2023-04-21 06:23:21 <veltas      > KipIngram: To be fair, I am just now finding out Windows' ECHO command actually captures the exact command-line, leaving double spacing between 'arguments' and everything
2023-04-21 06:23:29 <veltas      > While expanding parameters
2023-04-21 06:23:59 <veltas      > But IMO this is worse, because none of this s*** is written down anywhere, the best docs you can find are for Windows XP, and it's inconsistent
2023-04-21 07:01:14 <KipIngram   > Ugh.  I've done nothing with Windows but run from it for over a decade now.
2023-04-21 07:02:11 <KipIngram   > Part of me wishes echo would give you full control over the output, but on the other hand not everything is meant to be a total Swiss army knife.
2023-04-21 07:02:18 <KipIngram   > That just may not be its mission.
2023-04-21 07:03:03 <KipIngram   > As was pointed out to me in the earlier discussion, there are other tools that do offer that - fair enough.
2023-04-21 07:03:23 <KipIngram   > I was just being lazy, trying to use the thing I was most familiar with.
2023-04-21 07:03:48 <KipIngram   > It's handy when life works out well with that approach, but I guess there are no promises about it.
2023-04-21 07:13:34 <veltas      > Just needed to read a .bat file at work
2023-04-21 07:14:01 <veltas      > Today I learned cmd.exe FOR can parse a file into fields like awk
2023-04-21 07:38:03 <veltas      > There's a watch called the Freak by Ulysse Nardin, which uses the entire movement as the minute hand
2023-04-21 07:38:10 <veltas      > Go look it up, it's sort of interesting
2023-04-21 07:38:30 <veltas      > What struck me though is how that's actually quite Forthy, to refactor the design to that extent
2023-04-21 07:49:23 <KipIngram   > I cracked open the first Spenser novel a day or so ago.  The godwulf Manuscript.  It's actually pretty good "noir style" stuff.
2023-04-21 07:49:35 <KipIngram   > Spenser's a good character, at least at the start of the series.
2023-04-21 07:50:02 <KipIngram   > I recall from my first pass through these that he gets a little hard to bear about seven books in, but it's ok up to there.
2023-04-21 07:51:01 <KipIngram   > He's one of these characters that has "a code," and there around book seven it just started to feel a little too in my face high and mighty.
2023-04-21 07:51:17 <KipIngram   > Like he'd read too many of his own press clippings.
2023-04-21 08:22:36 <veltas      > Ugh I'm enjoying blocks too much
2023-04-21 08:23:16 <crc         > is that possible? :)
2023-04-21 08:23:43 <veltas      > I don't want to like them but it's quite awesome how you can do so much without any filesystem
2023-04-21 08:23:50 <crc         > 
ACTION is working on a block editor for larger 64x64 blocks

2023-04-21 08:23:51 <veltas      > It's too forthy
2023-04-21 08:24:13 <veltas      > Honestly crc I think the main trick is supporting line endings, rather than padding all lines
2023-04-21 08:24:29 <veltas      > Much more efficient use of space, but code's a little more complicated
2023-04-21 08:25:14 <veltas      > But I know that's a highly unconventional take
2023-04-21 08:39:41 <veltas      > I can probably break someone's mind in here by pointing out that IRCv3 has an extension to support sending/receiving 1024 byte messages rather than the normal limit of 512 bytes
2023-04-21 08:40:12 <veltas      > So that may or may not allow a forth to harness buffers for that work in a slightly elegant manner?
2023-04-21 08:41:22 <crc         > an editor that handles line endings is something on my todo list, but won't be started for a while
2023-04-21 08:41:59 <Zarutian_iPad> veltas: hmm? why does that break somones mind?
2023-04-21 08:42:11 <veltas      > In the sense of nerd sniping
2023-04-21 08:42:38 <veltas      > Someone may think "oh okay that works" and go write an IRC client
2023-04-21 08:43:48 <Zarutian_iPad> the 512 byte limit was basically choosen arbitarly to prevent huge messages to go onto IRC networks
2023-04-21 08:44:26 <veltas      > Yeah but 1024 is the size of a forth block, that's all I'm pointing out really
2023-04-21 08:44:42 <veltas      > They can be other sizes but 1024 is the sort of 'conventional' size
2023-04-21 08:44:51 <Zarutian_iPad> the IRC clients I have written do not care as it just parses until a line ender is found
2023-04-21 08:45:01 <Zarutian_iPad> oh, right
2023-04-21 08:45:32 <veltas      > If I wrote an IRC client I'd probably start replying with errors if you sent something longer than expected
2023-04-21 08:45:36 <veltas      > Or ignoring
2023-04-21 08:45:53 <veltas      > Not sure what RFC says to do in that situation
2023-04-21 08:46:34 <Zarutian_iPad> not sure, but ircd implementations send back an error and then close the tcp connection
2023-04-21 08:46:46 <veltas      > Yeah that's probably correct
2023-04-21 08:47:52 <Zarutian_iPad> but clients? I just let the OS tcp buffering implementation handle the big buffer
2023-04-21 08:48:14 <veltas      > Sounds like a security risk, unless they have some kind of fixed size allowance
2023-04-21 08:48:47 <Zarutian_iPad> why? the tcp buffering should apply back pressure
2023-04-21 08:49:22 <veltas      > This is beyond my understanding
2023-04-21 08:49:54 <Zarutian_iPad> right, you have not read about how tcp works, yes?
