00:00:00 --- log: started forth/21.06.14 00:53:30 --- join: neuro_ joined #forth 00:53:44 Does anyone know a good example of an infix expresion parser for Forth? 00:56:48 http://www.complang.tuwien.ac.at/anton/euroforth/ef08/papers/haley.pdf 00:57:52 neuro_: I have a parser generator in Forth :) https://github.com/siraben/meta-yacc 00:58:16 It would generate an infix parser out of a grammar like https://github.com/siraben/meta-yacc/blob/master/arith.txt 00:58:28 then emitting postfix 01:01:40 Nice 01:15:10 --- join: gustaf joined #forth 01:27:15 --- quit: Glider_IRC__ (Remote host closed the connection) 01:34:38 I asked not becahse I care about infix notation, but because I want to answer better when others complain about Forth's RPN notation. 01:39:34 "Infix is for n000bzzz!!!!" Joking, there's a teenager somewhere in me who didn't die. 01:40:56 Speaking of which I started writing my Forth yesterday on Amstrad CPC with an assembler that doesn't have macros (Maxam 1.5 with Protext editor). 01:43:36 reminds me when I wrote a Forth for the R216 computer whose assembler doesn't have macros 01:43:57 so I just manually constructed the linked list via labels 01:44:16 hm, if I knew more Lua I guess I could have extended the assembler to do that 01:45:17 --- join: Glider_IRC joined #forth 01:46:06 For linked lists it supports a LET directive and there's a program counter variable $, so it seems okay. 01:46:27 Just a bit of repetition but not too much, and I like that everything looks more clear unlike macros. 01:48:20 But Z80 didn't seem very friendly for implementing the address interpreter at first, so I used a memory location for every Forth register for the time being. 02:04:35 --- join: gravicappa joined #forth 03:37:32 --- quit: cp- (Ping timeout: 272 seconds) 04:11:39 --- quit: neuro_ (Ping timeout: 245 seconds) 04:39:51 --- join: zolk3ri joined #forth 04:40:55 --- quit: zolk3ri (Client Quit) 04:52:57 --- join: crc[work] joined #forth 04:52:57 --- mode: ChanServ set +v crc[work] 04:55:30 as a minor update to last night: my login is not yet locked, and my work ip is not blocked currently. I'll use crc[work] when connecting from this. 04:57:52 proteus-guy: "upgrade from IRC" other chat tech seems to be a downgrade 04:58:13 I don't know what matrix is like, that might be better 04:58:17 Or that GNU chat thing 05:00:46 IP bans are dumb, too many people share the same IP 05:03:44 Writing a Forth with a macro-less assembler is a real drudge. 05:04:03 Building that linked list by hand gets really boring. 05:04:38 I suppose you could think of it as "paying dues." 05:04:41 KipIngram: Use a separate preprocessor 05:05:01 Like cpp or m4 05:05:13 Or get a good text editor 05:05:16 True - these days I'd look for a way to do that, or put something together in Python or something. 05:05:38 But when I did it I didn't have any of that or the experience to know to go that way. 05:15:07 veltas, xmpp is actually nicer than irc but I get your meaning. It's not a simple obvious thing. 05:15:34 Also, IRC does have an "appeal" that those more sophisticated chat systems don't. I moderate in a reddit sub, and I have Discord installed to stay in touch with the other mods. It's fine, but it just feels "heavy." And I have Slack installed to stay in touch with work peeps. It also feels heavy. And they both feel like things that used IRC as their starting point and then layered on bloat. 05:16:03 What's most important is that we support the great #forth community we built here. To be "forth-ish" it shouldn't be massively complex or require big web browsers and such just to participate. 05:16:28 KipIngram, yep - understand and agree with your sentiment. 05:16:31 We're connected on two IRC networks, seems alright 05:16:38 ^ That feels right to me. 05:16:38 Latency could be a bit lower between the networks though 05:16:51 It would be coolest if we could all write our own chat clients in Forth. 05:17:12 veltas, except we can't seem to count on these two networks. So we need to have optionality. 05:17:17 Which I think we could, given that crc wrote the bridge generally in Forth. 