Talk:UxROM

@lidnariq : Seriously, this paragraph was a bunch of shit, it is speculatively mentionning non-existant oversize variations, speculatively mentionning a multicart that never existed and speculatively talked about a non-existing mapper that isn't even mapper #2, cloning the SUROM behaviour but without a MMC1. Why did you re-insert it ? Is it any valuable when it comes to doccumenting the existing games/hardware and mappers ? Is the goal of this wiki to doccument existing things, or to speculate about what could have been done ? Bregalad (talk) 05:09, 18 March 2014 (MDT)

By now I'm starting to agree with you that because emulators have standardized on the behavior with the additional 7432, SUROM-style behavior belongs in a separate page, just as other multicart mappers get their own. Let Farid do what he does best. --Tepples (talk) 11:37, 18 March 2014 (MDT)

The entire point was that the cost of the second 74'32 can be exchanged for having every 16th bank be identical.

This is no different in scope than using an inverter to prevent bus conflicts.

Since UxROM uses discrete logic, it continues to be a perfectly viable target for physical production.

Saying the sole purpose of the wiki is to describe what was is to say that the sole purpose of the wiki is for archiving details necessary for emulation.

Finally, if you want me to take you seriously, try not cussing. —Lidnariq (talk) 12:56, 18 March 2014 (MDT)

Even if describing only what was, if it was used in a known homebrew, that makes it relevant (though it is stupid that the "known" homebrew was not named in the text, which seems in contradiction with "known"). I do think it's worth mentioning practical extensions of a mapper, provided they are clear-cut and natural extension, and not merely one of a billion other irrelevant possibilities. In this particular case, I think 512k UxROM is worth mentioning, but I don't think it's good to offer 5 different recipes for building the hardware. That's more confusing than it is helpful. We should not give users advice and alternatives based on the potential cost of chips, for example. That's a discussion to be had on the forums for the benefit of someone who actually wants to build it. For the wiki, this kind of stuff is just noise inhibiting comprehension of the page. - Rainwarrior (talk)

What exactly do you think the sections of a wiki page are for? It's not like the average person visiting the page is going to care about the Hardware section.

For that matter, why leave the comments about the 7402 or 74157 if you're worried about confusing the poor little naïf of a reader?

Nintendo released three boards (UNROM, UOROM, UN1ROM, and m180) of the relevant type. All used one 74161 and one quad 2-input logic gate.

Emulator authors decided that the platonic ideal of mapper 2 is "one bank switched, one bank fixed to the last one". But this is emulator-centric thinking.

Because BK2 was developed in an emulator, of course it counts as works like the emulators say it does.

So it's equally plausible that the concept represented by UxROM is "one quad OR gate and one latch", the same as the vast majority of the other discrete logic mappers with more than 4 bits of state.

Note that I'm arguing about UxROM (hardware), not mapper 2 (emulators). —Lidnariq (talk) 21:08, 19 March 2014 (MDT)

My objection has nothing to do with preferring emulation information to hardware information. I think it's great to have good info about the hardware. My objection is to needless speculation and hypothetical implementations. The hardware diagram and description on this page is great, at least where it's describing the actual board used. The moment it starts speculating about different ways to build it are where I think we've gone off track. "A 74HC02 quad NOR gate can be used instead..." begins a hypothetical design situation. "If an actual multiplexer (74HC157 quad 2:1) is cheaper..." is not relevant to how this board works. I would be similarly opposed if someone were to add to the page 4 different C++ implementions of the mapper, along with commentary about the strengths and weaknesses of each approach. The 512k variant is relevant, and it needs to be mentioned, but this is not the place to hash out all the permutations of possibly ways to design its hardware. Do that on the forums for the benefit of someone who is actually having trouble building it. - Rainwarrior (talk) 23:27, 19 March 2014 (MDT)

Exactly what Rainwarrior said. Anything with the last bank which is not fixed in not mapper #2 and even less UxROM and have nothing to do on this page. "Oversize" variants are worth mentioning and they are already mentioned outside of this useless and confusing paragraph. Anyone with any tad of knownledge of electronics will know several ways of implementing them, there is no need to explain how to implement everything that "could be done" on this wiki. Bregalad (talk) 01:07, 19 March 2014 (MDT)