00:00:00 --- log: started forth/08.02.24
01:59:16 --- join: forther (n=forther@c-67-180-150-67.hsd1.ca.comcast.net) joined #forth
02:00:10 --- quit: proteusguy (Remote closed the connection)
02:00:10 --- join: proteusguy (n=proteusg@ppp-124-120-222-20.revip2.asianet.co.th) joined #forth
02:00:10 --- quit: forther (Read error: 110 (Connection timed out))
02:45:39 --- quit: sauvin (Remote closed the connection)
03:29:26 --- quit: Raystm2 ("Should have paid the bill.")
06:12:53 --- quit: nighty^ ("Disappears in a puff of smoke")
06:13:28 --- join: forther (n=forther@c-67-180-150-67.hsd1.ca.comcast.net) joined #forth
06:38:35 <Bushmills> tackling poor man's defining words in efte macros without having anything like "create" or facilities to build any such ... :)
06:39:36 <Bushmills> sub variable { here cell allot }    ;   sub foo { variable new }      ... would you consider that tolerable syntax?
06:40:03 <Bushmills> { foo fetch  ... n foo store } to use var then ...
06:56:03 --- quit: forther (Read error: 110 (Connection timed out))
08:44:56 <Quartus> Bushmills, what is there in      sub variable { here cell allot }   that defines a new word?
08:46:04 <Quartus> Oh, does { variable new } send a 'new' message to 'variable' or something?  In that case, wouldn't each invocation of 'foo' allot a new cell?
08:46:35 <Bushmills> not really "send a message" ; hi quartus. 
08:46:46 <Bushmills> "new" just executes ...
08:46:57 <Bushmills> and patches up the macro (defined with "sub")
08:47:36 <Bushmills> replaces "variable" against "push_pointer_to_data", and  "new" against  "variable" for later reference, if ever needed.
08:47:58 <Quartus> Gah. Ok.  It looks confusing.
08:48:15 <Quartus> What is this weird setup?
08:48:18 <Bushmills> so execution of "new" would also be the "not yet initialized" detection
08:48:59 <Bushmills> that's latching on a sequential command executor, to make that one forthish
08:49:16 <Bushmills> don't have shared name space, for headers.
08:49:44 <Bushmills> in fact, until shortly configuration compilation/parsing was a separate process
08:49:52 <Quartus> 'forthish' for some very small value of 'ish'
08:50:27 <Bushmills> other forth components map better than defining words
08:50:48 <Bushmills> defining words is one point were the underlying model is a bit resistant
08:51:01 <Bushmills> so I need to work around it at this point
08:51:15 <Bushmills> until we made the macro executor map forth better
08:51:28 <Bushmills> for example, I don't have access to return stack.
08:51:50 <Bushmills> my macro "return" stack is purely a loop and  temp >r r> stack
08:52:05 <Bushmills> but I need some variables now
08:52:14 <Bushmills> base, dp, such things
08:52:31 <Bushmills> therefore I'm willing to compromise a bit here
08:54:49 <Quartus> Well, as for the above, it looks to me as though each invocation of foo will reinvoke variable, which will allot a new cell.
08:55:43 <Bushmills> idea is that "new" patches foo, by overwriting "variable" with token of command "push the address supplied by "variable" to stack.
08:55:57 <Bushmills> thus, on first execution
08:56:00 <Quartus> 'new' strikes me as badly named, then.
08:56:22 <Bushmills> after that, "variable" will not be at that slot in macro anymore, and "new" not at all anymore
08:56:53 <Bushmills> thought of calling it "create" first .. but as the function is so different, it didn't seem to be a good choice
08:57:05 <Quartus> 'redefine' perhaps
08:57:58 <Quartus> or 'repurpose'
08:58:06 <Quartus> or even 'patch'
08:58:40 <Bushmills> yes, that is what it is *doing*. the  "inner working"
08:58:55 <Bushmills> I didn't think the name has to reflect "inner working"
08:59:02 <Quartus> Point being, it is doing nothing that involves the word 'new'.
08:59:02 <Bushmills> but more what it is used for
08:59:09 <Bushmills> in that case, "new variable"
08:59:18 <Bushmills> or, " variable new"
08:59:54 <Quartus> Yes, and 'variable new' says to me 'make a new variable'.  
09:00:01 <Bushmills> which is what it does
09:00:05 <Quartus> On each invocation.
09:00:59 <Quartus> What it sounds like you're trying to convey is that 'variable new' redefines foo, and only once.
09:02:20 <Quartus> From this it seems as though 'variable' and 'new' have some kind of status akin to a Forth IMMEDIATE word.
09:04:11 <Quartus> ... yet the syntax doesn't flag them as such, or does it?  Or are all words immediate, in this setup?
10:32:30 --- join: edrx (i=edrx@189.25.90.199) joined #forth