00:00:00 --- log: started retro/09.02.01
01:58:39 --- join: crc (n=charlesc@c-68-80-139-0.hsd1.pa.comcast.net) joined #retro
02:59:23 --- quit: crc ()
03:50:04 --- quit: nighty^ ("Disappears in a puff of smoke")
03:59:52 --- join: nighty^ (n=nighty@x122091.ppp.asahi-net.or.jp) joined #retro
07:32:44 --- join: crc (n=charlesc@c-68-80-139-0.hsd1.pa.comcast.net) joined #retro
08:48:25 --- part: nighty^ left #retro
12:43:09 --- join: nigol (n=nigol@gprs4.vodafone.cz) joined #retro
12:43:37 --- quit: nigol (Client Quit)
15:38:18 --- join: nighty__ (n=nighty@210.188.173.245) joined #retro
16:53:41 --- join: dpeschel (n=dpeschel@63-231-49-187.tukw.qwest.net) joined #retro
16:53:58 <dpeschel> Hello crc
17:51:00 <crc> hi dpeschel 
18:06:42 --- join: dpeschel_ (n=dpeschel@65-101-140-190.tukw.qwest.net) joined #retro
18:09:42 --- quit: dpeschel (Read error: 60 (Operation timed out))
18:31:21 <dpeschel_> n
18:45:03 <crc> n?
18:54:39 <dpeschel_> oops, a typo
18:54:41 <dpeschel_> glad you're here
18:55:19 <dpeschel_> (well, were here when you typed that...)
19:04:00 <dpeschel_> did you get my e-mail?
19:04:41 <crc> I recieved one message yesterday; I replied to it a while ago
19:11:28 <dpeschel_> oh, interesting...
19:11:33 <dpeschel_> I haven't checked the Google group lately
19:11:41 <dpeschel_> let me check my e-mail again
19:11:54 <crc> I get the google group messages via email
19:13:16 <dpeschel_> good to know
19:13:29 <dpeschel_> now I see you referred to a personal messaage to me...
19:13:32 <dpeschel_> which I finally read just now
19:13:59 <dpeschel_> and I need to try your patch
19:14:40 <dpeschel_> though I wanted to talk to ask about larger issues, namely...
19:14:59 <dpeschel_> deciding on a more detailed spec for how the VMs should do I/O
19:15:05 <dpeschel_> and making them all work the same
19:15:12 <dpeschel_> if that's important to you
19:23:30 <crc> more detailed specs would be good
19:25:17 <dpeschel_> OK, but how big of a problem is backward compatibility?
19:25:28 <dpeschel_> or having some way for Retro programs to detect VM capabilities w/ version #s?
19:25:46 <dpeschel_> it depends how many Retro programs there are
19:26:05 <crc> there's not many as far as I know
19:28:03 <crc> I have a moderate amount, but nothing I can't rebuild.
19:28:20 <dpeschel_> does anyone else have any?
19:28:32 <crc> not that I'm aware of
19:28:50 <crc> Mat may have some, but he's also working with an extended vm implementation already
19:29:54 <dpeschel_> OK, that makes things much easier
19:31:16 <dpeschel_> now do you favor using ports or emitting characters as a way to change colors?
19:31:25 <crc> ports
19:32:21 <dpeschel_> are strings made up of 8-bit or 32-bit characters?
19:32:29 <crc> 32-bit characters
19:32:44 <dpeschel_> but do any VMs really allow Unicode? maybe in the Java or JavaScript ones
19:32:45 <crc> the current vm only allows for 32-bit memory cells
19:32:52 <crc> at present, no
19:33:06 <dpeschel_> what happens if you type Unicode into the Java VM?
19:33:10 <crc> No idea
19:33:44 <dpeschel_> all right, then I'll concentrate on a design with only 256 chars.
19:34:08 <dpeschel_> of course Unicode has ISO-8859-1 as a subset, but the encoding is not the same
19:34:19 <dpeschel_> but I mean I'll just work with some variant of ISO-8859-1 directly
19:35:04 <dpeschel_> just a sec
19:36:48 <dpeschel_> have you thought about adding some way for programs to detect VM changes?
19:37:23 <crc> not yet
19:37:29 <dpeschel_> like more capabilities you can get through port 5
19:37:36 <crc> I'd planned to use the capabilities port to check for future additions
19:37:44 <dpeschel_> hmm, interesting
19:38:03 <dpeschel_> but the more changes are made, and the more programs are written, the less it helps to add detection
19:38:06 <crc> but since I haven't built an extended vm yet, the core image doesn't check anything other than presence of framebuffer and memory size
19:39:24 <dpeschel_> does SDL do alpha blending?  I want to change the text color...
19:39:32 <dpeschel_> but I noticed that the font is antialiased
19:39:46 <dpeschel_> so you can't just change white to another color
19:43:47 <crc> I don't believe so
19:45:35 <crc> actually, there is an SDL_SetAlpha function (http://www.libsdl.org/cgi/docwiki.cgi/SDL_SetAlpha), but I haven't used it
19:46:58 <dpeschel_> do you have an "aliased" (two-color) version of the font?
19:47:01 <dpeschel_> also more characters
19:47:09 <dpeschel_> 160-255 for example
19:47:11 <crc> nope
19:47:26 <crc> I seldom use the framebuffer vm now
19:48:00 <dpeschel_> you use the curses one then?
