robwired

I remember reading this and thinking we would all come back to this discussion. When the other user posted I thought it might be a good time to revisit. If we aren't in the "longer term" yet, that's fine. I think I do understand from our discussion today that it's not just configuring the emulator cores but rather writing all the other stuff that's in an emulator that's the issue. Launchbox looks so beautiful and well done so it's easy to forget that the emulators are handling the input mapping, shaders, audio back end, etc.. and NOT Launchbox. I do understand that this is quite an undertaking LordMonkus and appreciate you explaining it. I can see the amount of work involved in rewriting a simpler RetroArch. I guess the only way to realize the vision that I have of drag and drop cores would be if RetroArch were to release two programs. The existing program and a separate version of it's program that would just be a core without an installer or its own interface. This core could then be imported into a Front-End such as LaunchBox. Then all of the code you are talking about would be there for front-end developers to work with. Otherwise, we are back to Launchbox allowing the drag and dropping of RetroArch and managing the installation process, which is what it does now for all emulators. Maybe that process could just be made easier and seem more integrated into Launchbox. Thanks much.

I actually just posted the question on their threads so we are on the same page. I'd love to hear what Jason thinks though, let's give him an opportunity to weigh in. If it's correct that it wouldn't be that hard to get the roms working maybe things like the input mapping, shaders, etc.. could just be things that could be done in time.

Don't give up on us yet my friend, what you are saying is making sense. I'm just not as technical as you are... but sometimes that's not a bad thing. When I ask my own developers to do things for my company it's the fact that I'm not a developer that sometimes gets them to think outside of the box. At the end of the day, if you all decide that it's too much work then I really appreciate at least getting an understanding of why. Let's put aside the licensing issues because I'm not suggesting that the cores be distributed with Launchbox, just that they could be imported the same way that bios and rom files are and we also don't know if Libetro would or wouldn't want their cores to be used with Launchbox, let's just say that's a separate discussion for now. It's actually an interesting concept they have though, isn't it. The idea that you would just drag and drop emulator cores, roms and bios files and any compliant front end could run them? While it might be a little more competition for Launchbox the upside to the community would be incredible and this would create a standard. I could imagine the entire emulator community just writing to be retroarch compliant and that being used for many front-ends. Now onto the time consuming things like input mapping, audio back-end, shaders, etc... could we get an idea of the man-hours that we are talking about? If let's say 50000 hours were going to be put into Launchbox this year, are we talking about a few hundred hours or a few thousand? I'm also curious if the Libetro coders might donate their code for those things or if that project could be open-sourced as well?

I agree that there would be a little time involved as there is with anything that Launchbox decides to do but allowing the use of Libetro cores would probably be worth it on so many levels. Ultimately this is probably the roadmap for emulation in general, to me it's all about having a standard. I think if you asked newbies, they would echo this logic. So what we come down to is simply whether or not it would make sense to take advantage of these open-source cores or not. Jason, could you please speak to how much time it would take to make Launchbox work with the Libetro cores?

Let's break it down: * OpenEmu and RetroArch are fully contained emulators which have the Libetro Cores already installed, right? Are the Libetro cores available for download by themselves? If so, could Launchbox simply allow those cores to be dragged and dropped? If so, then they wouldn't have to be distributed with Launchbox so there aren't any legal issues, but Launchbox could make it dead simple for them to be imported. If they are available separately then that's the end of that issue for those cores. As for the other emulator developers, if they want to write to Libetro compliant, that's up to them, no one has to chase anyone and I'm not suggesting that Launchbox pull its support of other emulators or get into fights with anyone, just to make those cores import easily. This is something that Launchbox is already working towards, right? I saw a video of Jason showing that he was going down this path already so what's the big deal with making this even simpler? To me, Launchbox is the perfect front-end and Libetro has the cores solution so they seem like a match made in heaven. Again, this isn't to say that other emulators shouldn't be supported or that Jason should be fighting any battles. Nor does this affect roms or bios files, these are all separate issues.

If the emulator cores were distributed with the understanding that they would be used as plug-ins then it would be the responsibility of the emulator developer to write them so that they would work with the various front-ends. This would also deal with any legal issues as the purpose of the emulator core is to work with the front-end and would be appropriately stated in the terms. If this plug-in eco-system existed then not only would Launchbox benefit but the entire community would including other front-ends, users, developers, etc...

Sounds like what we need is an Open Source Plug-In format for Emulator Cores that could then be plugged into any Front-End. Would something like this work?

I have to admit that I didn't understand the ramifications of what I was asking for previously. Not only would it be a completely different project to "own" an emulator but there are a lot of other considerations behind doing so, including legal issues. It's with this understanding that I actually no longer think that in the short term that the focus of Launchbox should be to become an emulator. My New Hope (Sorry, Star Wars Geek here) is that even if the Libetro cores don't get integrated into the code, essentially creating an emulator, that at a minimum the focus on Launchbox making it easy to setup emulators is taken to a point where no one has to think. Launchbox already does this to some extent but it's not seamless. I'd love to see it as simple as dragging and dropping the Libetro app into Launchbox and it just takes what it needs from the zip file and we are done. What I'd like to see is that no one has to actually install Libetro, just allow Launchbox to use the files contained within. Other options to dragging and dropping would be to have an import emulator screen that does the same thing, but again, without the actual installation of Libetro, just importing the cores. This would expand the userbase significantly since for everyone of us emulator enthusiasts there are a hundred newbies. Maybe another project could be started that would act as a Front-End Installer! That program would allow you to download the emulators, it would collect the files it needed, sort them into the appropriate folders and then let you choose a front-end. We certainly have gotten granular here, haven't we?

