Open discussion: What game devs need from Delphi

[WILL]

But they could atract a lot of new developers by offering free versions for students similar as microsoft is offering free editions of Windows to be used by students. More newbies atracted means bigger chances for some of them to actually buy Delphi in the first place. Having 30 day trial is just not enough.

Microsoft's dev tools are shit. I know, I use the damn thing every day much to my begrudgery. I miss the days I coded intently in the Lazarus IDE, but it doesn't do what I need it to, so life goes on.

Free is not a viable option and I don't want a free version. I like tools that I can make games to sell. And I like software that has dedicated developers not volunteers that slowly make their way to a new version on their free time. It's that nice guarantee that tells me that I'm getting something of quality that will pay me back later for my hard earned cash.

That said, if I threw some free candy at a crowd of people you always have someone running after it. Doesn't mean there is a lot of value in cheap dollar store candy.

Another problem is lack of third party components and packages for Delphi. There are more and more excelent third party components developed for FPC/Lazarus but most of them doesn't support Delphi. Why? Becouse if someone wants to make their components Delphi compatible they need to purchase latest version of Delphi as trial edition time is just to short to be usefull.

Game developers don't need components. This is a concept for old Windows GUI applications so that any shmoe can make one with little to no code. Game programmers need to know how to program.

[SilverWarior]

I like tools that I can make games to sell. And I like software that has dedicated developers not volunteers that slowly make their way to a new version on their free time. It's that nice guarantee that tells me that I'm getting something of quality that will pay me back later for my hard earned cash.

I also like software that has dedicated developers instead of just volunteers that is why I have bought Delphi XE2 even thous many pepole advised me othervise. I'm also anually paying for "Support & Maintenance for Delphi" which gives me acces to all the latest updates for Delphi.

Now this is acceptable to me since I have a regular job and earn neough money for this.
But what about students for instance? They have no jobs to earn mone for this.

Embarcadero would say there is a Starter edition which is viable option for this. I have to strongly disagree with them on this. Why?
For the price of Starter edition they are offerin way to less:
- Only 32 bit Windows compiler
- FireMonkey components which have much worse perfomance than VCL. And since you don't have need to use FM on windows as you have on other platforms, probably only few pepole will even use them on Windows
- no code formater and and beutification for source code
- no class explorer
- no live code templates
- no symbol insight
- no code folding
- no class completion
- no resource DLL wizard
- no refactoring
- no UML modeling
- limited integrated debugging
- no database support
- and many other things stripped out
In short the starter edition even have some features stripped out even thou they were available in Borland Delphi 6 Personal edition which was in the end offered for free.

Game developers don't need components. This is a concept for old Windows GUI applications so that any shmoe can make one with little to no code. Game programmers need to know how to program.

True we don't need components we need libraries. These two concepts are quite similar.
So tell me how many Pascal based Graphical engines do we have available and how many of them have full support for the Delphi?
How many Pascal based Sound libraries do we have available and how many of them have full support for Delphi?
How many Pascal based Network libraries do we have and how many of them have full support for Delphi?
How many other Pascal based gaming libraries are available and how many of them have full support for Delphi?

So as you take a look closely you will see that most of them doesn't support Delphi or have verry limited support for Delphi. I know since I'm having dificulty finding suitable gaming libraries for Delphi since I'm using Delphi for game development.
Now why is that? Most of them are being developed by FPC/Lazarus and their developers just don't have enough money to go and buy Delphi for making them Delphi compatible.
Now you might say 200€ for starter edition isn't much but when you take into acount that Starter edition only supports 32 bit compiler you realize that you need athleast Proffesional edition if you wan't your library to have 64 bit support. But that means you will have to pay 1000€ for it. And nowadays you will definitly want 64 bit support if you want to do any serious game so you can make use of more than 3GB of ram.
This results in the fact that more and more pepole are literally abandoning Delphi support for their libraries.
And without Delphi compatible gaming libraries noone will use it for game development.


Another thing that bothers me greatly is how Embarcadero is "evolving" the Delphy syntax by implementing features present in other programming languages.
Now don't get me wrong I'm not against the implementation of theese features, but I'm mostly against the way they are implemented. I don't like when some feature gets implemented almost the same way it is implemented in some other programming language.
The reason why Objective Pascal or Delphi was so popular in the past and the reason why it is still poplar as it is is becouse it always have certain feature implemented differently than other programming languages (better implementation in many times). That is what has been atracting pepole to use of Objective Pascal.
But what Embarcadero is doing now is simply copy-cating features from other programming languages and therefore turning Delphi into crosbreed of other programming languages. ANd this by my opinion is not caled "maturing" of programming language.
What they need to do is learn how theese features work in other programming language and then implement them in new uniqe way, preferably better way.
But I'm afraid that guys at Embarcadero might not be capable of doing this. Why? As some computer scientist once sad:
"Most of the greatest discoveries are not discovered by scientists, but instead they are being discovered by stuedents, hobbyist. Why? Expirienced scientist already learned some path of how things are being discovered and they usually keeps folowing that path. But students, becouse they still didn't learned any particular path, still expiriment in serching of good path and have therefore better cahnce of finding compleetly different path which can lead to great discoveries. Same goes for software development. When you program long enough you develop some patern you folow when developing your application. And the more you stick to your pattern and less you expiriment, the less cahnce you have to discover something new and uniqe."

