State of the FE devs?

[screaming]

So why don't all of us who don't like the current FE choices get together and build one that works? Would any of you be interested in either starting fresh with a language like C# or picking up where PacManFan left off with Kymaera (C/C++)?

I'll go either way - they will probably both take the same amount of time to get up to speed.

There are a few threads around here or MAMEWAH's board where I describe what I think is important in an FE.  Hopefully we can use these to kickstart some sort of "mission statement".

An excerpt of a PM I sent some time ago:

1) Easy to configure. You should be able to hand-edit the configuration files and there should be an integrated interface to guide you along. In addition, just about everything should be as automatic as possible.
2) As flexible as possible but not at the expense of #1.  By "flexible" I mean:
    - Built-in supports for a range of emulators
   - Can add support for more emulators easily
   - "skinnable"
   - multiple emulators per "game list"
   - the possibility, via configration, of a "tree" view of the emulators/games.
3) It should be fast.
4) It should have integrated support for the most common hardware: ArcadeVGA, GP49s, KeyWiz, IPAC, etc.

In addition:

5) It should support older hardware wherever possible, but not at the expense of *useable* features.

I'd love to get some discussion going, if anything, to give the existing FE devs an idea of what we want.

[youki]

What your are diescribing is where all current active fe are going.   no?   Just let us time to implement features and enhance it  . We can not do perfect FE in one shot.  It is why most of FE on the Scene are still in beta.


Even if you start your dream FE now. You won't have it as you want before lot of time. I think.

[Havok]

1) Easy to configure. You should be able to hand-edit the configuration files and there should be an integrated interface to guide you along. In addition, just about everything should be as automatic as possible.
2) As flexible as possible but not at the expense of #1.  By "flexible" I mean:
    - Built-in supports for a range of emulators
   - Can add support for more emulators easily
   - "skinnable"
   - multiple emulators per "game list"
   - the possibility, via configration, of a "tree" view of the emulators/games.
3) It should be fast.
4) It should have integrated support for the most common hardware: ArcadeVGA, GP49s, KeyWiz, IPAC, etc.

I think that pretty much describes the state of the most popular FEs available today... Why re-invent the wheel?

[screaming]

I think that pretty much describes the state of the most popular FEs available today... Why re-invent the wheel?

  I have two intersecting opinions on this.

  On one hand, I feel like I'm forced to reinvent the wheel because no one else is willing to step up to the plate and manage thier front end like a real distributed project.  Like I said, I would be very happy to help code even the mundane stufff (like a config program), but I don't have time to waste.  If I do it again I want to know that the author and I share the same goals and/or I have some influence in the direction of the program (or at least the config files).

  On the other, I don't think we would be reinventing the wheel.  It can be argued that not a single front end out there has all those design goals, and given the longevity of the hobby I think it's a shame.  I'm constantly frustrated at the lack of interoperability between the front ends and the hassle to set them all up.

  Why can't they just WORK?  No display issues. No path issues. no artwork issues. No emulator issues. It can, and SHOULD, just work out of the box. 

  youki has the right idea and is getting there with his modules, but where's the automation to back that up?  Why can't his config program recognize that I have an SNES emulator in the same directory tree as every other emulator and automatically download, install and configure the module for me?

What your are diescribing is where all current active fe are going.   no?

  youki, you're a special case.  Out of everyone here you're the closest to being on the same wavelength as me with this. I mean no disrespect (to anyone, for that matter).

  That said, no, I don't think so.  Maybe in the back of thier head they sort of like the idea, but no, I don't think those are the standards to which they are coding.

Just let us time to implement features and enhance it  . We can not do perfect FE in one shot.  It is why most of FE on the Scene are still in beta.

  Most of the FEs in the scene are beta because they don't want to share the workload to speed up thier development.  They take many years to develop into something

The developers manage thier product (yes Howard, it's a product) like a hobby (i.e. they don't manage it much at all). This is all well and good, and I respect that, I'm just calling for something more polished.

And i know I'm not alone.

[Havok]

  Why can't they just WORK?  No display issues. No path issues. no artwork issues. No emulator issues. It can, and SHOULD, just work out of the box. 

I believe Atomic is heading for that, but I'll let Youki elaborate on that...

[youki]

youki has the right idea and is getting there with his modules, but where's the automation to back that up?  Why can't his config program recognize that I have an SNES emulator in the same directory tree as every other emulator and automatically download, install and configure the module for me?

