So whatever happened to the whole PGDCE thing?

[AthenaOfDelphi]

Let's get this thread back on track.

First up, to those suggesting we might like to use other languages and tools.... this is a site about developing games with Pascal. I don't want to use Unity, C#, Game Maker etc. etc. etc. I want to use Pascal, just like many of the other visitors here. Suggesting we're wrong for this (and with some of the comments, intimating we're somehow mentally deficient because of this) is wrong. There are no technical reasons why you can't use Pascal to make games. Performance wise it's right up there, learning it is pretty straight forward and there is a good range of tools.

The project failed because people couldn't agree, it was overly ambitious and ultimately only one person contributed any sizeable amount of code.

There are posts that suggest we should make games not engines, and having read the post by Ian Mallett I generally agree... but what are you going to use to make a game? What's out there that is well documented, under active development and isn't reliant on a single developer?

This is the hole I was hoping we might, as a community, be able to fill.

[laggyluk]

What I was trying to say was that technology is secondary choice in making a game once you decide to do it. You should judge features and pick whichever engine fits your game idea best. Trying to do it in non mainstream tech basing on sentiment is wasting resources.

On other hand If you want to make game engines then pascal is perfectly fine choice.

[LP]

The project failed because people couldn't agree, it was overly ambitious and ultimately only one person contributed any sizeable amount of code.

No, the project failed because it was doomed since beginning - the article "Write Games Not Engines" that I've linked explains why.

What's out there that is well documented, under active development and isn't reliant on a single developer?

Maybe you could actually state why do you feel that existing engines/frameworks do not suit YOUR needs? That is, name a specific framework/engine name and the problems that you have with it. There are authors of frameworks/engines participating on these forums, myself included, so we'll likely be able to answer for ourselves and address the issues that you may have.

Granted, there are some engines like ZenGL that were officially discontinued (and it's sad as Andrey is a very talented developer, I hope he's alright and going strong), but please, there are plenty more that are recent and active.

This is the hole I was hoping we might, as a community, be able to fill.

I think this is where you are wrong - in fact, I think, there is a market saturation, so there are TOO MANY framework/engines to choose from, it'll be quite hard to find any hole to fill. I can't say for anyone else, but at least in my case, the framework that I've published and have been maintaining since 1999 (for almost 18 years already) is just part of the common code that is shared among my own projects, and it is a part of my business plan (consulting), it wasn't made just for the fun of it, but it came as side-effect (result) of developing other products. In addition to helping anyone else, I actually make good money with it.

[AthenaOfDelphi]

As much of what I'm saying is clearly wrong, educate me.

Assume I'm a beginner (I can code Pascal well, but in terms of graphics APIs I'm a newbie), I want to use OpenGL because I want cross platform (I'm using Delphi and I'm planning on targetting Windows, MacOSX and Android). I have a game idea that is 2D tile/sprite based (run jump platformer with some top down world exploration).

Which engine/framework should I use and why?

The selection criteria are:-

• Must be actively maintained and up to date in terms of API implementations
• Must be maintained by more than one person (I don't want to be waiting for bug fixes, when the only dev needs a break for his/her honeymoon for example)
• Must be well documented (for beginners)
• Must have a good suite of examples (for beginners)

So point me in the right direction.

[laggyluk]

Firemonkey? sad.

[SilverWarior]

Firemonkey? sad.

You are suggesting Firemonkey. Seriously? It doesen't ment any criteria that AthenaOfDelphi specified.

• Yes it is actively maintained but the problem is that FMX is not focusing on game development but instead of standard desktop/mobile application development. Also FMX isn't up to date with latest API's. And since we are talking about API's the main problem of FMX is that on windows it uses DirectX and on other platforms OpenGL, meaning that there is no guarantee that applications developed with FMX will handle the same way on all supported platforms.
• While FMX is maintained by multiple developers the problem is that their response time to bug reports is quite slow. Just check Embarcadero Quality Portal for many FMX based bug reports where some of them are even more than a year old.
• As for FMX documentation. I'm afraid to tell you that it is terrible. I mean how can you say that certain documentation is good if checking documentation for some component properties actually shows you documentation similar property of one of the parent components. Or even worse when you get "Embarcadero does not have any additional information on this subject"
• As for examples goes. Those that are available are again not focused on game development but on standard application development.

And main reason why I personally would not recommend FMX for game development to anyone. It is its rather poor performance.
Wanna see how poor performance it has? Create a new FMX project, place 100 panels on the form, compile, run and while running just do some rapid mouse movements across the form and observe the CPU usage.
On my 7 years laptop with 2 GHz AMD Turion X2 CPU and ATI Mobility Radeon HD 3650 graphic card having 100 panels on FMX form and doing rapid mouse movements I can cause full usage of one of the cores and FPS drops below 20. Now imagine what FPS drops I would be encountering in actual game where I would have to do a lot of other processing. And before you start saying that this is just because my laptop is worth s*** I should mention that on this laptop I can play Far Cry 3 on medium settings with average of 40 to 50 FPS .

[laggyluk]

You are suggesting Firemonkey. Seriously?

Seriously no but it looks like only one meeting the criteria and performance was not one of them

[LP]

[*]Must be maintained by more than one person (I don't want to be waiting for bug fixes, when the only dev needs a break for his/her honeymoon for example)

Unless the framework is provided by a large corporation such as Intel or Microsoft (and even in case of those companies, the following could apply), you might be very surprised to find out that a certain portion of an important framework or section, a single person is typically responsible of, and many times is the only fully qualified to do the appropriate maintenance. At least in the companies I've worked in, a senior developer responsible of large portion of specialized code, in the area of 2D/3D graphics, is difficult to find and/or replace (so naturally, it is typically a well paid, important position). And now, with introduction of new, complicated APIs such as Vulkan and Direct3D 12, a senior engineer for such position, that also happens to know Delphi, is even more rare, practically nonexistent. Therefore, I think that the quoted criteria point is unrealistic, sorry.

[AthenaOfDelphi]

Granted, I was responsible for a big chunk of my teams product line up at my former employers, and I had little to do with other sections, but we were all capable of picking up the others portions (as demonstrated by the fact I handed over that huge chunk of work - 7 years worth - to another member of the team when I left). People are not irreplaceable, the people who think they are make themselves more replaceable.

I'm talking about the framework/engine as a whole... just like lots of other open source projects out there.

So with all that in mind, sounds like an excuse not to give me a good candidate for an engine I could use as a newbie. But, as I'm feeling generous, if you can't meet that criteria, lets discount it and instead go for a 75% requirements fit.

Any takers?

[SilverWarior]

If you can compromise for meeting 75% percent criteria I think you should check Platform eXtended Library. It supports multiple platforms, multiple API's through multiple providers (DirectX 7, 9, 11, OpenGL, OpenGL ES). It works both with Delphi and FMX. Supports both 2D and 3D rendering. And finally it is open-source project so anybody can contribute.
If you chose to use this one I might be able to help you get started since I have some experience with its predecessor Asphyre Sphinx 3.

Another option that you could chose is Castle game engine or nxPascal (might need a bit work to compile with modern Delphi).