@HammerD
Ralph said Hyperion never asked him any doubt. Hyperion never reported Ralph any problem about his scsi drivers. They never asked him about documentation or help. BTW, I think no documentation would be required, perhaps simply asking him something like "we change these registers to that value like we did on some PCI cards but it doesn't work like it should, do you know if we have to modify any other register to init the scsi correctly?" may be enough.
BTW, they could also take a look at linux APUS, AFAIK BlizzardPPC SCSI works in latest Linux Apus versions.
I don't have to act as a "mediator" since I'm not selling OS4 and since they are the interested party in making OS4 succeed they should ask Ralph for help firstly. If they ever had received a "no" from Ralph I would act as a "mediator" and ask politely to Ralph, but since they don't seem to be interested in making their own OS successful and forgot about classic users I won't waste my time trying to get information about something that won't be released. I bought a copy despiting I had it for free due to betatesting though.
Some people in the dev team liked to claim OS3 scsi drivers were "broken" but they never informed about these "terrible" bugs so Ralph will hardly fix them if you never report them. I guess it's due to OS4 broken virtual memory system not being able to use proper amiga DMA (the same reason most zorro dma cards won't work...). IMHO when everything works and you release an incompatible system that breaks everything that worked without problems, claiming that all amiga dma drivers are broken is kinda funny.
I don't know the reasons why the cybppc.device was developed as it was for OS4. I know it has to do some magic to get around the embedded cybppc.device in the cyberstorm flashrom so you get about 12-13MB/sec maximum out of it. On the same machine with OS 3.9 I get about 18.5MB/sec (as a reference). So the OS4 cybppc.device is a bit slower under OS4.
I wasn't involved at the time that it was being developed (for OS4). And I don't know the history behind the bug reporting, nor of any discussions with (or without) Ralph Schmidt.
I wouldn't say OS4's virtual memory is "broken", just that for OS4 Classic paging is not (yet) supported and older drivers would obviously need to be updated. Hyperion has not made any official statements on the future of OS4 Classic. So I would say if you are interested in further updates to OS4 Classic then let your views be known by e-mailing Hyperion and/or Acube (which is the distributor).
Having said that, it's next to impossible to find developers willing to work on Classic still, especially hardware drivers. Unless maybe if they were (well) paid and still had working Classic hardware (which is rare). I personally don't have the money to pay them.
Most active developers have moved on to AmigaOne/Sam, or for the MorphOS team Pegasos 2/Mac Mini etc. Most don't have Classic hardware anymore at all.