05:17:29 You can say that about anything, except nobody does it. Should write your web browser in Forth too 05:17:54 veltas, web browsers are the opposite of low complexity because of the protocols they support. 05:18:06 True - it's like shopping at Walmart or Amazon. Convenience has a way of overpowering philosophy. 05:18:16 Life gets busy. 05:19:12 Though browsing has become so horribly complex. Yes, we could write a browser in Forth, but we'd still be implementing a mountain of complexity if it was really going to function broadly on today's web. 05:19:36 Apparently gitter has an irc bridge. Curious if it's any good. I like the concept of gitter and the things it integrates with. Downside is it seems focused on Javascript/node.js technology. 05:19:59 That thing crc posted about a few weeks ago was pretty interesting. The protocol meant to sit between gopher and other mainstream systems? 05:20:02 veltas: I'm working on improving the latency as part of my ongoing rewrite 05:20:03 I forget its name right offhand. 05:20:06 KipIngram: gemini 05:20:19 Yes, that. 05:20:35 using gopher as a chat protocol? that's interesting... 05:20:36 Other than the TLS requirement I really like gemini so far. 05:21:56 crc[work], is there some link/docs for this? 05:22:16 both gopher and gemini are easy to implement (apart from TLS, which I handle through external tooling) 05:22:27 https://gemini.circumlunar.space/docs/ 05:22:27 proteus-guy: gemini info at https://gemini.circumlunar.space/ 05:22:32 thanks will check it out. 05:23:24 I understand WHY the TLS. One of the weaknesses of the way the internet evolved was that security and privacy weren't integrated in as ground floor requirements. 05:24:01 When that all started with Arpanet and so on it was just a bunch of scientists learning how to share stuff with one another. There was a level of community trust that's long gone these days. 05:24:35 And technology was slow, so they didn't really have the spare bandwidth and processing power to implement sophisticated security. 05:24:45 Just getting it to work at all was an achievement. 05:25:56 You see the other end of that spectrum in the bitcoin blockchain - it's rather remarkable in its ability to create a shared authentic ledger. But wow - it's got a lot of overhead. 05:26:37 Too much for it to really scale. It's cute and it's fun, but it can never become a planetary currency. Doesn't scale right. 05:26:58 * veltas braces for impending Bitcoin argument 05:27:11 If we tried we'd wind up with the bulk of the planet's data storage capacity holding blockchain copies. 05:27:35 :-) 05:27:48 So someone would need to build/implement Gemini protocol chat client/server? Gopher was a pre-web content browser. Is a Gemini chat kinda like a web client chat ui? 05:27:56 Nah. I don't have an axe to grind. ^^ that's just what I think. I could be wrong. 05:28:30 They also talk about graphene substrate processors running at 500 GHz in the future - maybe processing power and storage will outrun even bitcoin. 05:28:50 It's dangerous to try to predict the future on things like that. 05:28:57 If it did it would break bitcoin as a concept 05:29:11 Well, true. 05:29:14 veltas, why would it break Bitcoin as a concept? 05:29:19 KipIngram: I understand why the TLS, it's more of a matter that it introduces dependencies (I'm not good enough w/crypto to roll my own yet) 05:29:22 It's core idea is to consume a majority of processing power. 05:29:46 If processing power became vastly greater, people could start computing their own false blockchains. 05:29:50 The proof of work idea would break. 05:30:07 proteus-guy: The truth is that really you can't break it, if bitcoin gets easier then the work to mine gets harder 05:30:12 No - people would incorporate the increased hashing power and raise the complexity of the block. 05:30:18 proteus-guy: I don't think anyone has used gemini or gopher in a chat system yet. it's something on my list of eventual projects. 