Last visit was: Thu May 01, 2025 11:49 am
|
It is currently Thu May 01, 2025 11:49 am
|
| Author |
Message |
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
draft
You do not have the required permissions to view the files attached to this post.
|
| Fri Feb 14, 2025 6:36 pm |
|
|
|
robfinch
Joined: Sat Feb 02, 2013 9:40 am
Posts: 2307
Location: Canada
|
Is that document scanned from 1969?
Back in high-school before I had a computer, I used to sketch out computer circuitry on graph paper. That was 1980 ish. I scanned and kept some of the sketches, still have them in a folder somewhere.
_________________Robert Finch http://www.finitron.ca
|
| Sun Feb 16, 2025 10:14 am |
|
|
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
robfinch wrote: Is that document scanned from 1969?
Back in high-school before I had a computer, I used to sketch out computer circuitry on graph paper. That was 1980 ish. I scanned and kept some of the sketches, still have them in a folder somewhere. That was more like 5 minutes before the post. In 1969 I had more interest in the US space program. I had a used IBM 2310 disk drive for a while around 1983 and several other parts from a 70's IBM-1130. does that count? Todays update.
You do not have the required permissions to view the files attached to this post.
|
| Sun Feb 16, 2025 11:11 pm |
|
|
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
Still plugging away with the disc interface and file system.
I need 16Kb for disc data and stack, with ~ 48Kb for the small C compiler.
Block size will be 1Kb and 4 sectors per track. Since I plan to use the IBM 2310
as a model I can only get 1.5 Mb per disc using a better encoding than from 1965.
A tad larger than a 1.4Mb HD floppy and about the same speed.
|
| Sun Feb 23, 2025 8:59 pm |
|
|
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
Added a compare instruction, in the style Skip if AC <> (Y).
Here the Carry := AC == (Y).
Now to cleanup the software for a more standard use of registers.
Ben.
|
| Tue Feb 25, 2025 8:08 am |
|
|
|
robfinch
Joined: Sat Feb 02, 2013 9:40 am
Posts: 2307
Location: Canada
|
Quote: Added a compare instruction, in the style Skip if AC <> (Y). This is in addition to other compare instructions? Would it not be better to skip if AC==Y for counted loops? Maybe an example of usage would help. _________________Robert Finch http://www.finitron.ca
|
| Mon Mar 03, 2025 4:57 am |
|
|
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
robfinch wrote: Quote: Added a compare instruction, in the style Skip if AC <> (Y). This is in addition to other compare instructions? Would it not be better to skip if AC==Y for counted loops? Maybe an example of usage would help. The real instruction is CPE Compare if Equal. The carry flag is set if AC == Y, with no updating the AC. CPE A # '% JNC A # NORMAL LD.B A X++ CPE A # o JC A # DOOCTAL CPE A # d JC A # DODECIMAL ... // FALL TRU NORMAL // PRINT CHAR
|
| Mon Mar 03, 2025 6:01 am |
|
|
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
Starting to bring up the small C compiler. Right now I am doing basic code generation testing, using the bootstrap
i/o. The library printf routines is crashing somewhere, it could be byte alignment or a file handle problem.
Most of the file io routines are written, I just need a simple shell written as well. This is 1/3 of a milestone to having a
high level language and file io. Total memory looks to be 80 Kb, with ~16Kb for the FAT table, 1280 byte blocks, 1656
blocks per disk, 2Mb per disk. This I hope will be ample for self compiling 1975 ish.
Ben.
PS:Found a micocode bug,this I suspect is why printf was crashing wrong data was being read.Will test later
but not at 3am.
|
| Wed Mar 12, 2025 11:45 pm |
|
|
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
Still need to finish the disc routines, but got side traced by more hardware ideas after reading the paper.
*New Insights from Old Programs - The Structure of the First ALGOL 60 System.*
The X1 computer was 27 bits, I could do a 32 bit cpu (no byte addressing) and have 18 bits of addressing.
|
| Sat Mar 15, 2025 6:56 pm |
|
|
|
DockLazy
Joined: Sun Mar 27, 2022 12:11 am
Posts: 58
|
Haven't heard of the Electrologica X1 before. Has some interesting features.
Getting an Algol 60 compiler to run on a machine that limited, is something else. I might go as far to say it's the programmers equivalent of the dark arts. Should be an interesting read.
|
| Sat Mar 15, 2025 10:04 pm |
|
|
|
BigEd
Joined: Wed Jan 09, 2013 6:54 pm
Posts: 1821
|
oldben wrote: Still need to finish the disc routines, but got side traced by more hardware ideas after reading the paper.
