Mux

Neither of those tiers warrant an additional $2m. :-)

Somebody was asking earlier about Gamemaker and Unity and what not with regards to plug-ins. While GM might be able to roll something out, Unity for sure isn't. For one, they're going to ask 7 figures in order to even consider supporting your platform and for two, it uses c# which means you need some significant processing power to handle all that runtime managed code (garbage collection, etc). Nintendo licensed Untiy for an 8-figure sum whereas Sony and others have multi year deals for 7-figure deals. And that's for platforms that have ample resources and storage. So yeah, it would take more than the $3.3M in order to not be laughed out of Unity's offices. Anyway, software development is expensive. Considering they have ONE guy, at say a burn-rate of $10k/month (low estimate) they would need to spend a boatload of money on IP in order to not burn the guy out. Things like USB, IP stacks, that kind of thing. Alternatively, they could hire another programmer but then they would have to finish everything in about 4 months, so no... Also, for new cores... Who's going to do the API's, do the testing and support? While it's probably never going to happen, I would really like to understand how they came up with these numbers in some sort of post-mortum. I've been in game development my entire life and it's just mind-blowing to think they can pull this off. Mike seems hellbent to destroy whatever is left of his credibility, perhaps out of stress / being pissed off at the world but admitting that they completely under-estimated everything would go a long way, at least IMO, towards restoring that. -Mux

I was talking about the new 'prototype' for RVGS. It shows fuck all.. Just another render / picture of a case! FWIW with regards to their FB message: 27 years in game development with 10+ years in hardware development.. [Edit: Nevermind... That's the very first shell they had on the first page] So yeah, I've written 'a couple of lines of code in my life' -Mux

but but but... PCB? -Mux

One thing about cartridges though is that FLASH roms are dirt cheap these days. You can get a 16MB (byte, not bit) rom for about $1.50. Add in the cost of the PCB and a shell in semi decent quantity and you're look at most at something like $10 for COG, if that. So it's not that high a cost per cartridge. On the other hand, you'd have to either create a service for people to create carts or make it part of a dev system which adds to cost to the system as a whole. At a minimum, you'll need to add either a flash card or add stuff to the main board, neither of which are cheap. SD cards on the other hand make it a lot easier for people to just write games and share them, including the unwanted effects of piracy, not that that should be a big issue though. You could do BOTH but that takes a lot more work. Cartridges themselves are by far the easiest way to go as you don't have to deal with filesystems, load times, etc. So yeah, it's a bit of a trade-off. I'm really tempted to work on another 'retro console' with parts sourced from eBay though. I'm nowhere near Kevin's level but enjoy creating systems that hover around the 1980-1990 time era with equally old parts. That said, you're probably looking at $100 (retail) for the board itself whereas if you go with a single FPGA, some memory and an SD card (+ USB development interface), it's about half that.. -Mux

With regards to the question about joysticks and buttons on my 'stick-of-gum' board, they're actually on there! It's got a tiny 8-way joystick and 4 buttons. The joystick actually has a 5-th button which I used to insert coin(s). Going to back RVGS. I don't know guys.. I'm keeping tabs on it but they seem to be digging themselves into a hole. Maybe they've got more press lined up and will get a (large?) bump when Kotaku or Gamasutra or what not posts it, who knows. Like some other folks here, I also worked on games for the PS1 and 'reverse engineered' (i.e. disassembled the GTE codes to get better performance. Don't know if that's his claim to fame, but if it was... Well, that's not that great. Kevin, sounds like your cryonic stuff is pretty awesome. Pretty sure they've got plenty of quality control on both the hard- and software! Btw, on the XMOS version I actually managed to get color NTSC out *in software*. -Mux

It's a lot smaller than that actually but yeah.. I went a bit overboard and stuck the joystick and buttons on the PCB as well. From a construction point of view probably not the best idea given the abuse controllers get... -Mux

Never heard of him but yeah, I'm starting to think AtariAge is actually a really good place to meet people who've got access to machinery and tools that I don't :-) I've been saving up for a small CNC router so I can do my own stuff but for injection molding stuff and what not, I could really use some contacts! Thanks! -Mux

