SpritesMind.Net

Some members told me the YM2612 (wonder mega) topic became too long.

If you think something should be done, let me know

At the moment that topic is the source of information for YM2612 of the whole internet. We have to be careful about messing with it as there are links pointing to it all over the internet...

As I said before, leave discussion and create separate topic for pure checked info only.

We need somebody to scourge through it and make proper reference documentation from the data there.

The problem isn't even the length, it's that it's a disorganized mess where unless you know the exact page ahead of time you'll be screwed if you're looking up for something in particular (not to mention having to remember which information is valid and which had been proven wrong, which you can't even know until you've gone through all the thread).

For what it's worth, I did make an index of some of the more interesting pages in my aggregation thread. There's probably some more pages that should get added to the list (especially towards the end), but most of the core stuff is linked. It's obviously no replacement for proper documentation, but it's better than sifting through 56 pages by hand.

"something else"
I agree with HardWareMan.

Mask of Destiny nice thread.

Sik wrote:We need somebody to scourge through it and make proper reference documentation from the data there.

The problem isn't even the length, it's that it's a disorganized mess where unless you know the exact page ahead of time you'll be screwed if you're looking up for something in particular (not to mention having to remember which information is valid and which had been proven wrong, which you can't even know until you've gone through all the thread).

I made a document summarizing the thread as part of the Wiki of JT12. Check it out here. I actually spent a whole Saturday reading the thread and extracting the information. I assume that items that I have not document in that page can be extracted from simulation of JT12 (using iVerilog or other Verilog simulation tools) or just examining the Verilog code. Am I saying that JT12 is an exact YM2612. Well, with the available information on the internet, it is the closest thing. I will try to keep improving it.

I would like the present thread to be preserved but closed and then open a subforum for YM2612 hardware information where we can have separate threads for each aspect of the chip: PCM, LFO, EG, PG... Eventually a proper documentation collecting all the information has to come up and my Wiki is a first attempt to collect the information.

I plan to take exact data measurements from YM2612 using the test modes and a dedicated board. It would be great if we could host the data files I will generate as part of the forum or the main SM site.

I think the YM2612 does not really need a dedicated section, even for the technical part. And the very nice resume you made is a proof of it Maybe just having a topic which summarize acquired knowledge from the big topic is enough And i think we should have that kind of topic for each part of the system (68000, Z80, VDP IO, VDP rendering, BUS arbitrer, PSG, YM2612..)

Edit: bit of shame i made a typo by forgetting the important 'not' word here :p

I disagree with closing YM2612 topic. It should stay as discussion thread.
Also I disagree with making LFO, EG, PG threads.

Threads should made when someone have question about it, and he make topic with this particular question. As any other thread.
Sound implies not too many themes at all: YM2612, PSG, PCM, sound engines, theory, and little bit more.
I don't see reason for making YM2612 dedicated section. Not so many people making YM2612 emulation.

r57shell wrote:I disagree with closing YM2612 topic. It should stay as discussion thread.
Also I disagree with making LFO, EG, PG threads.

Threads should made when someone have question about it, and he make topic with this particular question. As any other thread.
Sound implies not too many themes at all: YM2612, PSG, PCM, sound engines, theory, and little bit more.
I don't see reason for making YM2612 dedicated section. Not so many people making YM2612 emulation.

The reason is that the current thread is too long and it will keep getting longer.

I agree with separating out particular discussion points into separate threads.

It makes it a hell of a lot easier to fish through specific material, than shuffling through a monolithic 50+ page thread.

Having a sub forum for YM2612 dedicated topics is a good idea but seems like a hard task to initialize and maintain. Each subthread should start with a post summarizing the current 'verified' informations we have on that part of the chip and kept updated according to the following thread discussion.

The thing with the current thread is that most of the 'good' informations (apart from a few long detailled posts by Nemesis) are indeed scattered all over the thread but also generally integrated in on-going discussion, so it would be hard to split that thread and pick parts to put in dedicated subthreads. It also still has on-going discussions or unanswered questions so I don't think closing it is a good idea. We know it contains wrong infos or misassumptions, some of them that were corrected later in the thread but some might be proven wrong or completed sometime by someone, who knows?

I mean, it always was more a general discussion thread about YM2612 than really an 'authoritative' thread anyway (maybe the topic title is a bit misleading and should be changed, as once suggested by Steve Snake) so let be it like that, it will eventually die once dedicated subtopics will be created since new discussions will more likely happen there. Unpinning the old thread and changing its title might also help.

The problem with threads is that they're for discussions. They aren't technical references, they're people talking to each other showing progress on their research and what they're doing and trying to sort out what's going on, etc. I'd say jotego's link is much closer to a good solution than yet more threads.

By all means keep doing research in threads (that's what they're good for), but as far as an actual place to be used as reference, we need a proper doc, not threads.

So if anything could be done, it must come from individuals ?
like each subject "work in progress" to its dedicated topic ?

It is a one of possible variant.