05:30:30 Yes, it was designed to be able to consume as much processing power as you throw at it. 05:31:09 proteus-guy: The idea is to make it computationally impossible for someone to calculate a fake block-chain, because they'd have to re-calculate the whole chain after the point they tried to edit. 05:31:19 I think bitcoin's the most brilliant invention of the 21st century but yeah there's a million issues with it 05:31:23 crc[work], OIC - so Gemini as a chat system isn't a thing yet. Seems like it could be something to provide a low cost low complexity front end to some of these web chat systems like gitter, however. 05:31:23 And the system is designed so that no one can have enough processing power to do that. 05:31:52 yes 05:31:52 veltas: I tend to agree - the idea of a non-tamperable currency is pretty fantastic. 05:32:02 It's like using gold again. 05:32:06 It's digital gold. 05:32:07 KipIngram, correct - right now there's not enough CPU processing power known in the solar system to reverse more than a few blocks in BTC. 05:32:08 Block chain on its own is a fantastic idea 05:32:29 Now adopted as a confusing and misused buzzword but as a concept it's genius 05:32:56 It is. And yes, for a few years in there everyone wanted to attach "blockchain" to their ideas. 05:33:05 crc[work], is that your thinking? Use Gemini to connect to existing systems like a bridge/client or build an entire chat network using Gemini as the core protocol? 05:33:23 It's better than gold 05:33:36 agree 05:33:42 Yes, someone always could find some underground mountain of gold. 05:33:53 We just ASSUME that gold has the scarcity needed to be a currency. 05:33:58 KipIngram, or one in space... actually quite likely eventually. 05:34:00 Bitcoin guarantees it. 05:34:04 Mathematically. 05:34:18 proteus-guy: I'm not sure on this yet. it's not something I've done much work towards yet 05:34:29 And my understanding is that we at least think that bitcoin is quantum-proof. 05:34:40 No one is aware of a quantum algorithm to reverse SHA-256. 05:34:52 twice even! ;-) 05:35:17 But we don't really understand the boundaries of quantum capabilities in a sound theoretical way. We've just stumbled on a few interesting algorithms. 05:36:11 proteus-guy: I had waited until recently to start working with gemini to give the protocol time to stabilize. then the freenode issues began right when I was starting to work on things using gemini. 05:36:41 ACTION likes it that there's already a gemini community out there. 05:36:51 crc[work], ah - makes sense. Are there pre-built Gemini clients that you're aware of? Just need a server - or we gotta build it all? 05:39:22 proteus-guy, for graphical targets, https://gmi.skyjake.fi/lagrange/ is nice. https://gemini.circumlunar.space/clients.html has some terminal ones, but I haven't used those yet. (I'll roll my own terminal one in Retro soon) 05:39:22 Regarding the TLS requirement - my generic issue with modern encryption in general is that it's complex. Ideally you'd have technology that a majority of people can understand completely. So that lots and lots of people could roll it up themselves. But encryption is a deep topic, and most people don't have the math chops to understand it. So they have to trust others to give it to t hem. 05:39:51 And do we KNOW that these techniques haven't been cracked? We don't really - the institutions with the finances to crack them aren't going to tell us they've done it. 05:40:36 If you have to more or less have a PhD in math to "truly believe in" your encryption, that's not a happy place. 05:41:51 crc[work], I'd prefer a text-mode client myself. I'll check it out. 05:42:27 It's unavoidable with technology that people have to rely on experts and use unscientific 'filters' to figure things out 05:42:58 I know. There's always going to be a "cutting edge" of tha tnature. 05:42:59 Stuff's complicated. Even if we really reduced bloat you'd still have that problem 05:43:32 But for example, Forth reduces the complexity of "a full operating system" to a level that's much easier to understand. 05:43:39 It would be nice if a "Forth of encryption tech" could emerge. 05:44:01 I mean that in a loose way - a "method" of encryption that's sound but easy to understand and implement. 05:44:13 Ideally a high school graduate could completely understand it. 05:44:20 We just don't have that today. 