« first day (575 days earlier)      last day (874 days later) » 

Anonymous
12:01 AM
Time for another rep cap :D
 
12:32 AM
So weird, QFTASM doesn't work on Chrome for iOS either...
 
12:48 AM
@Mego Even better: I can just change the database entry.
 
1:01 AM
reddit.com/r/programming/comments/701tzd/… Should we point this guy at StackOverflow and answer the question ourselves? :P
@PhiNotPi I was able to reproduce the bug easily.
 
1:20 AM
@El'endiaStarman there might be multiple bugs?
 
Maybe? I was able to stick it to the wall on both sides.
 
1:39 AM
@El'endiaStarman uhh... the code in the permalink is the old code apparently
 
hiiii
I was mostly wondering how the rotation system works
Is it ARS? (Sorry didn't try it out myself yet)
 
@PhiNotPi I can confirm this one works.
 
2:01 AM
@PhiNotPi Could Cogol benefit from a pre-initialized RAM as well? I definitely see some stack initialization at the beginning
probably not as much as GCC will
@Avery I believe it's Sega rotation?
 
@PhiNotPi Surgery complete!
 
@quartata hm, interesting
 
I'll be honest, I'm not much of a Tetris enthusiast so I'm probably wrong
I recommend trying it out yourself in the QFTASM interpreter
scroll down to the ram display
and put in for the addresses to display "3,5-31"
then run and watch the display
control is using the "direct write to ram"
put 1 in for the address
value     motion
   1      counterclockwise rotation
   2      left
   4      down (soft drop)
   8      right
  16      clockwise rotation
there's no wall kicks if that's what you're wondering at any rate
 
More simply: click this: play.starmaninnovations.com/qftasm/#jllHdnBGSP , hit "Run", scroll down to the RAM display, and then follow quartata's instructions after that.
 
oh shoot that was the old permalink
we should update the post
 
2:13 AM
No need.
 
@El'endiaStarman Maybe make the ROM display collapsible by the way
 
I updated the program that the permalink points to in the database so everyone that clicks on that permalink will now have the bug-free code.
 
it takes up a ton of vertical space
@El'endiaStarman ahh nice
 
@quartata I've considered that, yeah.
 
2:34 AM
Petition to remove rep-cap.
 
Anonymous
@PhiNotPi Seconded :P
 
nuu
can't have el'endia overtaking me in rep
 
@quartata Your days are numbered.
 
n-nuuu
This is what I found for truth machine btw:
. SUB A1 2 0;
. MNZ 0 0 0;
There's several ways of doing it with two instructions I suspect
I just picked the strangest one
 
@quartata how are you defining STDIN/OUT?
 
2:44 AM
Input in address 1
basically the same as taking input from a variable
which I allowed if there's no other way
2 instructions is 14.5 bytes
oh wait a second
I'm so dumb
how are we going to "output" writing to a register isn't good enough
Maybe I could set each address to 1 one at a time?
I guess truth machine is a bad pick
 
> truth machine is a bad pick
lol
 
I guess I'll try to golf primes then instead
the trouble with that is no multiplication
I mean multiplication is easy to do but it's long
 
Flashback to architecture design, where we had the idea for I/O ports for the processor.
 
Anonymous
@PhiNotPi That's still not a terrible idea
 
Anonymous
We have spare opcodes
 
2:54 AM
We could hook the busses up to different sink options in the interpreter perhaps
 
The original idea was to have the ports memory-mapped with RAM.
Using the space opcodes could also work.
 
actually
This could be useful for Pong
If we make it so that in the interpreter you could either redrect the busses to a console OR to a Websocket, it could make for better controls and online multiplayer
 
Anonymous
@quartata That's a good idea
 
Anonymous
We'd still want the ports represented in the pattern, but the interpreter can watch those and do special things with them
 
Obviously the computer can't do that in practice but in theory the busses could be hooked up to other computers
all we need is @KZhang
 
Anonymous
3:01 AM
So for Pong, we could work on networking as the next big step, and then showcase 2 QFT computers playing Pong against each other
 
!!
Yeah I'm liking this
 
I'm not 100% on board with ports. Could someone tell me what should happen for ports?
 
Oh gosh, websockets.
 
So.. memory mapping or special instructions?
 
