cr1901 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://libera.irclog.whitequark.org/~h~yamahasynths
<NiGHTS>
William D. Jones sur Twitter : "Okay, it's not done yet, but I'm excited! I can mark something off my bucket list: I wrote an MS-DOS Device Driver (the thing you put in CONFIG.SYS) and created a HELLO device. When read, it streams "HELLO, WORLD!!\n^Z\0" to the screen: https://t.co/fVBpm3QysS"
<sorear>
old school linux approach, assuming the 32kHz and 12MHz crystals are independent?
<cr1901>
Idk if the IBM PC has two async clocks
<cr1901>
32khz == RTC? and 12MHz is from?
<sorear>
PIT/main bus clock
<cr1901>
PIT is derived from the main clock I think. The main clock is divided to feed a 33% duty cycle clock to the 8088 at 4.77 (4/3 colorburst), and "something close to 2MHz" for the PIT.
<cr1901>
(IIRC for the second half of that statement)
andlabs has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<cr1901>
sorear: You have more info about how old school linux did it? I was gonna use an LFSR to make a PRNG :P
<sorear>
"poorly"
<cr1901>
Okay, how would I do it "not poorly" on hardware from the early/mid 80's? :)
<sorear>
the PIT cycle count observed at RTC interrupts, or vice versa, is a source of entropy but the old linux "SHA-based PRNG" is not something to emulate, maybe look at yarrow or fortuna or related projects
<sorear>
might also be worth looking at the discussion surrounding the new linux "ChaCha based PRNG" (I don't have a link, sorry, and the algorithm change is less important than the redesigned overall architecture)
<cr1901>
I got a headache looking at the Fortuna pdf lol... but something to keep in mind
<qu1j0t3>
PNG is nice if relevant at all
<qu1j0t3>
(Melissa O'Neill's)
<qu1j0t3>
simple to implement
<cr1901>
PCG?
<sorear>
link? i am getting very poor search results for "melissa o'neill" PNG
<NiGHTS>
foone sur Twitter : "Silly idea for a game: The fictional meta narrative is that the game is a pay-to-win mobile game but it has been abandoned by its creators, so the pay to win parts no longer function, as it's impossible to purchase them."
<cr1901>
When I still played stuff like that, I was patient enough to wait out the deliberately long cool down periods :P
<cr1901>
So at least that's how I would win your hypothetical game
<cr1901>
Oh that's clever. Darker colors are created by overstriking already laid-down chars
<cr1901>
Also CW: nudity for that link
<ZrX_NoMs>
wrote a little printersim to see how it plays out.
<tunixman>
Oh man I was on some random gig site that turned into basically helping college kids with their homework and one of the projects was writing to a virtual dot matrix printer.
<tunixman>
The exercise was difficult because the printhead was vertical but the data was horizontal.
<cr1901>
Can't wait to see art like this embedded in the tracks of a floppy. Though I suppose 40 lines isn't enough to create elaborate art.
<ZrX_NoMs>
this typewriter worked with 250k 3.5" floppies that were mostly incompatible between units. cheap dumb drives that had whatever the alignment from factory was.
<cr1901>
1. Why does a typewriter need a floppy drive?
<alexisvl>
I just assume it's one of those "word processors", not a true typewriter
<cr1901>
2. If you need a data xfer solution for a typewriter, wouldn't lack of alignment between drives make xfer nigh-useless?
<cr1901>
ohhh, like the Brother machines w/ the built-in CRT?
<emeb>
"word processors" were for ppl who were scared of real computers.
<alexisvl>
those were everywhere for a few years
<alexisvl>
scared? that's bs
<alexisvl>
we had one when i was a kid growing up, couldn't _afford_ a proper computer. it was way cheaper
<emeb>
Tell it to my grandpa who refused to go near a computer but didn't mind having a "typewriter with a floppy disk"
<alexisvl>
surely that means that's why everyone had one
<cr1901>
ZrX_NoMs: What model is that typewriter/word processor?
<cr1901>
emeb: I can think of a few advantages to a word processor w/ CRT compared to a luggable.
<ZrX_NoMs>
Brother something 30
<cr1901>
Admittedly, I'm biased by the fact that my experience w/ Compaq Portable I is... less than positive
<emeb>
cr1901: sure. a lot of the low-end WPs were way smaller/lighter. Could be folded up and kept in a closet when not needed.
<alexisvl>
yeah, ours was really small for having a usably sized crt in it
<cr1901>
lighter was definitely a big one. Only put in the circuitry you need to edit documents :D
<cr1901>
Wonder if anyone has RE'd the ROMs on those Brother machines
<emeb>
the ones I remember didn't even have CRTs - they had a few lines on an LCD.
<NiGHTS>
Mark Janello sur Twitter : "toasters are computers with special floppy disks made of bread… "
<emeb>
rofl
<tunixman>
nice.
<ZrX_NoMs>
can't remember the cpu. disks were gcr like apple and commodore, but with very primitive fat-like filesystem.
<Lord_Nightmare>
ZrX_NoMs: victor 9000?
<Lord_Nightmare>
aka sirius 1?
<Lord_Nightmare>
that uses commodore-style GCR but on 1.2MB 5.25 HD disks
<Lord_Nightmare>
and has a weird proto-FAT filesystem which is like some horrible combination of FAT12 and the pre-fat filesystem used by some versions of cp/m
<Lord_Nightmare>
with the worst of both worlds
<andlabs>
nice @ ZrX_NoMs's art
<andlabs>
lol @ cr1901's tweet
<andlabs>
victor 9000? isn't that the computer made by chuck peddle?
<andlabs>
*that* chuck peddle?