Angelscry is working on a next generation of Plugnplay module , we are close of what your describe here. I call that "magic module"...  You put the module,  and browse your drive to find emulators , roms , etc... and it create dynamically the config for Atomic and then it appear in Atomic directly usable.    Some prototype are already working. But it is a very had work to do. Angelscry did a terrible work on that.   But i can not say when we will release that officially.

youki, you're a special case.  Out of everyone here you're the closest to being on the same wavelength as me with this. I mean no disrespect (to anyone, for that matter).

The "plug'n play" was in my head from the beginning.   But it is really hard to find solution like that . Because each user as a different need and expectation. 

The developers manage thier product (yes Howard, it's a product) like a hobby (i.e. they don't manage it much at all). This is all well and good, and I respect that, I'm just calling for something more polished.

Did you try Maximus arcade , i think as you have to pay for it should be more polished.
Concerning Atomic, i try to polish a little more at each version  and hope to have a saying "commercial level" version in 2007 . (i mean with real Documentation,  real integration of configarution tool , a better website..etc..etc.. )   . But when i say "commercial" it does NOT mean you will have to pay !!!

[screaming]

Just out of interest, what 'direction' did Mamewah take not to your liking?  Really, little has changed since you made the config program...in fact no official releases yet since v1.61.

  That's a large part of it, actually. I was put off by the lack of progress. MAMEWAH is a great great program, and the perfect choice for a front end for a very large part of this community, but the lack of progress is causing it lose momentum and the gap has started to close.

  Unless development speeds up, MAMEWAH is going to end up being simply be the best option for outdated hardware instead of the best option overall.  It's that decision that turned me off.

[screaming]

... But i can not say when we will release that officially.

  The mantra of every FE developer out there.  You guys must have that plaquered on the doorway to your secret clubhouse.

The "plug'n play" was in my head from the beginning.   But it is really hard to find solution like that . Because each user as a different need and expectation. 

  I know it's hard.  It's very hard, actually.  It's harder than most people realize.... but it's not impossible, and once you do it you will set your FE apart from every other FE out there because you'll be the only one who will give the users what they want.

Did you try Maximus arcade , i think as you have to pay for it should be more polished.

  No, I've never heard of that one before. I  just took a look at the website and looks decent, but still doesn't look like it has any easy way to set up the emulators.  Either way, there's no way I'm paying for it.

[MustardTent]

No, I've never heard of that one before. I  just took a look at the website and looks decent, but still doesn't look like it has any easy way to set up the emulators.  Either way, there's no way I'm paying for it.

Just to be clear:  There's no way you're paying for Maximus Arcade, or for any FE ever?

[screaming]

No, I've never heard of that one before. I  just took a look at the website and looks decent, but still doesn't look like it has any easy way to set up the emulators.  Either way, there's no way I'm paying for it.

Just to be clear:  There's no way you're paying for Maximus Arcade, or for any FE ever?

  I am a proficient enough programmer that I don't feel it is necessary to pay for a front end just for myself. Personally, I would not pay for any Windows front end currently in existance (ok, that I know about), but that does not reflect on Maximus Arcade's value to the community as a whole.

  Are you asking for my opinion on whether or not it's worth $25/license to the community?

[jelwell]

This is all in the wiki.
Joseph Elwell.

[headkaze]

That isn't the same thing. 

With the old vs6 stuff you can do an array of controls, and put controls in an array.  Now you can just put controls in an array. 

The first is very useful for click events. 

I can have an array of labels (label(0) to label(13)) and they share common events.  So the same function can be called if any of them are clicked without any special code or for/case statements. 

I haven't figured out how you can do that in .net, I think it's impossible.

Of course you can do that in .NET! Just have a label array like in VB6 and add the same function to the click event of each label:

Code: [Select]

Label[] myLabels = new Label[13];

System.EventHandler labelEventHandler = new System.EventHandler(Label_Click);

for(int i=0;i<13;i++) {
    myLabels[i].Name = "Label" + i.ToString();
    myLabels[i].Click += labelEventHandler;
}

private void Label_Click(object sender, System.EventArgs e)
{
    MessageBox.Show(((Label)sender).Name); // Display label name of clicked label
}
You can draw a control onto your form and in your form's code there is a box called "Windows Form Designer generated code" which is a region you can expand and see the code generated by the IDE.