Anonymous
We would only need websockets for remote play. For the Pong demonstration, we could have 2 QFT computers connected in the same pattern. One for the player, and one for the AI.
 
3:05 AM
I'm personally against memory mapping
 
Anonymous
I think special instructions would be better in the long run
 
We technically could do both
 
SLZ and SNZ for sending
 
Anonymous
RLZ and RNZ for receiving
 
I was thinking just GET and AVA for how much is available
This is the part where KZhang tells us that buffered IO is way too hard
 
3:10 AM
I mean, we could just use a single synchronizer
The issue with that is if too much data is sent, some might get lost
 
That's OK IMO
The question is how big a buffer is feasible
 
I could do it so that only the most recent piece of data is stored
Although, I would need to figure out something else for buffering
It might be doable
 
You could perhaps provide a way for a clock signal of one computer to pause and wait for the other computer.
 
Anonymous
@El'endiaStarman I think that's a bit too complicated. The computers will be running in sync anyway, so we can just program/design around that.
 
Anonymous
Essentially we're building one big computer made out of two smaller computers. The same design principles that we employed for the original computer will still work for this.
 
3:19 AM
@Mego Not with the signal taking different amounts of time through the ROM. (It does take longer to get the last instruction, doesn't it?)
 
Anonymous
@El'endiaStarman I believe so, but we can manage that through software (especially with PIC)
 
Anonymous
Idea: introduce a delay component to the ROM so that every instruction takes the same amount of time to fetch, regardless of the address
 
Hrmmmmmm
Blocking will be tough too
We'd need a lock
We should probably do the interpreter first
@El'endiaStarman I was thinking a drop down menu that controls where the input bus reads from and where the output bus writes to
1 would be a simple console you could type into
2 would be the websocket
3 could be reading from a file (not for output obviously)
 
3:49 AM
Also yay I just realized I've been doing this wrong: reg doesn't mean a hard register
Our GH repos are a total mess :/
 
Anonymous
4:13 AM
@quartata Yep, I've been thinking about cleaning them up
 
7:51 AM
It would be cool to have it so we can have a console interface for the project
 
 
2 hours later…
9:43 AM
@Mego I wouldn't be surprised if that happens.
@quartata expect some new traffic
 
:D
Feels so good to have all these people acknowledging us
 
 
1 hour later…
10:52 AM
I like how the site analytics show where links to the project were posted
 
@muddyfish nuuuuu I don't have enough rep to see
 
@ASCII-only ^
 
11:17 AM
> yahoo 0
Lol
 
11:39 AM
I'm surprised you guys haven't made the news yet.
 
I wouldn't really expect that
 
the GoL clock did
 
it did?
 
well niche news, obviously. but still good traffic
I think that wasn't the only one
 
 
2 hours later…
1:39 PM
I'm surprised at how many Python modules there are for reading ELF...
decisions
 
 
3 hours later…
4:46 PM
@quartata ELF?
 
In computing, the Executable and Linkable Format (ELF, formerly named Extensible Linking Format), is a common standard file format for executable files, object code, shared libraries, and core dumps. First published in the specification for the application binary interface (ABI) of the Unix operating system version named System V Release 4 (SVR4), and later in the Tool Interface Standard, it was quickly accepted among different vendors of Unix systems. In 1999, it was chosen as the standard binary file format for Unix and Unix-like systems on x86 processors by the 86open project. By design, ELF...
 
Anonymous
5:00 PM
@El'endiaStarman The best Will Ferrell movie
 
Every time I google it I swear
I have to do "elf format" or "elf object"
If I just do Elf I see Will Ferrell's smug face every time
 
Should I add a link to this chatroom to the overview post?
 
I think that would be a good idea for people with questions, although most people seem to be finding it on their own
 
5:20 PM
The overview post is already #9 of most upvoted CG.SE answers ever.
 
guess which question has the most upvotes now
 
this one
 
correct
 
We should submit this whole thing to Code Review
3
 
@PhiNotPi *cackles maniacally*
 
5:35 PM
hrmm so this doesn't really let me modify elf files
why aren't there just bindings to libelf
pure python is for plebs anyways
 
Anonymous
6:29 PM
@quartata :(
 
Anonymous
8:26 PM
Could someone update the GoL MC file in the GitHub repo with the new ROM? My computer and Golly are not getting along right now.
 

« first day (575 days earlier)      last day (874 days later) »