So that is why most of my suggestions are biased to atracting as many new pepole to Pascal development as posible becouse I hardly belive that only by new blod we have chance for something new to be discovered.
Take a look at big software companies like Microsoft, Google, Aple, etc. All of them are reagulary organizing varous "taletn sercing" events like "Imagine cup" where pepole are given some task and they must find best solution for it. Most often the winners are students and not some expirienced programmers.

So if Embarcadero truely wants to evolve the Delphi as programming language and prevent Delphi comunity to simply fade away they need to start atracting new pascal programmers fast. And yes making Delphi game development friendly is good choice since gaming development is far more popular than buisnies application development and gaming industry is making much more money. So it is no wonder that about 70% of all new programers wants do develop a game first.

Also Embarcadero mist not forget about keeping existing developers in using Delphi. Why? If they lose existing developers from the comunity, there will be noone to teach the very basics to newbies, there will be noone to encourage newbies to reach their boundaries.
And here the Embarcadero isn't doing verry good. Lately I have seen many Delphi developers switching to other programming languages. And main reason doesn't seem the lack of features of the Delphi programming language itsef but the lack of atention of Embarcadero to listen to complaints and suggestions of theese pepole.
Now the rumor for this is that on top of Embarcadero sits Marketing guys and not programmers so they don't view thigs the same.
If that is the truce could someone please remind theese marketing guys of this cruicail marketing saying: Customer is always right!
And yes we developers are your customers, so why are you refusing to listen to us?


I'm sorry if my post has turned into rant.
What I want is best for Delphi wich in turn would be best for me.

[phibermon]

Complaints about having to use Windows - Let's be clear about this one, having to compile on one platform whilst targetting another is an everyday occurrence.

It certainly is and I'd like to do that from Linux.

Edit : that sounded a bit confrontational wanting to avoid any comparisons between platforms best used for a development workstation, there's a myriad of reasons why you'd want the IDE to work across multiple platforms. Maybe you've got a mac and you don't want to buy a PC, maybe you work in a certain government department (certainly in our country Athena) and you can't even run Windows. Maybe you simply don't have the available hardware making dual-booting just to test your builds a hassle.

The top reason however is debugging (I retract this if Delphi supports remote debugging on every platform). Whilst it might be acceptable to say if it works on A then it works on B, that's under the provision that you're only using officially supported abstractions.

Even if remote debugging *is* supported, unless you can afford multiple machines in which to run concurrently with your windows development environment (Virtual machines not being acceptable as you require direct access to the GPU or at least some people do) you're again stuck.

Mobile phones, game consoles, arcade cabinet hardware - Remote debugging or 'virtually assured' operation is the only viable alternative because those devices are not suitable for use as workstations. Em. expect Mac owners to buy PCs to dev? virtual machines? Windows licences?

Yes their primary market are development studios that have tons of hardware but this is a thread regarding what game devs like us need from Delphi.

I happen to multi-boot various windows versions, Linux and OSX Lion on hardware I selected for that purpous, I purchased my copy of OSX etc and happily run FPC+Lazarus on all envrionments. It's a perfectly valid setup that many people choose.

I could not do that with Delphi in it's current state. It'd take me half a life-time with all the re-boots just to debug my Cocoa framework interface and Linux isn't even supported.

That really doesn't seem like a good thing to me, I'm not buying it am I? I don't see how that makes good business sense, I mean who's choosing not to buy something because it *does* support other platforms?

And if it's just a case of time + money and they can't afford to do it? well I'm not jumping on a sinking ship. And if it's for technical reasons that can't be overcome? well then Lazarus+FPC is superior (for my needs) and there's no reason to switch.

Now I don't really believe that those suggestions are the case, there's many factors at play I simply don't know anything about.

But the question what game devs need from Delphi?

The answer is simply : What we already get from FPC+Lazarus plus a ton more that will justify the cost.

[AthenaOfDelphi]

It certainly is and I'd like to do that from Linux.

Edit : that sounded a bit confrontational

Yes, it does sound a bit confrontational (I'm assuming you were commenting on my comment, on re-reading it, I'd agree)... apologies for that.

On the whole, I'd like to see a Linux version and a Mac version of Delphi (possibly in that order). I currently take two laptops with me everywhere I go... my MacBook Pro and my XP laptop. I have all my creative software (Cubase, Scrivener, Tinderbox for example on my Mac) and my business software and development tools on my XP laptop. I know I can run Parallels or Bootcamp etc. but for various reasons I choose not. Having Delphi on the Mac would be one step closer to me being able to ditch the XP laptop which would be great, so in essence I agree, but personally I think there are more pressing issues that Embarcadero have to deal with, hence my 'this is how it is, get used to it' stance.