*New Insights from Old Programs - The Structure of the First ALGOL 60 System.* Very interesting - this is the landing page you need, for the 360 page PDF.
|
| Sun Mar 16, 2025 9:20 am |
|
|
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
Revised the order code, to one I wanted for the longest time. The original software was used with a FPGA design, that would only route correctly with one opcode decoding. The other decoding gives me a better order code layout. Have a bug with the disc routines, with some hard coded constants, that I want relocate for better memory usage. On a side note, I got the prices of 4K dram in 1975, $20 each from Godbout. I am guessing a 16Kb card is about $750 with $400 for the dram. Other cards $250 each I guess, as ball park figure, for me to make.This version has no MMU. A similar computer is here. Code: THE ORIGINAL WOFFORD WITCH - 1975 to 1989
The original Wofford Witch was a DEC PDP-11/40 computer, running the RSTS/E timesharing operating system, and was in operation from 1975 until 1986 in the academic computer center at Wofford College in Spartanburg, South Carolina.
View the original Witch purchase order and shipping papers from 1975
The original Witch had core memory, LA36 DecWriters, VT50 DecScopes, ASR-33 Teletypes, an HP7200A Plotter, and RK05 disk drives.
Here is a summary of the original Witch specifications, extracted from the 1975 purchase order:
PDP 11/40 SYSTEM with two VT50 DECscopes, an LA36 console, and another LA36 for the computer room.
11/40-BK 11/40 16K PAR LA36 115V 60HZ 16,800
MF11-UP 16K SEN MEM LOGIC + 16K IMPL 6,300
MM11-UP 16K SEN MEM BOARDS ONLY 5,600
KW11-L REAL TIME CLOCK 390
BM873-YA RESTART/LDR-DSKS,MT,DT,C&T,P 600
KE11-E EXP ONST:MUL,DIV,ASR,ASMC 1,470
KT11-D MEMORY MANAGEMENT 2,600
RK11-DE CONTROL + 1ST RK05, 115V 60HZ 11,000
RK05-AA 1.2M WORD PACK 115V 60HZ 5,100
DD11-B SYSTEM UNIT PREWIRE FOR DF 275
QR430-AE RSTS/E, LIC,BIN,SRV,DECPACK 5,000
DH11-AA 16CH PROG ASYNC MPXR 115V 4,700
DM11-DA (2) LINE ADAPTER $200x2 = 400
H960-DA CABINET W/1 EXPANDER BOX 60HZ 3,000
VT50-AA (2) KBD CRT 20MA 115V 60HZ 1250x2 = 2,500
LA36-CA (1) DECWRTR II, 20MA, 115V 50/60 HZ 1,650
CREDITS -16,400
TOTAL 50,985
This proves you could buy a computer, but not any more in 1975. Godbout had surplus 8K x 18 core for $275 (Byte Sept,1975). http://woffordwitch.com/Witch.asp
|
| Fri Mar 21, 2025 1:50 pm |
|
|
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
Found the logic bug, a typing error in a ld instruction.
Now I need 64Kb + 8K rom.
|
| Fri Mar 21, 2025 10:10 pm |
|
|
|
robfinch
Joined: Sat Feb 02, 2013 9:40 am
Posts: 2307
Location: Canada
|
Quote: ound the logic bug, a typing error in a ld instruction.
Now I need 64Kb + 8K rom. I've found it is almost something simple like a typo that really causes issues. The stuff one thinks is hard turns out to be working. _________________Robert Finch http://www.finitron.ca
|
| Sat Mar 22, 2025 4:02 am |
|
|
|
oldben
Joined: Mon Oct 07, 2019 2:41 am
Posts: 768
|
robfinch wrote: Quote: ound the logic bug, a typing error in a ld instruction.
Now I need 64Kb + 8K rom. I've found it is almost something simple like a typo that really causes issues. The stuff one thinks is hard turns out to be working. My biggest problem is rollbacks is I don't have version control software. If a design idea fails or I make a bad mistake like updating the or deleting the wrong version it hard to get back from a save a few generations before. Was that foo98 or foobar23? The other catch is need my software at some point to be self hosting, and Small C from Dr Dobbs is the only portable compiler that I could find that has a simple code generator and still only needs 16 addressing. I have the chicken and the egg problem, I need a OS to port a compiler, but I need the Compiler now to write the OS. I was wishing many many moons ago, to get something like Minux ported but it uses X86 memory management rendering it non portable for a non intel cpu. A fat indexed style of file I/O is my best option something like here: File Integrity in a Disc-Based Multi-Access System see Operating Systems Techniques, C. A. R. Hoare and R. H. Perrott, Eds., Academic Press, New York, 1972, 227–248.
|
| Sat Mar 22, 2025 5:11 am |
|
Who is online |
Users browsing this forum: claudebot and 0 guests |
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum
|
|