If your referring to rewriting Jonny5 using C# I wouldn't use form controls, I would use GDI+ (aka the System.Drawing namespace). .NET is very flexible, and I can assure you that anything you can do in VB6 can be done in .NET. I uploaded one of my VB.NET assignments for you a while ago to check out how to use GDI+ in .NET. You can still download it from here. It was the basis for the GameEx Theme Editor I wrote, but I must admit I now write in C#, but you can convert the two quite easily and the resulted binaries are practically exactly the same.

Check out ConvertCSharp2VB and you can convert C# to VB.NET by pasting code into the website. I think there is a similar site somewhere that does the reverse.

[Howard_Casto]

Manually adding the same function for each label and having it automatically done for you are not the same things. 

There are a ton of things you can't do in .net that you could do in vs6 automatically, the most notable one being controls with transparent backgrounds.  Mind you, you can do it manually with gdi functions, but I don't understand how a platform that makes you do teh easiest of things the hard wasy is supposed to be better.  Especially considering you've got that 20 megs of bulk that has to be installed for each program. 

Mind you that doesn't always apply to fe development as hopefully we aren't using standard controls for everything, but it seems like for simple "windowsy" apps .net is 2 steps forward and about 10 steps back.  The main reason I'm not willing to upgrade yet is not because for my bg apps (dk j5) but my small ones.  It'll increase the bloat and make normally simple apps a lengthy build. 

[JoyMonkey]

Oh right!  I forgot about Kymea-- Kymar --  Kymaera..  At one point I tried getting everything together to compile it but I never managed to.  Then, wasn't he working on a big update?

Yeah, the update was going to add support for PNG-24's with their beautiful alpha transparencies. With that in there, you could make some really slick looking skins.

[headkaze]

Oh I think I get what you mean now, you mean when you use the IDE to draw a control like a label if you copy and paste it in VB6 it automatically places the control into an array for you. Yeah okay, I've never tried that in .NET but I'm writing an application as we speak that uses dynamic controls, and I just draw the control on the form, copy the generated code, then create a function like Label drawLabel(int x, int y, int width, int height) then I get it to return the Label object it creates and place that into a ListArray, so if I want to delete all the controls I can just loop through a foreach and object.Dispose().

But seriously, if your not taking on .NET because it doesn't create controls into arrays automatically, or that controls can't have transparent backgrounds then that's a pretty lame reason. For a start you shouldn't be using form controls for things like Johnny5. That should be done using GDI+. I think your coming up with anything to avoid the inevitable realisation that it's time to upgrade your coding skills. VB6 is dead, it's time to get over it and move on (IMHO)

And that 20 Meg of bulk isn't memory resident bulk, it's all the namespaces, dlls etc. and when you consider the amount of automation and code they've put into it, 20 Meg is a reasonable expectation. I also expect Vista to have .NET Framework installed by default. I remember the VB6 runtimes were never in Windows by default either. And really where in the day and age of dual core processors and high speed ADSL2, so really we can't say that 20 Megs is a big download, or that it's slower than a C++ program. I know youki keeps saying .NET is only for fast machines, but it's not true. If you look at GameEx as an example there are many people who run it on slow machines. I tried the demo of AtomicFE and to me the scripted themes were very slow, so it's funny how people can call .NET slow just by an assumption.

Well, I don't want to keep trying to convince you to move with the times, if your gonna keep coding in VB6 your gonna keep coding in VB6. Dosn't really bother me really. I know as a fellow coder your passionate about the language you use. But I used to love VB6 too, and it's not after you undertake a big project in .NET do you start falling for it like VB6. And once you go dot net you never go back... bah who cares! O_o

[youki]

I tried the demo of AtomicFE and to me the scripted themes were very slow, so it's funny how people can call .NET slow just by an assumption.

That Demo uses the old engine, if you try the demo using Atomic 0.17-2 engine it will be faster.
And you know the scripting use VBScript... it is why it can be slow for heavy animation.

You will see very soon  a demonstration the power of my engine in term of layout and speed. 

[headkaze]

I tried the demo of AtomicFE and to me the scripted themes were very slow, so it's funny how people can call .NET slow just by an assumption.

That Demo uses the old engine, if you try the demo using Atomic 0.17-2 engine it will be faster.
And you know the scripting use VBScript... it is why it can be slow for heavy animation.

You will see very soon  a demonstration the power of my engine in term of layout and speed. 