The JTAG interface is so that you don't have to mess around with burning new configuration proms all the time. That said, I'd lean more to having a 'fixed' configuration and give programmers the flexibility to do awesome stuff.That's why the 2600 had such a long life IMO. -Mux

Somewhat off topic but here's an arcade emulator I made the size of a stick of gum.. The idea was to stick (no pun intended) it in a tiny, custom arcade cabinet but alas.. I suck at that kind of stuff without a CNC router. I've got some videos somewhere if anyone's interested but basically it's emulating pacman and ms-pacman and has / had a main menu where you could select multiple games a la 60-in-1. Same problem as my other board is that each game would need a separate executable and thus be developed separately. Single 8-core XMOS chip and SMT parts. Unfortunately, XMOS did a bit of a 180 and hiked up their prices and (probably) is on their way out. That said, I did some fun things with it, like my first VDP core.. -Mux P.S. I guess my point is that prototypes show better than shiny 3d videos and cardboard cutouts with parts glued on them...

Figured I'd chime in since I got quoted and all that :-) So the main reason for using ancient / old technology was that it was through-hole, which means I can hand-solder it myself without screwing around with 0.5mm pins and magnifying glasses. I initially had the crazy idea that people might want to assemble the boards themselves, as well as write cores (that's why the JTAG connector is partially on there). Anywho, I realized a couple of things: 1. The vast majority of people just want a pre-assembled board 2. Very few, if any people are going to write a new core. 3. While sourcing parts from eBay works for now, it's not exactly something you can scale. 4. I created a daughter board that was basically a rom emulator, which increased costs and (again) isn't something most people would want. While still available, CPU's (z80 an 6502 are somewhere around $5) at some point they're going to disappear. Soft cores would solve that but FPGA's seem to be replaced with more powerful ones every day. So with that in mind, at a minimum, a board will need to: 1. Be pre-assembled (surface mount) 2. Have a USB interface for development and an SD-card for storing games 3. Have a decent software interface to program the games. 4. Keep the cost down as much as possible 5. Make it open source Contrary to Kevin's comment about HDMI and partially because I (honestly) haven't looked into it and want to keep things simple. Most monitors still take VGA as well, so yeah.. why not? To Emehrs point, keep it simple and at least at a NES level with some additional bells and whistles for people to tinker with. -Mux

Nintendo will *NEVER* release their IP, ever. They're incredibly protective and while they tolerate emulators and what not, anyone who's trying to make a buck of their older technology will get a cease and desist letter. -Mux

Kevin's point about chasing rainbows is a valid one. Friend of mine a long time ago did an atari 2600 core and kept adding stuff to it until eventually Jax Pacific released their 2600 clone. He was all pissed off about it but they just hunkered down and built it while he was going down the road of LVDS and God knows what else. At least Ouya didn't make that mistake. Btw, so wait... They're not adding an FPGA for $1.9m in dev cost but only for the 'higher tier'? One might expensive FPGA that is... -Mux

As I mentioned earlier, and Kevin backed agreed with, the software part is going to take a very, very long time. From what I understand, they're going to outsource it? That's a recipe for disaster. For the amount of money they're asking, they should be able to staff up and get their hands dirty as well. Just throwing it over the fence isn't going to work. Another point that's completely unaddressed is what 'the toolchain' entails. Drivers, debuggers, assemblers, linkers, etc. Are you going with GNU? LLVM? Who's doing that? What's the timeline? Granted these are all areas that don't need to have been disclosed up front but at least it'd be good to have some insight in. From where I'm sitting, they're hopelessly underestimating what it takes to bring a product to market.. -Mux