05:44:49 A high school graduate could understand most complicated things, but not all of them, and not with the time that high school students devote to learning 05:44:52 As simple as possible but no simpler... crypto is pushing boundaries of number theory. That's gonna require a lot of computational work no matter waht. 05:45:07 It's not really pushing boundaries of number theory 05:45:24 I agree - most of it is really just fairly simple number theory. 05:45:28 Some of it might be, most algorithms people use aren't 05:45:29 Understanding it - yes. Designing and implementing new techniques that are harder to crack is rocket science. 05:45:35 But number theory just isn't something most people learn. 05:45:57 And nor should it be, people need to specialise 05:46:00 I'm talking about the design parts. Once one is designed typically it's not too difficult to understand. 05:46:02 Right - I just don't think encryption as it exists today is a "grass rootable" tech. 05:46:03 You can't get around that 05:46:25 One might have said that about compilers at one point in time. 05:46:29 But we have Forth. 05:46:33 :-) 05:46:42 Forth doesn't contradict anything about compilers though 05:46:48 Forth is an "easy way" to do something that can also be done in very hard ways. 05:47:00 No - it's just an easier way. 05:47:30 You *can* write a compiler in Forth without knowing any convention compiler theory. It won't do anything that compiler theory teaches. You can also write an optimizing forth compiler learning about compiler theory 05:47:36 conventional* 05:47:48 dinner's here folks... ttyl. 05:48:19 A lot of Forth's genius is trying to make people rethink what their requirements are 05:48:25 +1 05:48:28 Semiconductor tech is the same situation. We have these these mega-expensive foundries and so on. I wish we could evolve toward a world where kids in junior high are playing with silicon and implementing things. 05:48:44 I'd like to see all these technologies NOT gated and controlled by big corporations. 05:49:30 Kids used to play with chemistry in quite advanced ways sometimes. You could get all the stuff to do it. These days you can hardly buy a really "competent" chemistry set. 05:49:34 We've evolved in the wrong direction. 05:50:09 We've pushed things toward regulation and control and economic centralization, instead of toward grass roots independence. 05:51:05 Because yeah, some kid might mix the wrong chemicals and blow up his house. Ooops - can't let people buy that stuff... 05:51:52 recently, you can't even buy battery acid anymore 05:52:08 When our response is to outlaw more things every time an accident happens... it's not the right path. 05:54:08 --- quit: cbridge_ (Remote host closed the connection) 05:54:23 --- join: cbridge_ joined #forth 05:54:30 And his point was that the path to greatness never is. 05:54:44 I was so moved by the thing that I cut it out and laminated it. 05:54:50 It's around somewhere, but I couldn't put my hands on it right now. 05:55:08 It encapsulated what I think of as the right philosophy of cultural advancement. 05:55:16 It's NOT the way we're doing it these days. 05:55:48 He talked about "whole ragged populations boiling westward." 05:56:00 Without thinks like feasibility studies and safety assessments. 05:56:14 And he closed by saying that those people understood that to live is to dare. 05:56:36 without "thins" 05:56:38 :-| 05:56:40 THINGS 06:00:40 there's a good book about evolution of liquid rocket fuels, it is called Ignition 06:01:19 many times author notes (being also rocket fuel scientist himself), how it is possible that this and that guy managed to do his research without blasting his hands off 06:04:02 Ah, I remember what the editorial was called. 06:04:08 "No Glory for Shrinking Violets." 06:04:35 It had a cartoon picture of a guy hiding inside a box with a security blanket and sucking thumb. 06:18:04 ventyl: That does sound like it's probably a fun book. 06:18:19 Have you read Longitude? 06:18:37 It's basically a history of navigation, with special emphasis on the longitude problem and the history of solving it. 06:18:47 KipIngram: it definitely is. even for someone who doesn't know organic chemistry 06:19:00 nope, I have come accross this book by an accident 06:19:06 My dad was an organic chemist. 