That's what happens when you don't keep up with the development of other FE's you don't know how better they are since you last looked at them! Thats why when people (won't mention any names) say things about GameEx it's quite often about the state of it 6 months ago!

[Howard_Casto]

But seriously, if your not taking on .NET because it doesn't create controls into arrays automatically, or that controls can't have transparent backgrounds then that's a pretty lame reason. For a start you shouldn't be using form controls for things like Johnny5. That should be done using GDI+. I think your coming up with anything to avoid the inevitable realisation that it's time to upgrade your coding skills. VB6 is dead, it's time to get over it and move on (IMHO)

No not really.  Considering anything I do graphically I do via directX now I'm more than competant when it comes to coding skill.  But I am NOT willing to re-write all of my existing apps just because m$ was too stupid to keep all existing controls/control properties from vs6.  J5, in particular uses transparent labels for the captions.  Now yes I can do that via gdi, but gdi doesn't nicely format the text, you have to do all the centering and alignment manually. I can port the whole thing to dx, but the newer dx text functions don't work on older video cards, so again, I'd have to do it manually. With a transparent label, it's all done for you.  I'm not willing to upgrade because it's dumb to spend hours re-writing functions to get the application to do exactly what it did before, only now it takes more resources (.net controls are huge and bloated as is the .net framework, it doesn't work like runtime files did). 

I'm more than willing to write NEW apps in .net (I think my mame output handler might be made in c#).  But upgrading for the sake of upgrading would only remvoe features, not add them. 

I'm not arguing over the new suites merits, but it has a ton of flaws.  My guess is you never really used some of the built-in stuff in vs6 so you don't miss it.  Trust me, I could make you a list of things they've left out a mile long.  I'm not opposed to change, but some of these functions don't have an alternative in .net. 

[headkaze]

No not really.  Considering anything I do graphically I do via directX now I'm more than competant when it comes to coding skill.

What I meant by that is update your coding skills in .NET. That's all I meant, I'm not assuming anything about your programming skills in general.

I think you make some valid points, there are some annoying things with .NET, but I also felt the same way when moving between VC6 and VB6. I find that doing a bit of research on the web usually reveals a solution to my problems.

And your right, I don't think you should re-write all your applications for the sake of it, you were just asking about control arrays and I remembered checking out J5 and assumed you were playing with the idea of rewriting it in .NET. It's not terribly difficult to do it all in GDI+ once you write a few classes to display things how you like.

I guess all I'm saying is be open minded and don't write off .NET because of a few annoying limitations. Quite often the limitations can be removed by inheriting the controls and overriding the built in methods. I know that sucks, but it's a work around. For example at the moment I'm writing something for GameEx, there were two annoying things that didn't work the way I wanted out of the box. 1. Resizing a label's height based on the multi-line text inside. I wrote a nasty unmanaged GDI class to do that, then I posted a message on a .NET forum and a nice guy posted managed code to do it. 2. I couldn't get tooltips with multi-line text. Again I found a great class in on the Net that would automatically place line breaks into the tooltip's text at a width you specify. So while the limitations can be frustrating the good points outweigh the bad by a long shot.

[AllisterFiend]

Lots of good information in this thread.

I want to thank everyone who helped me sort through everything.

I have learned a lot in the last few days.

I'm looking into VB 2005 or Visual C# 2005 and I'm leaning toward Visual C#.

Thanks once again.

Allister Fiend

[Minwah]

  Why can't they just WORK?  No display issues. No path issues. no artwork issues. No emulator issues. It can, and SHOULD, just work out of the box. 

Because everyone has completely different setups, making this very difficult to achieve, without having to assume some things or hard code things for certain emulators.  If everyone had just a Mame setup with the default paths, then I agree this is possible and would be easy.

The developers manage thier product (yes Howard, it's a product) like a hobby (i.e. they don't manage it much at all). This is all well and good, and I respect that, I'm just calling for something more polished.

And i know I'm not alone.

You're are not alone in that you want everything for nothing.  If I thought making a highly professional polished frontend was worth it then I would give up my job and all my hobbies to do it.  But frankly the return for doing it is practically non-existant.  I purely work on my FE because I want to use it myself, and obviously I add other people's ideas to enhance other peoples enjoyment of it - and it is nice for me to know that other people want to use my FE.

