2024-07-23 09:49:05 <pkal        > who is responsible for https://forth-standard.org?
2024-07-23 09:49:29 <pkal        > as in individuals one could message directly?
2024-07-23 10:31:50 <veltas      > I don't think anyone here is
2024-07-23 10:35:10 <olle        > is what?
2024-07-23 10:37:20 <pkal        > responsible for forth-standard.org, I presume
2024-07-23 10:38:02 <veltas      > olle: http://forth.chat/logs/libera/forth/2024-07-23
2024-07-23 10:41:02 <olle        > Oh ok
2024-07-23 10:41:15 <olle        > Should I use Matrix instead of irssi to preserve IRC history.....?
2024-07-23 10:41:27 <olle        > Don't wanna leave my computer on all the time :d
2024-07-23 10:41:45 <veltas      > Just use the logs
2024-07-23 10:41:46 <pkal        > a bouncer might help
2024-07-23 10:41:54 <veltas      > http://forth.chat/logs/libera/forth/
2024-07-23 10:42:29 <veltas      > If you have a server just run irssi or some IRC client in tmux or screen
2024-07-23 10:42:33 <veltas      > That's what I do
2024-07-23 10:42:57 <olle        > veltas: logs for all channels? mm
2024-07-23 10:43:12 <olle        > I do have a vts
2024-07-23 10:43:44 <veltas      > A lot of channels have public logs
2024-07-23 10:43:49 <pkal        > (one day libera might implement the IRCv3 channel-history extension and allow you to query the last n messages/messages written in the last t minutes)
2024-07-23 10:44:16 <veltas      > I sort of doubt it, and if they do it will probably be configurable in chanserv
2024-07-23 10:45:11 <pkal        > yeah, I don't belive they would enable it by default either
2024-07-23 10:45:34 <veltas      > I mean channels would be able to opt out, and would have to opt-in if they wanted it
2024-07-23 10:45:47 <veltas      > But I think the legal consequences of logging would put libera.chat off
2024-07-23 10:46:10 <olle        > Right
2024-07-23 10:46:38 <pkal        > true, I haven't considered the legal point.  perhaps only storing the last t minutes might make that easier to moderate
2024-07-23 10:47:28 <pkal        > just enough to give people like olle some context when joining
2024-07-23 10:48:20 <veltas      > Well I hope olle knows now there's a public log of this channel!
2024-07-23 10:48:44 <veltas      > crc might consider putting it in the topic more explicitly although probably nobody will read it
2024-07-23 10:49:37 <pkal        > fwiw I was surprised there was a public log as well.
2024-07-23 12:11:01 <crc         > I've added a mention of the logs to the topic and channel entry message
2024-07-23 12:27:26 <olle        > Nive
2024-07-23 12:27:29 <olle        > Nice*
2024-07-23 21:53:51 <olle        > Can `if` be defined as a word based on other words?
2024-07-23 21:57:43 <darkling    > You could do it with CASE, I guess. I don't think you could do it if you didn't use _something_ that didn't do some kind of conditional exexcution.
2024-07-23 21:58:03 <dlowe       > it's often defined in terms of a word like branch0 and branch1 which sets program execution to a particular point in memory
2024-07-23 21:58:11 <dlowe       > but these are usually implementation dependent
2024-07-23 21:58:12 <darkling    > So at some point, there's got to be a branch in the implementation.
2024-07-23 21:58:25 <GeDaMo      > Jump table
2024-07-23 21:58:40 <dlowe       > the IF word handle computing the memory locations for the branch instructions
2024-07-23 22:00:10 <olle        > Yea makes sense
2024-07-23 22:03:12 <KipIngram   > I think I've seen 0branch more often than the alternatives.
2024-07-23 22:03:46 <KipIngram   > Often that will be the only primitive conditional operation you have (not counting machine code conditional sections inside primitives).
2024-07-23 22:08:29 <pkal        > apparently leave doesn't work in begin ... again?  is there some other way to break out of a infinite loop?
2024-07-23 22:17:42 <dlowe       > divide by zero :D
2024-07-23 22:18:55 <dlowe       > I think leave is supposed to work with LOOP, not AGAIN
2024-07-23 22:18:58 <dlowe       > could be wrong
2024-07-23 22:30:54 <olle        > Looping is another primitive thing :d
2024-07-23 22:54:06 <veltas      > Branchless BRANCH ... write colon word, get condition into 1 or 0 depending on true or false using arithmetic/bitwise/0<> or whatever's available, and add (0|1)*offset to R@ before exiting the word
2024-07-23 22:54:41 <veltas      > But I can't think of a legitimate reason for not just implementing it in machine code, given that it's probably easier than that
2024-07-23 22:57:41 <pkal        > dlowe: from my testing that is the case
2024-07-23 22:59:42 <veltas      > For example, if 0= is allowed, something like:  : BRANCH   0=  R@ @ AND  R> CELL+ + >R ;
2024-07-23 23:02:01 <zelgomer    > 0= is going to return 0 or -1, so you would need a "1 and" in there or something
2024-07-23 23:02:29 <zelgomer    > or negate i guess
2024-07-23 23:02:53 <veltas      > If it's -1 then I AND and get the offset
2024-07-23 23:03:00 <veltas      > If it's 0 I AND and get 0
2024-07-23 23:03:31 <veltas      > -1 is actually more useful than 1, I found when writing it in IRC
2024-07-23 23:03:41 <zelgomer    > yeah
2024-07-23 23:03:56 <zelgomer    > : branch  0= cell and r@ + r! ;
2024-07-23 23:04:35 <zelgomer    > well, no. that doesn't jump over the parameter in the 0 case. whatever, forth is hard.
2024-07-23 23:04:44 <veltas      > Agreed
2024-07-23 23:05:07 <veltas      > Did you see my code?
2024-07-23 23:05:26 <zelgomer    > i saw this   : BRANCH   0=  R@ @ AND  R> CELL+ + >R ;
2024-07-23 23:06:13 <zelgomer    > and now i realize i misunderstood it
2024-07-23 23:06:38 <zelgomer    > because i forgot r@ is going to give you the address of the offset, not the offset. sorry, i've been away from forth for a few weeks.
2024-07-23 23:29:18 <veltas      > I don't blame you, I'm currently mucking around with Lua
2024-07-23 23:29:23 <veltas      > On a Lua holiday