Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation

33 Excellent

About Kaide

  • Rank
    Chopper Commander
  1. To put it bluntly, because of things like vertical integration and Amazon's volume getting them deals with carriers they work with, Analogue will never be able to offer shipping as cheap as Amazon. For Amazon, their game is really fungible goods. You don't care how you get your copy of Sakura Wars, since they are all equivalent, for the most part. You care that you get a copy of Sakura Wars. So it makes a lot of sense for Amazons to play games with margins and integrate vertically to cut costs, and then convert shipping into a hidden cost baked into the MSRP when possible. It lets Amazon reap extra profits from customers that are cheaper to ship to. Analogue isn't a retailer in the sense that Amazon is, doesn't deal with the volume, and doesn't own their own logistics chain. They could roll the logistics costs into the advertised price like Amazon, but I guess for some reason they like having the advertised price be lower. It is an old-school way of doing things. The Super NT and Mega SG are both "under 200$". Roll in the costs (at least for US customers), and they can't claim that anymore, but they would get fatter margins from folks that are cheaper to ship to in the US. I can't really say if they should or not, they are both trade-offs. Would their sales drop at "$219" instead of "$189"? No idea. But the fact that they tend to run lean on stock in general suggests that they probably aren't hurting too much for a boutique company doing it as they do. Another option is that they could use Amazon as the logistics company, but it looks like they'd have to bump the price up to around 230$ anyways just to cover the FBA fees (ignoring the costs of returns for a moment). So not really any better than using their existing logistics company.
  2. It’s more a comparison that just because a rejigger of a circuit design is being done by the same folks that designed the thing originally, or have access to the original documentation and/or circuit designs, doesn’t mean it’s going to be functionally identical. The key difference between an “official clone” and a “3rd party” clone is mostly how much reverse engineering you have to do, and thus the risk of introducing bugs to the circuit because of that work. Yeah, this is where I (as an engineer) tend to clash with folks in marketing. From an engineering perspective, kevtris is cloning hardware. But on the marketing side, I’d agree that it’s hard to market something like the Super NT as a “clone” because of bootleggers and emulation boxes like the Hyperkin stuff. Really the difference is that Analogue is selling a boutique product rather than trying to produce it for the cheapest price possible, and has kevtris doing the engineering work which helps a lot.
  3. I notice that you ignore the rest of my post here. But to put it bluntly, an FPGA is no more a simulation of a electronic circuit because the circuit links are re-routable than an EV is a simulation of a car because it uses batteries as the power plant.
  4. What makes a Z80 real? Is it because it has Z80 stamped on the package? Is it because it uses Zilog’s circuit design? Is it because it is an ASIC? Is it because it is a circuit that implements the Z80 architecture as defined by some dead tree document that specifies how the Z80 works? Would custom Z80 designs like Nintendo’s variations used in the Gameboy line count as a real Z80 or not? Is it real if I can drop it into an existing circuit without the rest of the circuit knowing that it’s been swapped out? For a more philosophical question: What’s the difference between someone who got the HDL for a Zilog Z80 and put it on an FPGA versus someone who reverse engineered the Zilog Z80 into HDL and then had an ASIC made at a factory? What’s the difference between using that Zilog Z80 HDL to make an ASIC vs putting it on an FPGA? How far do we want to take this particular Ship of Theseus? But it’s not like an FPGA is running a software program though. The programming is just the configuration of signal routing, and is effectively fixed once programmed, until you program it again. It’s more akin to EEPROM (FPGA) vs ROM (ASIC). So once you’ve loaded in the configuration that recreates a circuit diagram, it’s going to have a recreation of that circuit until told to reconfigure. I will point out that compatibility issues aren’t a great argument towards calling something an emulator or not. AMD CPUs are not considered x86 emulators, despite running into the occasional compatibility issues with software. Intel CPUs are not considered x64 emulators, despite the spec being made by AMD. Sony shoved PS1 hardware in the PS2 and yet couldn’t maintain perfect compatibility with PS1 games. Nintendo had compatibility issues with the 1-Chip SNES. Just tweaking circuits using existing designs can introduce compatibility issues, so a reverse engineered circuit is going to be at least as prone to issues. As for the CRAM dots, I’m like other folks who’ve commented that I don’t really go looking for them, or care to have them enabled.