With that said, I do not mean that I do not *want* my FE to be very polished etc., and hopefully it will eventually be, but as for development speed, well that is limited by time I have and whatever else I am doing.  Most crucially, IMHO, time is of absolutely no relevance in the emulation scene - emulators and fe's will come, develop and go, and the user will use whatever they want.  There is always more than enough choice to keep everybody happy.

[Minwah]

VB6 is dead, it's time to get over it and move on (IMHO)

If I were a commercial programming company then I would agree.  But for our purposes VB6 is more than capable. 

Does the programming language used actually matters as long as the FE works??

[screaming]

  Right, and that's kind of my point:   It could be polished if you weren't the only one programming it. You know, say you devote 5 hours a week to it (an hour after work each day, for example). If you had two people doing that, you'd have 10 man hours. 3 people, 15 man hours. 4 people and you have 20 man hours - or a good solid weekend of programming.

  All you put in was your hour after each workday.

  It's not just you, man, it's every developer out there with an existing maintained front end. With very few exceptions, they all have the same goals and are going the same direction! Why not collaborate instead of doing reinventing the wheel every time you update your FE?

[screaming]

  Why can't they just WORK?  No display issues. No path issues. no artwork issues. No emulator issues. It can, and SHOULD, just work out of the box. 

Because everyone has completely different setups, making this very difficult to achieve, without having to assume some things or hard code things for certain emulators.  If everyone had just a Mame setup with the default paths, then I agree this is possible and would be easy.

  Nah, it's not hard. Look at youki's modules! He has maintainers who have volunteered to maintain certain modules that apply to him.  Heck, you have jcrouse who posted his brazilian config files for his brazillian emulators.  What's there for you to maintain?

  If you programmed your FE in such a way that it provided a framework for a generic emulator, and implemented some kind of plugin, or "module" infrastructure, you could let someone else do the work of keeping up with the emulators leaving you to keep going with the backend and still provide your users a customized working and easy solution.

  I've said all this already about a year ago. It's nothing new.....

[swindus]

But the MAME module of AtomicFE is nothing else like a default MAME installation with a default path on the same drive as the FE ...

I have to agree with Minwah here. It's not easy to create a framework that will work for all users out there. It works if the user will use exactly the given FE structure, but I know a lot of users which are not willing or can not use this structure for example when using more than one harddrive.

Correct me if I'm wrong.

[MYX]

I am about done with the build of the cab. I always hate when I see so much talk about writing this stuff and it flys over my head. So I decided a while back that once the cab was done to start to learn to write some flavor of code. On a vacation about 9 years ago I bought a book about HTML. I read it and when I got home I wrote my first web page in note pad. Was it pretty, um... no. But I picked it up fairly quickly. And now occasionally I have to dink with some Java script when I am working in After Effects to write animation expressions. I understand that things like VB6 and C++ are far deeper than this, but I would assume that the process is similar... Write... Doesn't work...de bug until you find that you typed a , instead of a . then bang head on wall.

I asked Randy T about it and he recommended VB6. When I went to find info about that I learned that it is not only a language, but you have to buy a really frakin expensive program to do it in. I then found Visual Studio Express It is free and said to be geared towards the new programmer or hobby programmer. I really want to understand what the heck Howard is actually saying. I would like to be able to give at least a somewhat educated question instead of trying to ask questions that probably are not even close to how things work. I want to be able to add my own ideas to this FE business. I was wondering if you have any knowledge about the express version of VB2005?

The following, even though it may sound like total criticism, is actually said with the highest of respect to you guys. You have done what we can't. You have stepped up to the plate and given us FEs for MAME.
I think in a lot of ways, the reason why people are getting frustrated with FE a over FE b is because one writer may develop towards things that would personally benefit him because he uses a specific brand of hardware vs .another programmer who writes towards another type of hardware. Recently with the LEDWiz stuff coming to light (sorry, no pun intended) Some people are not at all interested because the original cabs may not have used this type of hardware, while others see it as a new gaming experience. I tend to fall into the later category. I lived through the 80's. I lived in arcades. My friends would ride our bikes for an hour to go to the Castle Golf and Games to go play Gauntlet and waste all our birthday money on the damn thing. But I am letting MAME be the creators of original game play. To me the FE and it's abilities never existed in the arcade. Nor did the flashy animated backgrounds. Heck if you want authenticity, you build a FE and when you flip a switch the screen comes to life and you wait for it to boot up, then, you got one game.
Naw, that was then and this is now. I say lets ("lets" is assuming I pick this stuff as well as I hope) freak the thing out and make it jump through all the hoops.