<Lord_Nightmare>
yes
<Lord_Nightmare>
and that's why it uses the same gcr format as commodore used
<andlabs>
yes it was lol
<Lord_Nightmare>
since chuck peddle designed that too
<andlabs>
I'm still amused by that
<andlabs>
if you can't beat em, join em™
<andlabs>
anyway FAT12 on GCR
<andlabs>
how to somehow be even less useful than the CP/M cartridge for the Commodore 64
<andlabs>
I wonder if the C128's CP/M mode can read it
<andlabs>
also fun fact: there was actually commercial software for the CP/M cartridge!
<andlabs>
I own two of them, just gotta wait for them to come in the mail
<andlabs>
Also known asACT Sirius 1
<andlabs>
ManufacturerApplied Computer Techniques
<andlabs>
wait isn't that the company that made the Apricot
<andlabs>
it was lol
<andlabs>
way to go from one extreme to the other
<Lord_Nightmare>
i think the apricot first models are maybe somewhat derived from the design of the victor 9000/sirius 1
<andlabs>
heh
<Lord_Nightmare>
since kidde (who owned victor/sirius) wanted to make the computer capable of reading MFM formatted 360k disks, but the way the original hardware was designed made this difficult
<andlabs>
lol
<andlabs>
see this is the problem with floppy disks not being standardized from the start
<Lord_Nightmare>
I suspect apricot may have started out by attempting to make some sort of conversion board to replace the floppy control board in a victor 9000/sirius 1
<andlabs>
but oh well
<andlabs>
I still need to set up a proper flux dumping system but I get distracted with other things
<qu1j0t3>
andlabs: Are they standardised now? ever? :)
<Lord_Nightmare>
and this may have eventually mutated into the first apricot model
<andlabs>
qu1j0t3: yes that's my point
<andlabs>
they are forcibly standardized now, at least
<andlabs>
because USB doesn't have any provisions for anything other than the IBM MFM format with the two most common DSDD/DSHD CHS counts for 3.5" disks
<andlabs>
and also the weird NEC/Fujitsu one used in Japan because Japan
<andlabs>
but also because no one uses them
<ZrX_NoMs>
Lord_Nightmare: No, the Brother typewriter.
<cr1901>
According to Wikipedia, proto FAT was "FAT8"?
<cr1901>
It is not possible to format a floppy disk with the NTFS file system; Windows NT formats all floppy disks with the FAT file system because the overhead involved in NTFS will not fit onto a floppy disk.
<NiGHTS>
Overview of FAT, HPFS, and NTFS File Systems - Windows Client | Microsoft Docs
<cr1901>
That's... hilarious to me
<tpw_rules>
i mean they lied the opposite way about FAT on big drives
<qu1j0t3>
cr1901: hahaha meanwhile iirc apple defaulted to HFS for 800k disks
andlabs_ has joined ##yamahasynths
andlabs has quit [Ping timeout: 256 seconds]
<cr1901>
tpw_rules: I thought the idea was that on big drives, FAT was slow b/c you had to traverse it strictly in order to get to other parts of a file
<cr1901>
Not that I understand how trees help w/ this
<cr1901>
(There seems to be FAT, simple linked list, and everything else, using trees of various complexity)
<tunixman>
tReEs ArE lOg N nOt N^2
<tunixman>
I guess.
<tpw_rules>
i mean yeah, but windows makes all this big deal about how 32GB is too big for that and forces you to use NTFS or exFAT despite the fact that mac and linux don't run into crushing performance problems on 2TB FAT drives
<cr1901>
Ahhh, well, might be funny to format a floppy as NTFS, but I don't think anything could read it
<qu1j0t3>
tpw_rules: which is why HFS has existed since 800k floppies and 20mb hard drives. (people ran 5MB hard drives before that, with MFS, and the results were very uninspiring)
<qu1j0t3>
and yeah, it's about B-trees having lower access time complexity
<cr1901>
I remember seeing from Glasgow experiments that 3.5" 1.44MB floppies don't bother using the index hole to figure out where to start laying down sectors. Does that mean that each track has a different starting point relative to index?
<cr1901>
Because I thought most spinning media will try to optimize reads/writes across tracks/cylinders to minimize waiting time
<cr1901>
And you'd need to know where sectors begin relative to each track to know how to optimize that
<tpw_rules>
i thought that only applied to hard disks
<tpw_rules>
but idk i'm young enough to never have used a floppy in anger
<qu1j0t3>
tpw_rules: you mean the efficiency of access problem?
<tpw_rules>
no i mean you could set up interleaving on your hard disks to optimize that. never heard of it for floppies
<qu1j0t3>
oh.
<qu1j0t3>
yeah it applies to floppies but probably not often seen
<ZrX_NoMs>
PC cares about index. Amiga doesn't. can't remember what macs did.
<cr1901>
interleaving is for within-cylinder accesses
<cr1901>
Well, Amiga doesn't because you can't read just a sector on Amiga
<cr1901>
(I think)
<ZrX_NoMs>
and then theres famicom disk. /o\
<cr1901>
spiral floppy disks don't exist in my world :)
<cr1901>
whitequark: Do you still have your Glasgow captures for 3.5" floppy disk tracks? I can't remember if you did multiple tracks or just one
<whitequark>
multiple
<cr1901>
If it's possible, could you tell me whether each track got the index pulse at different positions relative to sector 0 (or let me download the data so I can figure it out myself)?
<whitequark>
let me grab some new captures, i don't think i have any on hand
<whitequark>
how many tracks do you need?
<cr1901>
Track 0, 1, 2, 77, 78, 79 will be fine (or 37, 38, 39). Just looking for a pattern.