ChanServ changed the topic of ##yamahasynths to: Channel dedicated to questions and discussion of Yamaha FM Synthesizer internals and corresponding REing. Discussion of synthesis methods similar to the Yamaha line of chips, Sound Blasters + clones, PCM chips like RF5C68, and CD theory of operation are also on-topic. Channel logs: https://freenode.irclog.whitequark.org/~h~yamahasynths
Xyz_39808 has quit [Ping timeout: 244 seconds]
<Lord_Nightmare>
Foone: that echo pc isa card is very interesting, it uses the ECHO-1 SEC microcontroller (a tsp50c50? or tsp50cxx?) instead of a tms5220
<Lord_Nightmare>
because by 1987 tms5220c chips were getting harder to obtain presumably, or echo wanted a larger fifo and the weird fifo on the tms5220c/tsp5220c was too small
<Lord_Nightmare>
the 1988-1991 echo2 apple2 cards use a smaller tsp50c0x microcontroller called 'echo-2 sec' or 'echo-3 sec' depending on when it was made
<Lord_Nightmare>
none of them are dumped, we don't even know how to trigger the test mode on those microcontrollers
<Lord_Nightmare>
the tsp50c1x microcontroller is used on the milton bradley electronic talking battleship (3 mcu versions depending on when it was made! suspected oddball bugfixes, we don't have dumps of any of the 3!), on mb omega virus, and on mb dream phone
<Lord_Nightmare>
we'd LOVE to emulate all of those
<Lord_Nightmare>
the same tsp50c0x or tsp50c1x microcontrollers are used on a few vtech talking devices as well
<Lord_Nightmare>
although the vtech socrates used a toshiba speech chip, not a ti one
<Lord_Nightmare>
type design drawings for the first hp laser printer
<Lord_Nightmare>
the laserjet (retcon-named the "laserjet 1"), probably for the built-in fonts rather than the postscript upgrade cartridge
<Lord_Nightmare>
I actually OWN a laserjet 1
<Lord_Nightmare>
its in the attic, waiting for preservation and a new cartridge assuming one can be found
<Lord_Nightmare>
those type design drawings i had assumed were destroyed in the HP museum fire
<Lord_Nightmare>
I also have the postscript cartridge, which i've been told has its own 68k in it and a bunch of very large roms holding postscript type-1 fonts
<Lord_Nightmare>
its sonically welded shut
<Lord_Nightmare>
or so i recall
<Lord_Nightmare>
it fits into the font cartridge slot
<Lord_Nightmare>
i'm now holding it... its not sonically welded but held with snap tabs that are very difficult to undo, maybe i could open it with a spudger or an opener pick
<Lord_Nightmare>
its quite heavy too, so i don't doubt that there's two pcbs in here in a stack
<Lord_Nightmare>
the type design documents for the laserjet imply the fonts came from autologic, so it must be the internal fonts
<Lord_Nightmare>
i wish i had a service manual for the hp laserjet, i could probably safely get it apart and get the firmware roms out to dump them
<Lord_Nightmare>
i believe it uses a 68k for its main cpu
<Lord_Nightmare>
but may have a sub cpu as well
<Lord_Nightmare>
it might use a 68020 but i doubt it. it could use a 68010 though
<Lord_Nightmare>
the postscript cart is from 1991, so it may well have a 68030 inside
<Lord_Nightmare>
i don't doubt for a second that the postscript cart has a much faster cpu (and likely as much ram) as the main printer does
<Lord_Nightmare>
since its effectively soft-rendering vector postscript fonts to a bitmap font buffer for the printer to basically blit into a framebuffer
sorear has quit [Read error: Connection reset by peer]
sorear has joined ##yamahasynths
<cr1901_modern>
https://twitter.com/cr1901/status/1298581379253510144 I went down this rabbit hole yesterday and found one of the original LHA decompressor's source code. It still works on x64 Windows... sort of...
<superctr__>
i guess you found a really old version of LHa
<superctr__>
i think there are newer versions that work just fine on windows, even without CRC errors
<superctr__>
and 64-bit windows still has 32-bit int, unlike linux
<cr1901_modern>
7zip decompresses just fine. I was just curious
<superctr__>
it makes references to the source code though
<superctr__>
i suppose it should be possible to understand it with google translate + code examples for context
superctr__ is now known as superctr
<cr1901_modern>
superctr: Yes, that's true, but I would of course prefer an English version. And Okumura's English is infinitely better than my Japanese. So I figured I'd ask.