06:19:10 Taught at university. 06:20:38 They eventually "solved" the longitude problem by developing seaworthy timepieces, that would keep accurate enough time. But there were some clever ideas that didn't win, too. 06:20:45 it is interesting, that in rocket science, typical american way of naming things is almost an oficial nomenclature 06:20:55 KipIngram: Re. "evolved in the wrong direction", broadly +1; though perhaps more along the lines of Poul Anderson's "Snowball" ( http://gutenberg.org/files/59314/59314-h/59314-h.htm ) and "No Truce With Kings" (with a side dish of good old Cliff Simak, of course. Also, my personal opinion is that "gated and controlled" is precisely where GNU GPL is ought to give kids some leverage over Big Money. 06:20:56 like UMDH-RFNA combo 06:22:28 Well, the typical evolution pattern in industries is that some large companies appear, then we decide that the arena needs to be regulated, and lo, the regulations wind up "entrenching" the existing large players. 06:22:46 A primary effect of the regulations is usually to blockade small entity participation. 06:23:03 Often just by making it too expensive to surmount the legal hurdles. 06:24:03 Down here in Texas, for example, the big beer distributors use regulation to limit competition from craft microbrewers. 06:25:10 And of course the craft beers are worlds better. 06:25:34 But the market is dominated by swill. 06:25:47 oh wow ##linux got destroyed by lee 06:25:51 oops, wrong channel 06:26:06 nah 06:26:22 is there a channel discussing leenode decline? 06:26:44 used to be the observer channels, probably see #libera now 06:27:03 libera was full of pls cloak me spam 06:27:40 What exactly is a cloak anyway? In IRC context? 06:27:49 KipIngram: hides your IP address 06:28:02 Why would I want to do that? 06:28:07 from /whois commands at least, AFAIK 06:28:40 I mean, my nick is MY NAME. I don't have anything to say I'm not willing to say openly. 06:28:45 Well right now I can see your IP address 06:28:52 And? 06:28:57 How does that hurt me? 06:29:08 so standard risks apply, if you had open ports or a server running 06:29:12 or geographic location 06:29:21 Oh, you don't see my machine's IP address. 06:29:28 You see my router's IP address. 06:29:30 I understand, that it was kind of feature for ops of large channels to limit harassment & stuff 06:30:04 In fact, you don't see that - you see the IP of a virtual machine I use for the sole purpose of IRC connectivity. 06:30:18 I run weechat curses on it, and I use mosh to ssh into that. 06:30:28 But ok, I see your point. 06:30:35 And yes - that makes sense. 06:30:38 great then, haha 06:30:49 yeah it's internet hygiene to not expose your external IP, I suppose 06:31:12 IIRC in some places IP address corresponds closely to physical location 06:31:15 Cloaks are also sometimes used to show one's affiliation. 06:31:21 Yeah. 06:33:24 So out of curiosity, what IP is showing for me, siraben? 06:33:42 ~kipingram@62.151.183.147 06:33:58 Ok, yes - that is the IP of my virtual server. 06:34:05 Where weechat is running. 06:35:14 it's in Spain? 06:35:51 ventyl: ##freenode is the channel you want 06:35:58 open ports are 22, 80, 443 and 8443 06:36:05 That's what it looks like isn't it? 06:36:10 I just looked that up too. :-) 06:36:21 I'm actually in the Houston Texas region. 06:36:56 The IP that I get from whatsmyip.com shows Houston, but isn't really localized to my own address very well. 06:37:03 Probably 15-20 miles away. 06:37:28 that tends to be a function of connecting google products to your network 06:37:49 Interesting. I'm running Chrome. 06:37:50 huh is there ##freenode on libera? 06:37:57 for me, google also shows locations in ~100km circle 06:38:04 based on my IP 06:38:18 but I don't let any smartphone to connect to my router 06:38:25 I actually consider that a plus. I'd rather not have it be too easy for someone to locate my house. 06:38:51 Any old joe, I mean. 06:39:00 androids are able to use navigation based on wifi AP MAC addresses 06:39:00