@parrothead... I just don't get it.. Why is it so difficult to show a WIP? All Kevin did was show you the board(s) he made, explained how much time he spent on it and basically tried to explain to people that your numbers were high. You then immediately jumped down his throat and told everyone that you won't be soliciting cores from him. Really? First of all, if this is even true, I would expected to have Kevin mention part of this. Secondly, if you're scrambling for cores, then what the hell are you guys working on? The bad voodoo stems from the fact that you guys just don't show anything other than injection molded cases and slick 3d videos without showing ANYTHING that resembles the tech behind the console, despite the fact that the hardware guys had PLENTY of time to get at least SOMETHING up. I mean, seriously... So now you're making it sound that Kevin was in cahoots with you and is now effectively 'cut off' because he, like most of us here, have some serious doubt about the validity and capability of your team. -Mux

What pisses me off is that it's people like Kevin and others (there are plenty of people) who actually PRODUCE something who're going to get screwed down the line by these kinds of vaporware products. If anyone ever tries to get something out on IGG or KS they're not going to succeed because the public is going to be wary as they haven't got a clue what the difference is between a working prototype and a shiny 3d render of a vaporware product. -Mux

It's actually kinda sad that projects like these go live and erode the good will and trust from backers. While they still have plenty of time to get a bump, I really, really hope that if they DO get funded, the actually follow up with something tangible. It just boggles my mind, and I keep repeating myself, that with TWO hardware people on board, they're not showing ANYTHING other than a piece of cardboard with some components glued on it. -Mux

Seriously? This is what they came up with? The PCB isn't even rendered correctly NOR does it have any traces on there. And an additional $2m in order to 'increase the size of the FPGA'? It'll be fun to see and watch how many suckers consumers are willing to part with their money. Unbelievable. Edit: Love the comparison matrix. It's like comparing a tricycle to a race car. -Mux

Alright, one more thing... The whole 'courting us' thing rubs me the wrong way. I've reached out to some of the people (who shall not be named) about their game content and it's not like THEY begged them. More the other way around. I mean, their product isn't THAT special.. Edit: Also, the projects on KS that don't have a hardware project were most likely ones that existed before they changed their rules, and rightfully so! -Mux

come on guys.. I'll stay cautiously optimistic but if you spent a year fucking around with molds and what not, their hardware guy could have cobbled together SOMETHING. I don't care how lame it is. Just f'ing bring it tomorrow or you run no chance at all.. Then again, you might get enough money, fuck up the project and taint retro consoles forever.. who knows.. At least Andre LaMothe documented his entire process with Hydra despite the fact that he switched from a dedicated CPU + fpga to a parallax propellor. There's no reason these guys can't.. Anyway, I'm done.. Hope to be pleasantly surprised tomorrow.. -Mux

Additionally, if they're doing that, why not get an ARM micro and add an EVE2 to it? http://www.ftdichip.com/EVE.htm James Bowman did this for Gameduino 2 and all things considered, it's pretty beefy. Thing is, now you're straying away from what's considered, at least in my opinion, retro.. Ahhh, it's getting more confusing by the minute :-) -Mux

Sooooo, with that in mind.. Why not go for software emulation in that case? I mean, if you've got a beefy enough SoC that can output HDMI and all those bells and whistles, the case for including an FPGA becomes kinda moot IMO. Just a thought... -Mux

Absolutely. They're treading on thin ice here. While there are plenty of people emulating (hardware and software) consoles, they're not making money from it. It's somewhere in the same neighborhood as the Repton devices and what not. Given that they're in the US they'll be quite open to litigation. Sega's licensed their IP over to Jax Pacific (IIRC) a while back but at that point it becomes an expensive venture. Maybe that's why they're shooting for $2-3m, so they can bundle it with a few consoles? -Mux

Sure, if you're just showing the board then by itself it's not an accurate representation. On the other hand, if you show that board hooked up to a PC, with a custom debugger and file interface running Metal Slug and divulge information about your actual system, that's an entirely different story IMO. But yeah, I can see your point as well. Either way, they should be able show something and they haven't so far, so it's all moot.. -Mux

Usually an FPGA+ARM combo is just used for communication / re-configuration. So the CPU will take care of handling the USB stack, communicating with the host software on the PC and, in case a cartridge is there, configuring the FPGA. Once that's done, it's all up to the FPGA. -Mux