<rant>
Up until I started building my cab, I have always used MAME32. I had built a desktop control to play with my computer. Once I started with the cab, I knew I wanted to use a FE so I could skin it and make a part of the main design of the cabinet so everything has a cohesive look. As to the question about buying a FE that did everything that one would wish it to do... I saw many of you say you would never buy such a thing. But please remember, YOU ARE THE PROGRAMMERS WHO WRITE AND UNDERSTAND THIS STUFF. I had no knowledge of any of this stuff and it has taken a couple of months of weeping and gnashing of teeth to get to the point I am (which is still not very far) I'll tell you, when I have to dance through 4 different config files to finally find the place to change a 1 to a 0 and then I have to make a  batch file put it in here then tell some line in one of those 4 or 5 config files to look under the kitchen sink to find the batch file, then make this load that and so on and so forth is just silly. You are making the user become a dang programmer. Heck yes, I would pay $25 for a FE that said, ok where are the roms, thank you, where are the snaps, thank you, Have a page of supported hardware. Click here and the thing just works works. Instead of Here is my FE and if you want it to do XYZ then go and write a script to make it do ABC and then use this widget to convert to XYZ, then go to config a and switch the 1 back to a 0. I want to click on the stuff I want, and ignore the stuff I do not want. The FE then puts batch file XYZ in place for me. If it is something that has to depend on info from me, make a batch file wizard. Write, not in programese but in simple English what you want. In one FE I was something like "extract list from yaddayaddaXML" or DAT file. I have no idea what this means. I want another button that says "I have no clue, you choose."  There is a lot of language in these things that is written to someone who knows what you know. Write instructions that the everyday man can understand. Either that or make far better support docs (perhaps eve with a picture or 2). I see a lot of pride from each of you in your programs. And there is good reason to be. There is a lot of very cool stuff being done. So if you are this good then YOU be the programmer and let the user be the user. If you want your FE to be the best (which secretly you do want) and you want people to say Dang, "This IS" the best dang piece uh of dang software I ever saw man!!!, then write the best and make it easy to use. There are probably more complaints about difficulty of use than about anything else in this hobby. You are busy humming along building this cab, then bam! You hit a wall when it comes time to make it work with the FE.

Aight I am done </rant>

 

[screaming]

But the MAME module of AtomicFE is nothing else like a default MAME installation with a default path on the same drive as the FE ...

I have to agree with Minwah here. It's not easy to create a framework that will work for all users out there. It works if the user will use exactly the given FE structure, but I know a lot of users which are not willing or can not use this structure for example when using more than one harddrive.

Correct me if I'm wrong.

  Why does "not easy" mean "impossible"?  Are you all afraid of a little challenge? 

[swindus]

  Why does "not easy" mean "impossible"?  Are you all afraid of a little challenge? 

No, I don't think so.
The current FE's are exactly that what you want. A framework for a lot of emulators. You are right, they work all in a different way, some easy, some more complicated. But they are always a try of a developer to create the 'ultimate' way to work with almost every emulator in a easy way. Maybe for now there is no better way to do this.

Additional is not only up to the developer to create a nice software. It's also part of the user, especially when it is freeware. No bug reports, no wishes means, no development in the right direction for the user, thats it.

[screaming]

I have been giving Minwah suggestions for MAMEWAH for well over 2 years now because that's the FE I chose to use. Some he as implemented some he has not.  Some I've been able to directly contribute with (like a config program), but most I can't because it's not open sourced (or even "shared source"d).

  I am not alone by any means, and MAMEWAH isn't the only FE that I've given my input on.

[swindus]

No offence here, I just share my point of view. I know that you contribute a lot to the community so don't take any argument personally.

So why don't you try to create a FE dev group which will build the best FE ever?

[screaming]

So why don't you try to create a FE dev group which will build the best FE ever?

  Yeah, that would be the point of this thread

  Actually, I'm surprised at the lack of interest particularly from the non-FE coders out there.  *shrug*

[Havok]

Actually, I'm surprised at the lack of interest particularly from the non-FE coders out there.  *shrug*

Why not generate the framework, and then put out the call? I would be interested in seeing a fresh approach, with the ability to submit suggestions with some type of submission tracking system...