19:48:25 <crc> that, and the javascript implementation
19:51:53 <dpeschel_> I like the idea of seeing results after you hit space,
19:52:00 <dpeschel_> since it matches the Retro execution mechanism
19:52:19 <dpeschel_> but it means you can't paste a series of definitions separated by return
19:52:36 <dpeschel_> and I'm also not sure whether "ok" or any other messages should insert return
19:52:46 <dpeschel_> have you thought about those issues?
19:56:17 <crc> return and/or tab can be handled by altering (remap-keys), which is why I added that in the first place
19:56:47 <crc> as for ok and other messages, they are bits I've grown used to from my past work on retro
19:57:39 <crc> the 'ok' can be disabled easily enough:  here ] ; is ok
19:58:13 <dpeschel_> interesting points, but I was asking about the UI design
19:58:28 <dpeschel_> that's ideal, not how to implement various designs
19:58:51 <dpeschel_> wouldn't (remap-keys) remap every press of return/tab even if it was in a string?
19:58:58 <crc> yes
19:59:01 <dpeschel_> I thought of using (remap-keys) but decided to write word-accept instead
20:00:26 <dpeschel_> getting to another question on my list...
20:00:38 <dpeschel_> can I write a macro in toka  color: x
20:00:54 <dpeschel_> that transforms into  char: x # color
20:01:19 <dpeschel_> maybe by writing the macro in retro.forth after color is defined?
20:02:18 <dpeschel_> on the other hand, I don't need the letter format for colors once real colors are working
20:02:46 <crc> that should be doable:
20:03:02 <crc> [ ` char: invoke # color ] is color:
20:03:05 <crc> or something similar
20:04:02 <dpeschel_> interesting, I had tried
20:04:30 <dpeschel_> expanding the definition of char:, like this:
20:04:39 <dpeschel_> [ wsparse c@ >char compiler @ [ # newcolor # !, ] ifTrue ] is-macro color:
20:04:41 <dpeschel_> but it won't assemble
20:04:44 <dpeschel_> in fact it crashes toka
20:11:05 <dpeschel_> hold on...
20:12:24 <crc> [ 1 , ` char: invoke , ` color invoke ] is color:
20:12:36 <crc> assuming that 'color' is a retro word
20:18:53 <dpeschel_> does this mean your first version won't work?
20:19:33 <dpeschel_> hold on again...
20:19:53 <crc> [ ` char: invoke # ` color invoke ] is color:
20:20:15 <crc> you'd have to use ` and invoke on the forth word, but it work if you did this
20:21:00 <crc> actually I'm wrong... 
20:21:07 <crc> the first example would work
20:21:24 * crc hasn't done work on extending the cross-compiler in quite a while :(
20:24:39 <dpeschel_> and is the only one who really knows how it works
20:26:45 <dpeschel_> about UIs, I downloaded Retro 6.8 for Windows and played with the editor
20:27:37 <dpeschel_> and saw some similarities with 10 and some differences
20:27:50 <dpeschel_> maybe some features disappearing as Retro is developed, as well as appearing
20:28:20 <dpeschel_> but has it had a much fancier UI in the past, or just the interpreter loop and the editor?
20:34:07 <dpeschel_> with the interpreter loop being a pretty bare-bones typewriter interface
20:40:10 <crc> the interpreter loop has always been fairly minimal
20:40:45 <crc> each version tends to change things a lot
20:42:10 <crc> I don't recall 6.8 ever running on windows
20:43:42 <dpeschel_> you're right, it was a different version that I downloaded
20:43:47 <dpeschel_> maybe 8.2.5
20:44:09 <dpeschel_> I got 6.3 (not 6.8) to see if it contained the manual you were writing
20:45:11 <dpeschel_> I mean the internals manual, which I see references to but no copies of
20:46:32 <crc> perhaps this? http://retroforth.net/doc/RxBook.pdf
20:46:48 <crc> it was the documentation for the core of retro 9.x
20:49:13 <crc> it doesn't apply to the 10.x code though
20:53:05 <dpeschel_> http://retroforth.net/board/index.php?topic=53.0
20:53:11 <dpeschel_> is what I was chasing
20:53:22 <dpeschel_> probably even more out of date than yours
20:54:49 <dpeschel_> now I remember... I wanted to see if there were any SDL-specific words in the past
20:56:47 <crc> none that I recall
20:57:26 <crc> some early 6.x releases had a framebuffer using svgalib and/or sdl, but there was only minimal words provided for this
20:57:43 <dpeschel_> OK, I won't worry about it
20:57:43 <crc> I *think* that there was a clipping function, to limit the drawing area
20:57:50 <crc> but nothing significant
20:58:01 <dpeschel_> yeah, it sounds like that's the case
20:58:11 <dpeschel_> I forgot to ask what OS you run the curses VM on
20:58:44 <crc> OS X mostly, also Linux (my server boxes), and Windows (at work)
21:00:02 <dpeschel_> because on OS X, curses lets the real TTY driver interpret a few control characters
21:00:05 <dpeschel_> like ^V for literal next
21:00:32 <dpeschel_> and I believe the incantations to fix that are not portable
21:08:18 * crc needs to get some sleep; 5 hours till I have to get ready for work....
21:08:41 <crc> it's been good talking with you; I'll try to be back here tomorrow evening
21:14:04 <dpeschel_> thanks for your time, sleep well, I'll do my best to be on tomorrow
21:26:05 --- quit: dpeschel_ ("leaving")
22:07:58 --- quit: crc ()
23:59:59 --- log: ended retro/09.02.01