You guys are hilarious. How intense would you all be if I suggested that Launchbox and Libetro merged? Just kidding!

Just want to say that I wasn't offended at all which is why I made the joke to lighten up the mood. Your moderators are very passionate and protective of Launchbox and you. You picked a great staff, I'm honestly a little jealous. Getting to the truth of the matter at the end of the day is what it's all about!

I really like you guys.... you are VERY passionate! I'm not offended. If any of you are in the SF Bay Area let me know, drinks are on me! At the end of the day, it's clear that Launchbox is intended to: 1) Make Emulation more fun 2) Make Emulation elegant and easy I see that there are nods in the direction of making Launchbox easy and if the Libetro team has done most of the heavy lifting and will continue to do so, then it seems that this integration wouldn't be that hard. Would someone that is more tech savvy than me check to see if this is the case with Libetro cores?

Do I feel like I'm in the corner of your mom's basement holding up my shiny white iPhone as a cross to ward off a bunch of Lindows guys who are mad that I'm suggesting that we let a few of the pretty people join our birthday party? A little...

I actually started thinking about this when I saw the work that Jason was putting into in having Launchbox go through the trouble of recommending, downloading and attaching emulators to Launchbox. It's clear that there is an understanding there that the process should be simple but the amount of work in supporting all those steps and having to make videos explaining how to do it is not as elegant as simply including the cores and I'm not sure in the long run that it even saves any time. Of course no one is suggesting that roms or bios files be included but having a place on the cores screen where a user could drag and drop them is nice. I wouldn't say that this is any harder or gets my hands dirtier than dragging and dropping a game though. The real work is in configuring each emulator, it's frankly a pain in the butt and while many here are very tech savvy, this limitation on enjoying Launchbox shouldn't exist. Think of the children I assume that ease of use and elegance aren't features that some users would be willing to pay for, but there are users who would be happy to contribute financially, users that aren't even using Launchbox yet! Even Linus Torvalds uses his Mac hardware for Linux

I appreciate the feedback guys. You both make some great points. 1) Let's all agree that no one wants to lose the choice of using other emulators. 2) Emulators cores are VERY small so the additional space isn't really an issue, but I would agree that it should be easy to simply remove a core. 3) IF emulator cores are based on Libretro, which I don't know myself, then there wouldn't be much "bloat" in the code. 4) Retroarch is an absolute pain in the butt, its probably the hardest emulator to actually set up. Using it with Launchbox is like externally plugging a SCSI box into an iPad. It works but... Hopefully that addresses any downsides while allowing an elegant solution for many more users. Some people like to work on cars and I get that, but Launchbox seems like it's going for the turn-key approach rather than the build a car kit model.

Launchbox is an amazing piece of software, a phenomenal front-end and I think it has the potential to be the best game emulator there is. Yes, I know that it's not an emulator and there are those who would say that Emulator front-ends should stick to what they know like Hyperspin, RocketLauncher and MAME all being separate. Let's dispel the notion that users actually want to have three separate programs or that it's necessary to keep front-ends, launchers and emulators separate. On the Mac side, OpenEmu has done an incredible job combining these three together, all the while letting the emulator developers continue to work their magic and have found a model that gives credit to these developers. It's a win-win-win for users, the open-emu team and the emulator developers because there are probably a hundred times as many users than there would be otherwise and in some cases, it might be a thousand or ten thousand as many users, for example for a small system, how many users would actually take the time to search out a Fairchild system emulator. So the magic is there and it's not because development of cores contained WITHIN OpenEmu was some huge technological challenge, it was because the approach for the OpenEmu team has been to determine how to get as many users using the system as possible. OpenEmu is dead simple in this regard as a user does NOT have to separately download an emulator core, it's contained right within the program as are new cores with each version update. Yes, there would be a TINY amount of ONE-TIME work required to ask each of the Emulator developers to give permission for distribution, but fortunately, OpenEmu has already paved this road, so any developer that gave permission to OpenEmu would most likely not have an issue in giving it to LaunchBox. If LaunchBox were to take this approach it would immediately become the 800 pound gorilla in the room of Front-Ends, pushing aside HyperSpin and setting the foundation for collaboration between the Mac, Windows and Linux communities in a way that has never existed before. With a potential user base of 10x, 100x or maybe even greater the Launchbox community would attract new volunteers. It's all about the user experience. Emulator set up is just far too hard for many users, consuming hours and hours to set up each emulator. It's past time someone stood up and volunteered to lead. As a community we need our William Wallace and I would gladly contribute my $25/year in support of this well deserved cause. Warm Regards, Rob E.