2023-09-27 08:36:08 <drakonis    > https://www.uiua.org/ wow neat
2023-09-27 08:50:24 <vms14       > it's quite cool
2023-09-27 08:50:36 <vms14       > it transforms words to symbols
2023-09-27 08:51:02 <vms14       > and it says you don't have to type the entire word, just enough to disambiguate
2023-09-27 09:07:55 <vms14       > it makes me wonder if i should play with this
2023-09-27 09:07:56 <vms14       > https://pdl.perl.org/
2023-09-27 09:08:08 <vms14       > i can steal it xd
2023-09-27 09:08:35 <vms14       > it's funny how many implementations of languages i can find in cpan
2023-09-27 11:00:29 <KipIngram   > Oh, neat - that will be fun to play with.
2023-09-27 11:02:40 <KipIngram   > Stack of arrays - that's right in line with what I've been thinking about.  I figure I'll support more than just arrays, but arrays will be a primary data type.
2023-09-27 11:03:19 <KipIngram   > I also want to support "multivectors," as in geometric algebra.  That would actually be a "structure" of arrays of various dimensions.
2023-09-27 11:03:50 <KipIngram   > A single stack item could specify "parts" of different order (scalar, vector, bivector etc.)
2023-09-27 15:02:45 <KipIngram   > I think this blog post has some good ideas in it:
2023-09-27 15:02:48 <KipIngram   > https://www.sacrideo.us/source-code-readability-elegance-and-complexity/
2023-09-27 15:20:28 <lagash      > What is the quickest way to determine how many bits a cell is? I'm assuming it involves CELLS?
2023-09-27 15:28:25 <GeDaMo      > If it's a standard Forth, there's an environmental query ADDRESS-UNIT-BITS which you can multiply by CELLS https://forth.sourceforge.net/standard/dpans/dpans3.htm#3.2.6
2023-09-27 15:31:47 <lagash      > Ah, well, apparently it's not. I have been messing around with DurexForth, for the C64.
2023-09-27 15:32:20 <lagash      > 2012 compatibility is somewhat lacking it seems.
2023-09-27 15:33:11 <lagash      > The C64 is an 8-bit machine, but I was fairly certain that Durex uses 16-bit cells. Just wanted to confirm. 16-bit cells is the minimum for Forth systems these days, yes?
2023-09-27 15:37:18 <KipIngram   > I don't really like the way uiua puts underscores between numbers.
2023-09-27 15:38:06 <KipIngram   > lagesh: If CELLS is defined then you can just say 1 CELLS . and it will print out bytes per cell.
2023-09-27 15:38:52 <KipIngram   > 8 CELLS . if you want bits.
2023-09-27 15:39:30 <KipIngram   > Gforth 0.7.3, Copyright (C) 1995-2008 Free Software Foundation, Inc.
2023-09-27 15:39:32 <KipIngram   > Gforth comes with ABSOLUTELY NO WARRANTY; for details type `license'
2023-09-27 15:39:34 <KipIngram   > Type `bye' to exit
2023-09-27 15:39:36 <KipIngram   > 8 cells . 64  ok
2023-09-27 15:40:14 <lagash      > OK, that returns 16.
2023-09-27 15:40:45 <KipIngram   > Been a while since I used a 16-bit system.
2023-09-27 15:44:12 <KipIngram   > Oh, I do like how uiua lets you type names and then will render them into glyphs for you.
2023-09-27 15:44:44 <KipIngram   > That is actually a good idea - reduces the need for specific keyboard support and lessens the pain of the learning curve.
2023-09-27 15:44:58 <KipIngram   > I'm stealing that idea.
2023-09-27 15:48:50 <KipIngram   > It would be nice too if positioning the cursor on a glyph would (perhaps optionally) display the name somewhere too.
2023-09-27 15:58:09 <KipIngram   > I also want to be able to type those glyps directly without having to fumble down to that right alt key - that thing is a pain to get to.
2023-09-27 15:58:27 <KipIngram   > I'm likely to use ctrl on my caps lock key instead.
2023-09-27 15:59:53 <GeDaMo      > I saw an online APL which used pairs of characters to input the special APL characters
2023-09-27 16:00:25 <KipIngram   > Do you recall how it decided when to do that vs. just two separate chars?
2023-09-27 16:00:27 <GeDaMo      > https://tryapl.org/
2023-09-27 16:00:33 <GeDaMo      > It uses tab complete
2023-09-27 16:00:52 <GeDaMo      > Hover over the characters at the top
2023-09-27 16:02:41 <KipIngram   > Oh, I see.  Well, I'm planning a console interface, so "hovering" probably isn't in the cards.
2023-09-27 16:03:13 <GeDaMo      > You don't need to hover to input
2023-09-27 16:03:34 <GeDaMo      > It just shows the character combinations
2023-09-27 16:04:27 <KipIngram   > I see now.  That get the job done.
2023-09-27 16:18:55 <Zarutian_iPad> CELLS !
2023-09-27 17:18:13 <KipIngram   > Afternoon, Zarutian_iPad
2023-09-27 17:18:33 <KipIngram   > CELLS ! Segmentation fault
2023-09-27 18:05:58 <Zarutian_iPad> be back in a bit
2023-09-27 18:47:17 <Zarutian_iPad> 
ACTION watched https://www.youtube.com/watch?v=f4s1h2YETNY and found it nifty.

2023-09-27 19:29:03 <KipIngram   > I think this "glyph" approach a la APL and uiua is the right way to go.  Requires effort to reach a "mastery" level, but I think once you do you're a more powerful programmer.  Because of that "visual information channel" I talked about re: electronic schematics and so on.  One of those Aaron Hsu links I put up earlier makes me think he has similar ways of thinking.
2023-09-27 19:29:41 <KipIngram   > He actually mentioned  "more of the program in a smaller space" making his brain able to process it better.  That's exactly the same sort of idea.
2023-09-27 19:31:43 <KipIngram   > I mentioned to xelxebar that I think Forth has some elements of that, given its use of symbols like @ and ! and so on, but not as much as APL.  Also, Forth has the spaces.  Those spaces spread things out and would lower the information density on that visual channel.  Not terribly so, but I feel pretty sure there is a factor there.
2023-09-27 19:32:12 <KipIngram   > I'm leaning very strongly toward strengthening the parser in my next effort and reducing the spacing requirements.
2023-09-27 19:33:02 <KipIngram   > Fits in with my earlier talk on expanding the capacity for structured literals, too.
2023-09-27 19:33:34 <KipIngram   > I figure if I'm going to support arrays, I need to be able to write arrays in an optimally compact fashion.
2023-09-27 19:34:09 <thrig       > shutdown all the garbage mashers on the detention block!
2023-09-27 22:05:26 <xelxebar    > KipIngram: https://mlochbaum.github.io/BQN/doc/context.html
2023-09-27 22:06:52 <xelxebar    > BQN's ideas might make parsing a lot more parsimonious.