00:00:00 --- log: started retro/10.06.05
06:04:35 --- join: docl (~luke@216-161-87-126.ptld.qwest.net) joined #retro
06:05:24 <docl> good morning
06:38:23 <crcx> hi docl 
06:50:23 --- join: sixforty (~sixforty@pdpc/supporter/active/sixforty) joined #retro
06:54:17 <docl> hi crc, hi sixforty 
06:54:29 <sixforty> hi, docl
08:28:56 --- quit: sixforty (Ping timeout: 258 seconds)
09:00:42 <docl> crc: I like the way rxe.extensions.retro is put together.
09:29:28 <crcx> thanks
09:30:44 <crcx> I moved $$b and $$B there after experiencing a few crashes; I think they are overwriting code before/after the block buffer
09:33:34 <docl> hmm
09:34:21 <docl> toBlank hunts for a blank line
09:36:27 <crcx> if I edit in block 127, and try to shift down past the last line, it segfaults
09:37:59 <docl> putting bounds in toBlank seems to stop it from segfaulting but it then goes into an infinite loop
09:44:31 <docl> maybe because blk is shifted back by one, so it repeats the search over and over.
09:46:28 <crcx> I pushed a possible fix
09:46:56 <crcx> not sure if it breaks anything, but seems to keep things from crashing
09:57:54 <docl> it helps. I can still get it to crash by going one line above the text and pushing it down.
10:27:29 <crc> ok, home now so I can work a bit faster :)
10:29:29 <docl> :)
10:30:22 * docl is working on ideas for making a more flexible database
10:31:41 <docl> data types could be stored in a vocab or object.
10:33:32 <docl> they would be words like classes are. sub-classes of .data
10:45:02 * crc wonders if it'd be better to just have a swap-lines key instead
10:46:27 <docl> might be
10:47:18 <docl> maybe a seek-blank key too?
10:47:56 <crc> that could work
10:51:40 <docl> I just redefined $$b to seek blank and ran it at the end of the file. this causes segfault.
10:53:53 <crc> try this one: http://retroforth.com/paste/?id=3425
10:54:08 <crc> $$b to swap the current line with the following one; $$B to seek a blank line
10:56:05 <docl> cool
10:56:27 <crc> does this work ok?
10:57:50 <docl> yes, it works good
10:57:57 <crc> almost :(
10:58:09 * crc got it to crash with $$B in a full block
10:59:05 <docl> hmm
10:59:20 <docl> yeah I had that happen once just now, can't seem to replicate
10:59:56 <crc> only seems to happen in the last block
11:00:07 <crc> this seems to fix it:
11:00:08 <crc>   : seekBlank  ( -  )
11:00:09 <crc>     repeat line ++ guard isBlank? if; again ;
11:03:52 <docl> I can't seem to replicate the issue :(
11:05:56 <docl> I had it happen again, the first block had some text in it and the last block had text on the last line (offset by some). I hit B right above the last line and it exited
11:06:15 <crc> with my update?
11:07:31 <docl> no, let me try that
11:07:47 <docl> I seem to be getting better at replicating the issue although it is sporadic
11:09:08 <docl> ah, got the same result with the update.
11:09:29 <docl> created line at end of buffer, then another line above it. then hit B from the line above that.
11:10:53 <docl> from the last line do <tab>test<tab>i<tab>test<tab>iB
11:12:02 <docl> this is not a segfault though. stack element missing?
11:14:27 <docl> could it have to do with 64 chars in a row being blank?
11:16:31 <docl> from the first block, <tab>test<tab>k<tab>test<tab>iB
11:23:12 <crc> docl: isBlank? should only be looking at whole lines, not partials, so that's not the cause
11:23:34 <crc> though I do see a problem with it now (no rdrop to get rid of the loop index)
11:24:39 <crc>   : guard @line 7 >if 7 !line rdrop then ;
11:24:39 <crc>   : isBlank?  ( -f ) @line (line) 64 for @+ 32 !if pop 2drop 0 ;then next drop -1 ;
11:24:42 <crc>   : seekBlank ( -  ) repeat line ++ guard isBlank? if; again ;
11:24:49 <crc> this appears to fix both of your test cases
11:33:19 <docl> looks good :)
12:12:46 <crc> ok, pushed
15:44:10 * docl just pushed the database I've been working on
16:29:34 <crc> looks good
16:29:44 <crc> you can use ' prefix instead of char: now
16:45:11 <docl> cool!
16:48:52 <docl> pushed :)
16:50:19 <crc> pulled :)
23:35:12 --- quit: docl (Ping timeout: 264 seconds)
23:59:59 --- log: ended retro/10.06.05