<Lord_Nightmare>
where can the original src be found? i wonder if it was intended to compile on win16
<Lord_Nightmare>
on a pc88xx running windows 3.1
<Lord_Nightmare>
if that's the case, fixing it to compile properly on modern compilers is just a matter of changing some int to short in a few places
<superctr>
maybe the original version was for DOS/early windows
<superctr>
the Unix version really makes it clear it's "LHa for UNIX" so maybe that isn't the first
<superctr>
Although I think the modern releases are all based on the Unix code
<andlabs>
Lord_Nightmare: but does that font archive have phototypesetter fonts
<Lord_Nightmare>
who knows
<andlabs>
well I'll definitley be finding out
<Lord_Nightmare>
the person who ran the archive before it was done passed away 2 weeks ago, its now all in an IA storage unit
<andlabs>
getting the Graphic Systems C/A/T fonts back will bring a more than a decade long dream of mine to reality
<Lord_Nightmare>
it will eventually be gone through
<andlabs>
huh?
<Lord_Nightmare>
assuming IA still exists after the publisher lawsuit
<andlabs>
this is time limited?
<Lord_Nightmare>
huh? no. the archive was from someone in NYC who ran a typewriter repair shop and forensic type analysis business in the 60s-80s
<Lord_Nightmare>
his son took over afterward but passed away from metastatic cancer 2 weeks ago
<andlabs>
no you said [10:11:57] <Lord_Nightmare>it will eventually be gone through
<Lord_Nightmare>
before he passed away, jason scott got in contact with him and arranged the entire archives of the shop to be transferred to internet archive
<andlabs>
one sec though brb
<Lord_Nightmare>
IA now possesses the stuff
<Lord_Nightmare>
but they have a HUGE BACKLOG
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
andlabs has joined ##yamahasynths
<andlabs>
Lord_Nightmare: oh you're saying it'll be gone from the warehouse
<andlabs>
I thought you meant it'll be gone from the website
<andlabs>
but it's not even online yet
<cr1901_modern>
Lord_Nightmare: Ask him on hellsite if he has it? The copy I have came from txtfiles
<Lord_Nightmare>
andlabs: i think we're both misunderstanding
<Lord_Nightmare>
jason scott did a rescue of the entire contents of a physical typewriter store in NYC from a guy who was on hospice care for cancer, which he has since died of
<Lord_Nightmare>
that rescue included a massive number of books and resources about typography, including vintage typewriter typography going all the way from that up to modern digital fonts
<Lord_Nightmare>
this included the design docs for the fonts from the original hp laserjet (the internal fonts, not the postscript/external ones)
<Lord_Nightmare>
these artifacts, which include many physical typewriters and piles of books, manuals, and design documents, are now in a warehouse/storage unit owned/rented by the internet archive
<Lord_Nightmare>
EVENTUALLY the internet archive will go through all of it and scan the paper bits
<cr1901_modern>
Can someone be paid to expedite processing?
andlabs has quit [Ping timeout: 240 seconds]
<Lord_Nightmare>
cr1901_modern: i wouldn't be surprised if you could donate to IA and specifically earmark it toward a specific project
<Lord_Nightmare>
but you'd have to arrange that with brewster etc
<cr1901_modern>
My pragmatic opinion is "you can't save everything, but it bothers me far less if a valiant attempt was made and failed"
<cr1901_modern>
rather than nobody thought it important at all until it was too late
<cr1901_modern>
e.g. when a NES proto from May 2019 was lost b/c a collector outbid the crowdfunded bid
emeb has joined ##yamahasynths
andlabs has joined ##yamahasynths
<Foone>
Lord_Nightmare: I've been looking for a copy of that talking battleship game for a while. I had it as a kid, and it was hilariously bad, and I'd love to dump the ROM of it
<ZrX-oMs>
Lord_Nightmare: Those Eagle schematics...
<Lord_Nightmare>
zrx YES! I was just thinking of that last night
<Lord_Nightmare>
"didn't zrk offer to render those schematics?"
<Lord_Nightmare>
gotcha. Thank you so much for doing this!
<ZrX-oMs>
np
<cr1901_modern>
whitequark: I find that awardbios compression interesting because I'm looking at another LHA file I scrounged up that's plaintext when decompressed. I can make out some strings in the awardbios like what's supposed to be "GenuineIntel". >>
<whitequark>
ha, did you get sniped by that
<cr1901_modern>
Yes, very much so
<cr1901_modern>
I had been on a (entropy, like Huffman) compression kick recently, so I decided to dig.
<cr1901_modern>
In the file I have on hand, I see no discernable strings in the payload.
<cr1901_modern>
The awardbios is "dictionary coding, like Lempel-Ziv. But the dictionary entries and offset-len pairs are Huffman encoded in a "second" (not actually how its implemented) pass.
<Foone>
I wonder how much boot time and RAM is wasted having to decompress BIOSes rather than just buying a slightly bigger eeprom
SceneCAT has quit [Ping timeout: 256 seconds]
SceneCAT has joined ##yamahasynths
SceneCAT has quit [Read error: Connection reset by peer]
<cr1901_modern>
Interesting question. I guess a decent chunk of it gets out of the way after starting the boot to DOS, since I think the BIOS would be RAM shadowed at segment 0xe000 thru 0xffff