  5. Agree, virtual machines don’t really make sense in an FPGA perspective, really. Although there is some code running on the ARM SoC, IIRC. But I’m not really sure VMs make sense there either for these products, to be honest. That said, I’m going to do something dumb and wade into the emulator vs clone debate. If we think a clone is an attempt to recreate the hardware, and that emulation is the act of creating something that imitates the original hardware on some other hardware, FPGAs are in a weird grey area. I can understand the debate, honestly. On one hand, what kevtris does is the same thing that someone cloning the hardware would do: reverse engineer the original by analyzing it and creating a new description using HDL or circuit diagram. Someone writing an emulator is not seeking to create the HDL, but rather software that can run on some foreign CPU. But the FPGA is effectively “other hardware” that runs the clone, so what makes the Super NT different than BSNES? For my part, I think the distinction doesn’t really matter that closely as much of the quality of the final result, but the distinction should probably be what the output of the process is. You could take kevtris’ cores and with tweaks, produce ASICs from them and produce clone hardware that’s similar to the Super NT. You cannot do that with BSNES.
  6. In this case it looks like my shipping label entered the system on the 20th, but yeah, they probably shouldn’t attempt to estimate delivery dates based on “company printed a label” and instead wait for the package itself to enter the system.
  7. The one thing I miss with my OLED though is good motion resolution for film without resorting to interpolation. The lack of vertical blanking like with a CRT isn't great, but my 1080p LCD HDTV could emulate it with a rolling backlight and it worked great for film and 30fps games. Agree on the burn-in comments though. The Image retention is certainly a bit annoying because the TV keeps getting left on the streaming box's home screen, but don't have any actual burn in. But letting the TV go through it's process for helping clear retention every X hours is pretty important with these sets.
  8. Might depend on "redesign". If redesign is "we found a more appropriate edge connector and used it", then maybe not as much of a guarantee. But if it's an Analogue special like the GG connector, then I'd agree.
  9. I had to use a separate browser because I hit that issue. My iPad wouldn't let the order go in, but my Mac would. Weird. (And this will be my second NT Mini... sold the first one because I didn't lose any money and expected a plastic one to come out)
  10. That was only true of the original NT. The NT Mini was the first FPGA console Analogue sold.
  11. Would you if the risk of getting it wrong is causing component failure in a device that’s around 10x more expensive? Best case, a botched repair breaks the adapter completely. Worst case, it “works” until it damages the SG as if no repair was done.
  12. For grins, I examined the board of my v1.0 GG adapter that showed up today. I mostly gave up on disassembling it when I hit the tabs that hold it together. That with the black PCB meant it would take more time than I had to try to puzzle out the circuit. It can be disassembled, but the tabs are awkwardly positioned to get at. With only 4 components on the board, I suspect the bodge to apply the v1.1 fix may require physically breaking a trace or two, unless it’s lucky enough to be related to one of the 3 resistors on the board.
  13. That’s a fair point. However, not everything is or needs to be in the VHDL. Anything running on the ARM SoC can be GPL’d as usual. Depends on where this blend routine is implemented. Is it in the core or the SoC code? Although if it is in the core, there’s still something to be said about not yoinking IP from a community that is already suspicious (unwarranted as it may be). And at the end of the day, anything like this *should* be handled by a lawyer.
  14. There's the GNU/Stallman camp where the thinking is that if the software isn't free (as in speech *and* beer), then it's a violation of how intellectual property should work for software. I believe in a strong public domain, personally, something more in line with the original copyright act in the US, not what we have today (Life+X, where X is whatever Disney lobbied for last). But honestly, I see the value in letting creators experiment and decide how to share their work in the short term. Open, closed, shouldn't matter. At least not when people need to be able to get paid to put food on the table. It could also be some misplaced anger due to the behavior of companies like Hyperkin, that effectively profit off the OSS emulator work without contributing anything back. Or all of the above? I'll say that the work Kevtris has done has helped get me invested in holding onto a collection of actual cartridges again. The 16-bit consoles were some of my favorites, so having the Super NT and Mega SG really helps me preserve some of that nostalgia.
  15. I’ve already pointed out this is wrong. The only legally safe reimplementation is a clean room implementation. Not doing so risks it being marked as a derivative work. And no, this isn’t why GPL isn’t open source. It generally isn’t considered an OSS license because it requires derivative works also be GPL and it’s restrictive in that including GPL code in a project requires that the whole project become GPL (hence the need for LGPL which isn’t quite so invasive). And this is all by design. MIT or BSD licenses are easier to work with, and may be what you are thinking of. And unless the author is willing to relicense the blending code to Analogue under a different license, it doesn’t matter what they say on the internet. The license on the code still applies. And that may be tough if the code was contributed directly to a project like MISTer, since the maintainers of that project may also have say in where the code can be used. Code ownership in projects with multiple contributors gets messy. It is not as clear cut like you claim.
  • Create New...