[screaming]

I started this a while ago and was working on it pretty dilligently until summer came around   Now that it's getting colder I'm looking to get back into it.

http://byofedev.bluecamel.org/

It doesn't do anything "plugin" or "module"-wise, but it wouldn't be too hard to add into the infrastructure.

[screaming]

Now that I'm starting it back up again it seemed like a good idea to find out who would else could come on board.   *shrug*

[youki]

But the MAME module of AtomicFE is nothing else like a default MAME installation with a default path on the same drive as the FE ...

Once again you affirm things without really knowing the subject.  But ok, if you look only the mame module i made  it looks like that but if you look closer you will see  it is not just a default installation.  And have a look to others modules....  you will see if it is default installation.
Angelscry did a phenomenal work on that.  He coded his own wrapper , own list generator , tried tens of emultors for a same machine to find the best one. Made all the configuration of emulators and modules. Some modules uses multiple emulators and run the game on the best one for that game.   And update regulary module on his web site.  And you say "It is nothing else like a default installation"....    
 

[swindus]

Sorry youki, don't want to derogate your or Angelscry's work. The plug and play stuff of AtomicFE is a great idea. But please put my statement back into the context of the discussion from above and you will see I'm right here.

[Howard_Casto]

If you want your FE to be the best (which secretly you do want) and you want people to say Dang, "This IS" the best dang piece uh of dang software I ever saw man!!!, then write the best and make it easy to use. There are probably more complaints about difficulty of use than about anything else in this hobby. You are busy humming along building this cab, then bam! You hit a wall when it comes time to make it work with the FE.

Aight I am done </rant>

I think this is a major misconception of users out there.  The cabinet part is supposed to be the easy part and the software should take you a lot more time to setup.  Let's put it this way.  When you buy a new pc which is harder, taking it out of the box and plugging in all the cables or spending the next 8 hours installing all the programs you need and getting the settings just the way you like it?  It's the same thing for mame cabs, just multiply the time by about a thousand.  Computers are complicated, computer software is compliacted, emulators are infinately more complicated than regular software and a program that is supposed to manage/launch an emulator... holy crap, that thing is complicated. 

None of use expect for everyone to bow at our feet at how great our stuff is, but what would be nice is if users would be appreciative of how easy we've made it for them instead of constantly complaining when they have trouble getting one thing or another configured the way they would like.  People get on my case a lot because I am negative, I'm realistic.  I would rather point out hurdles that devs are going to run into rather then tell them how great it is going to be.  I do this because the users that say "yeah that'd be cool you should do it" are the exact same ones to complain un-endingly when you are finished with it and it isn't exactly as they wanted it. 

You learn real quick that you should please yourself first and the user second or else you go crazy trying to make everyone happy.

I'm not talking about you in particular, you haven't really been guilty of this I'm just saying in general. 

I understand the bit about being willing to pay for a fe that just works.  The thing is that would take an incredible amount of time to do, so much so that I don't know if anyone of us could afford to spend the time building, even if we charged. 

People in this thread keep throwing words around like "just work" but if you truely want to give the user the freedom to set things up the way they want, this isn't easily possible.  Let me give you an example with the easiest one to deal with MAME:

Ok first off where is it?  You can ask the user, but it would require them to actually know where they put mame and probably the use of a mouse.  Well that probably  isn't going to work on the cab itself.  You could do it over the network, but what about the cabs without nics?  Parallel port connection?  Floppy?
Let's assume we figured that much out.  Now for mame itself.  Mame has around 60 options and they seem to change every week.  There is absolutely no way to set them up optimally for every pc in an automatic fashion.  You can write a fe that allows you to access these options internally but that in of itself could be a fulltime job.  Let's skip over that part and assume the user can figure out how to setup mame on their own.  Now for roms.  Some people have multiple rom directories, some just have one.  Some are one different drives, some are on network shares, some are on cd rom.  You can have the program search for roms across the whole computer but even with a super fast algorythm that could take hours on larger harddrives.  You can let users manually browse but if they have a lot of paths to add it could take forever.  Manually typing paths leaves it open to user error.  Now we have the artwork paths, the same issue applies.  Also the user has to actually find and download the artwork themselves.  Finally it's time to generate a gamelist.  You can do that automatically, but there are three distinct case scenarios you have to check for as depending upon the mame version, there are three different calls to choose from to generate the list.  Once you get that mess sorted out, you have to deal with catvers and how the fe is going to display the list.  Just a big list of all the games?  How about clones?  Lists by cats?  Different users are going to want it different ways, sometimes in very custom ways.  Those are a lof of questions to ask and a lot of places where things can go wrong.  Oh and before you get some grand idea like "what about a drop in module that has the emu, roms and artwork" I'm not even going to begin to explain how illegal that is and how that violates mame's useage agreement. 