I'd just like to add this thought to the melting pot as well while I'm posting... not all games rely on graphics APIs etc. My web based game relied heavily on databases and the VCL for it's tool chain. So, dropping the VCL and stuff like that that 'game's development hardly use' isn't necessarily the right thing to do. My next game is also likely to rely heavily on such things, again for it's backend server system and the administrative and design tools used to build and manage the game. Sure, the client will talk to a server and will be OpenGL based so it will have no need for databases and the VCL etc. but the backend will definitely need those things. Splitting them out would cause me a problem if I were faced with the decision of picking elements from a list of bits and pieces.

[phibermon]

Oh sorry I mean I thought that my comment sounded confrontational

I agree with your point that mass multiplayer, web based and certain other server based scenarios would make greater use of the VCL.

Perhaps that's something that Em. could concider - a lot of Delphi strengths lend well to game server scenarios being that they can be similar in requirements to database/web applications.

They are well positioned to push that angle, a few comprehensive working samples, making good use of the VCL, geared towards such things wouldn't break the bank. It's right up their street and there's definitely a market.

[azrael11]

I must say ... i totaly agree with sascha thoughts...

If emb want to catch Lazarus/FPC it must create a Delphi7 cross platform IDE with the latest OpenGL,Sound,NetPlay,Input Headers with no BullS@#$t licenses and have a minimux community forum with IDIE developers...

Then i think a lot of people come back and soon money from this stage fill the emb pockets...

Thank you...

[laggyluk]

oh wait, they could also release free, stripped 'basic' edition and then 'pro' version with rebranded GLScene components for 2k euros

[Darkhog]

Free is not a viable option

Except it is. Borland did it for years. Then they've made shitty Turbo Delphi, crippling it from ability to add custom components, like it was up until Delphi 7 Personal (dunno if Delphi 8 had Personal edition), then they sunk and had to sell Delphi.

They should get back to releasing non-crippled (except in terms of out-of-the-box components and maybe database support) Personal editions like Borland did with Delphi 7. After all, 7 is the lucky number.

[pstudio]

People have pretty much covered it but here is what I believe is essential in a Delphi for Game Developers Edition:

Crossplatform compiler (x86, x64, ARM...) (Windows, Mac, Linux, Android, IOS, Windows Phone...) - Must be easy to target any platform without any (major) changes in code.
Solid list of libraries - You know one of the big reason I like Oxygene for Java? I get the entire JDK library and can find pretty much anything on the internet. I don't expect Delphi to compete with Oracle on the size of the default library, but any data container, file reader/writer, stream system etc. must be included in a preferable memory efficient and high performance version.
Game/Media specific libraries - I don't expect Delphi to be included with all the game frameworks/engines out there, but as everyone else is saying: "At least get up to date headers for OGL, DirectX, OAL...".
IDE - a really good editor must be included. Visual editor would be nice for tools programming, but the code editor must be great. Anything that can be automated should be (with possibillity to turn off) to improve code production. A great debugger with memory inspection etc. is also needed.
Profiling tool - If you want to be serious about your games this is a requirement. You need to be able to inspect your memory usage and find bottlenecks, so you know where to optimize your code and where you can add a bit of extra wow to your game.
Cheap - and no stupid licensing requring you to pay additional fees if you actually make money.

I'll admit it's been a few years since I last actually followed and used Delphi. I went with FPC since it was just as good for my needs (if not better) and a lot cheaper. After receiving a free license for Oxygene for Java I've been using that and plan on continuing to use that paying for an upgrade when I need it.
However I have a hard time seeing Delphi regaining customers who have switched to FPC/Lazarus. Delphi's force is database applications and that kind of things. They do it well and their prices may be fair for that kind of usage. However it's mostly useless for indie game devs. Yes there will be times when databases are needed for games, but for most people this isn't a must.
If you strip away the things we don't need from Delphi it doesn't really compete well with FPC. Yes Delphi will be crossplatform eventually, but they will need to do it so much better than FPC or Oxygene and offer a lot more, if they want to sell me a copy of Delphi versus downloading FPC for free.

IMO Delphi is just not for game developers. Sure if you need Delphi for anything else then you can use it for Game development as well. If you just want to make games in Pascal; FPC is fine, and if you're not set on using Pascal there are plenty of other free solutions out there. I'll always have a soft spot for Delphi. It was the first programming tool/language I ever used but I feel like we've grown apart and a reunion is not likely.

[WILL]

As much as people would like a free Delphi, it's not going to happen. They just won't do it anymore.

I won't even bother getting involved in any "make a free version" conversations, because I've talked to them directly and they said an emphatical no every time. I didn't even bring it up myself, one time! They tried it when they were in-between (when they were called CodeGear) and to my best guess Embarcadero squashed it shortly after.

I think that they could make a new Edition of the Delphi compiler that was more comfortable and behaved the same way that those who are used to THAT tool have grown accustomed to it and make their low-end market sales (involving the game developers) that way. Make one for each major platform IDE (code editor and core essentials only) and make them single platform sell them at a low affordable code of about $50 - $100. That way they could allow you to still develop cross-platform by buying one for each platform you want to develop on. A solution that they would probably agree to if they really are interested in selling something to game developers that we may buy.

Of course this is not taking into account where ever the compiler quality and stability state is right now.