this any other way. I don't think C++ smart pointers even existed when I started writing the current editor. So we will learn from this and move forward.

I know I go back and forth on this a lot, but I doubt there will be a paid update before Turbo arrives.

Share this post


Link to post

Sounds like a plan. Are you still interested in replacing the Code::Blocks file or is Leadwerks for Linux on legacy support going forward?

Share this post


Link to post

Yes, I just don't know whether to go Codelite or VSCode.

  • Like 1

Share this post


Link to post

I saw the Google Stadia announcement, cloud gaming, and its running on Linux, with Vulcan graphics, Havok. They seem to be talking about 2019 launch. Do you think this will be a game changer, will it make Linux more popular, and require more devs to be familiar with using a Linux platform?

Share this post


Link to post

There are a few companies doing this same idea. NVidia has it's own game streaming service in beta right now. However they work with your games that you own. You give them your steam/ubisoft/epic/whatever launcher id/pw and they validate you have the game and then it'll launch in the cloud for you to stream play. The benefit is this is a better transition from where we are then Stadia is. Not to mention NVidia should be able to figure it out well given they can make custom gfx cards for this if they want.

Share this post


Link to post

been  done a few time but the problem is bandwidth and not everyone got the bandwidth that this type of service need.

Share this post


Link to post

It doesn't need to be everyone.  I'm sure they've crunched the numbers of how many people have a good enough bandwidth, how many locations across the U.S. or world they would need to set up (so that latency is minimized), how many people would actually pay for the service to cover their expenses, etc. etc.  Not to say that they can't be wrong and it fails anyway but it's safe to say they have the numbers.

It used to be that people doubted things like Netflix would work out because people didn't have the bandwidth for that either.  And it became immensely successful.  Similar with YouTube 1080p (and 60fps and 4K) streaming, Twitch, etc.  The world is rapidly changing for the better in this respect and it's not gonna slow down.  More and more people will try to make dramatic things work that we couldn't have imagined even just a few years ago.

Share this post


Link to post
22 hours ago, gamecreator said:

It doesn't need to be everyone.  I'm sure they've crunched the numbers of how many people have a good enough bandwidth, how many locations across the U.S. or world they would need to set up (so that latency is minimized), how many people would actually pay for the service to cover their expenses, etc. etc.  Not to say that they can't be wrong and it fails anyway but it's safe to say they have the numbers.

It used to be that people doubted things like Netflix would work out because people didn't have the bandwidth for that either.  And it became immensely successful.  Similar with YouTube 1080p (and 60fps and 4K) streaming, Twitch, etc.  The world is rapidly changing for the better in this respect and it's not gonna slow down.  More and more people will try to make dramatic things work that we couldn't have imagined even just a few years ago.

It is my belief that stream playing games is the future and will be here in the next 10 years as a good quality (things are still shaky right now) at min. However watching video is completely different than what this is. The input lag has been the main issue with this stuff. Having enough bandwidth to send the result frames to a person was only a matter of time in speed/bandwidth which was inevitable as things are always getting faster. Not experiencing input lag, especially for fast twitch online games, can be a challenge since your input has to take an extra few hops now. Before it was from your PC to the game server. Now it's your PC to the cloud streaming your game, to the game server. In this model it would make sense if the game servers were actually in the same datacenter as the cloud streaming your game to make that step as fast as possible but this would be a lot of coordination with game devs and leave out the little guys game trying to compete in this space most likely.

Share this post


Link to post

“Her der, there are people who don’t care about graphics enough to spend money on a graphics card. Therefore I will target them to buy a service they will pay for to get good graphics.”

  • Confused 1

Share this post


Link to post

To me it's more about the convenience and perceived expense. If I can pay $20/mo to play top end games all the time and cancel at any time then that's a good deal considering a top end gfx card costs around $300+. I mean if I can play in a browser on a chromebook what cost me $200 total then I call that a win!

Share this post


Link to post

Ill think that the supporting of Linux distros will be more eficient if the core targeting will be not the Ubuntu as the official linux distro.. cost as mentioned before by "reepblue"

and actualy his words was "..There are lot of concerns in where Ubuntu is going..." As we see the release with realese have a bunch of differences as a problem with libs and brokens depedesies and as a performance isues and differences.. so i think this isues  will dissapere ar will dramaticaly dropt if the Debian core will by selected as a major distro target cost its a parent of many distros and builds like Ubuntu/LinuxMint/MX/Elementary, and etc..)

Share this post


Link to post

Ubuntu is the obvious choice , what matters to me is that it just works out of the box.

And ubuntu is debian at its core, lately they even droped that unity ui and using gnome as most other distro.

And fact is used all over the place in industry as development desktop, and as cloud os is good.

 

Is normal for distros to update the compilers and libs , applications need to take care to be updated is not os fault, and ubuntu is pretty conservative with their lts releases.

 

About Stadia think google will have problems to get top games in there , big studios and the console bunch dont think they like the idea.

Share this post


Link to post

Being a Linux user I certainly appreciate the effort  to support my platform and I try support as many commercial software companies that do so as I can, by purchasing their software. But I also understand the difficulty and frustration on the developers side. There has to some kind of compromise. Setting a certain LTS release as supported (such as Ubuntu 16.04) instead of trying to support every release should not be a big deal. As the end-user realizing the effort behind developing for a platform that hardly makes any money I'd rather be dual booting an older OS release to run the software in these cases. At least my platform is supported.

Share this post


Link to post

From a Leadwerks Pro perspective.

Linux is a hell of a lot easier to deal with.

From a Leadwerks perspective, the Leadwerks UI on Linux is no where near the quality of that in Windows. It's this fact alone that's probably hurting Leadwerks Linux. If I need to make a map change. I dual boot into windows to make changes to the map.

I have been on Leadwerks Linux for the last year

I currently use kubuntu 18.04.

Share this post


Link to post

The editor is gimped on Linux, and you can't collapse models (not even with the command parameter, the model will crash the editor). 

Share this post


Link to post

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...