I don't think anyone in their right mind would expect a developer to take care of all of this automatically, but the way some of you guys talk, we might think so. 

And remember, I listed the easy one, mame, which has built in tools to help us devs.  There are far more complicated ones like daphne and zinc, in which we can either write a slew of custom utilites, depend upon clrmamepro dats for data or other crazy means.  Then of course there's the custom launching. 

I'm not ranting back, or anything I just wanted to share the devs likely point of view. 

[Havok]

Sorry youki, don't want to derogate your or Angelscry's work. The plug and play stuff of AtomicFE is a great idea. But please put my statement back into the context of the discussion from above and you will see I'm right here.

I think you are missing the point of his plug and play modules - they are for someone who wants to quickly and easily setup their system with minimal configuration time, or for someone with little understanding on how to do so. So, in that context - it is what the user wants.

I use Atomic - and I don't use plug and play modules. I setup everything manually - I have a fairly complicated tree structure, with filtered lists and multiple emulators - I can (and do) use multiple drives.

So, to put this all in perspective: Youki's FE has the best of both worlds: quick and easy "it just works" configuration, AND power user "I want it this way" configuration...

[Havok]

You will see very soon  a demonstration the power of my engine in term of layout and speed. 

And it is very cool!!!

 

[screaming]

I'm not ranting back, or anything I just wanted to share the devs likely point of view. 

  I would like to state that while I agree with some of what you say, you do not speak for me.
 

None of use expect for everyone to bow at our feet at how great our stuff is, but what would be nice is if users would be appreciative of how easy we've made it for them instead of constantly complaining when they have trouble getting one thing or another configured the way they would like.

  The developer in me knows that it's very difficult to make customers happy. Not only that, but it's even harder to make customers of a software product happy because of all the variables we have to account for. When someone says something deceivingly simple like, "I want to draw a square on the screen right here.", any software program has to:

  - detect the mouse click
  - figure out where the mouse is
  - watch the mouse
  - detect when they let go of the mouse
  - figure out where the mouse is
  - what color is the square (need to allocate color(s))?
  - how thick is the border?
  - is it filled in or just outlined?
  - is it a perfect square or are they trying to build a rectangle?
  - What if they draw the square off the page?
  - what if they draw it backwards (i.e. bottom-right to top-left)? Everything is negative.
 
  That's a lot of stuff that I need to keep track of just for drawing a square, and that's not including all the initialization crap that you need to do.

  However, the user side of me doesn't care about this and just wants to draw a dang square. "Why can't it just be simple?", I ask.
 

I think this is a major misconception of users out there.  The cabinet part is supposed to be the easy part and the software should take you a lot more time to setup.

  If you're the only one who thinks this way then I don't think the misconception is "out there". I think it's with you.
 

I do this because the users that say "yeah that'd be cool you should do it" are the exact same ones to complain un-endingly when you are finished with it and it isn't exactly as they wanted it. 

  Personally I take that as motivation for expanding my programs, and transversly my knowledge, better. By "better", I mean more user friendly and automatic.
 
  So far no one has said that what i want to do is impossible, only difficult.
 

The thing is that would take an incredible amount of time to do, so much so that I don't know if anyone of us could afford to spend the time building, even if we charged. 

  You and Minwah keep proving my point about opening up your own FEs.

  As this hobby progresses, the user's demands are getting greater and greater (look at PowerMAME vs. regular MAME, or a keyboard hack vs a keyboard encoder! Yikes!).  As such, the demand for a more sophisticated FE grows and grows beyond what one can expect from a single hobbyist developer.  Open it up!  Look at the bigger picture and manage it like a project! YOUR dream of what you see DK being could be obtained so much faster....
 

People in this thread keep throwing words around like "just work" but if you truely want to give the user the freedom to set things up the way they want, this isn't easily possible.  Let me give you an example with the easiest one to deal with MAME:

  It CAN be done, and WILL be done, either by me or someone else (like youki has started to do) if the "old timer" FE devs don't step up to the plate.