2023-04-21 08:50:51 <veltas      > Not extensively
2023-04-21 08:51:09 <veltas      > It's not my expertise, I haven't done much networked stuff
2023-04-21 08:51:27 <veltas      > I know the low-level breakdown but not too much about the flow control
2023-04-21 08:52:26 <Zarutian_iPad> so, basically the window size sent back and it can go down to zero
2023-04-21 08:53:01 <veltas      > Makes sense
2023-04-21 09:02:23 <KipIngram   > veltas: I think it makes total sense to like them.
2023-04-21 09:03:17 <veltas      > I've read a little more about it, I do find networking stuff interesting I just haven't done it in the real world at all
2023-04-21 09:03:26 <KipIngram   > Re: newlines, I'm torn; I don't think I'm so sure that they're that important.  Yes, you'll be less storage efficient without them, but really so what?  It's not like you're going to use any large fraction of your storage space for source code.
2023-04-21 09:03:29 <veltas      > TCP etc all fascinating to me
2023-04-21 09:03:42 <veltas      > Don't claim to be an expert or anything close though
2023-04-21 09:03:46 <KipIngram   > So yes, you gain by supporting them, but it's a lot more complex and it's a very minor gain.
2023-04-21 09:03:55 <KipIngram   > That said, I am probably going to support them this time.
2023-04-21 09:04:17 <KipIngram   > I think the way I've written EXPECT will actually support them naturally."
2023-04-21 09:04:33 <veltas      > It just means a block goes from 16 lines to potentially many more lines
2023-04-21 09:04:44 <veltas      > Also KipIngram, crc: do you support tabs?
2023-04-21 09:04:47 <KipIngram   > The way I plan to do it is when I used EXPECT to edit a line of a block, it will actually be working with a buffer that goes all the way to the end of the block.
2023-04-21 09:05:20 <KipIngram   > But it will only print up to the first newline; the reason for editing the whole thing is so when I add and delete characters in that line being edited, it will move all the rest of the characters appropriately.
2023-04-21 09:06:45 <crc         > I do not support tabs in the interpreter
2023-04-21 09:07:26 <KipIngram   > I don't either; my EXPECT basically ignores everything below ASCII 32, except for escape and Enter.
2023-04-21 09:07:43 <KipIngram   > And those are both special case control keys.
2023-04-21 09:08:12 <KipIngram   > Well, that's not fully true.
2023-04-21 09:08:21 <veltas      > Tabs are another potential space saver
2023-04-21 09:08:28 <KipIngram   > I check for other things too; my editor commands are ctrl-<letter> keys.
2023-04-21 09:08:41 <KipIngram   > So are Chuck's compressed formats.
2023-04-21 09:08:57 <KipIngram   > So is just overtly supporting some kind of compression.
2023-04-21 09:09:28 <crc         > for end of token, I recognize space (32) and line feed (10)
2023-04-21 09:09:32 <KipIngram   > But like I said - I just don't see that it's a big payoff.
2023-04-21 09:09:56 <KipIngram   > We just all seem sensitized to it somehow.  It FEELS like waste.
2023-04-21 09:10:18 <KipIngram   > Well, it is waste.   But if FEELS more intolerable that it actually would be.
2023-04-21 09:11:23 <KipIngram   > It still might be worth implementing a block editor that used fixed length lines first; it's easier, and you'd have an editor to do further work with.
2023-04-21 09:11:30 <KipIngram   > Then use it to write your better editor.
2023-04-21 09:11:51 <KipIngram   > Or you could start with a super-simple line editor.
2023-04-21 09:12:49 <KipIngram   > Any of those would let you start developing in Forth instead of using whatever dev tool you're using.
2023-04-21 09:15:49 <crc         > I use a simple line editor most of the time. I have several others as well, of varying levels of complexity, but they aren't used much so far.
2023-04-21 09:17:03 <KipIngram   > Yeah, it's a good incremental improvement case.
2023-04-21 09:49:39 <KipIngram   > I liked the different editors they had (line and screen) in that doomsday Forth-based package.  I'm having trouble remembering its name right now.
2023-04-21 09:49:49 <KipIngram   > CollapseOS.
2023-04-21 09:50:16 <KipIngram   > You could edit either way, and I definitely felt like for small edits you might find yourself going routinely to the line editor.
2023-04-21 09:50:39 <KipIngram   > It was heavily search-based, which was an idea I hadn't thought of previously.
2023-04-21 09:51:20 <KipIngram   > That could be used in line or screen editors, and in a screen editor particularly it might wind up being faster navigation that cursor keys.
2023-04-21 09:51:54 <thrig       > I remove support for the arrow keys in vi
2023-04-21 09:54:17 <KipIngram   > Oh, I meant to include the ctrl- keys in that category too.
2023-04-21 09:54:21 <KipIngram   > Cursor "motion" keys.
2023-04-21 09:54:40 <KipIngram   > I don't ever use the actual cursor keys in vim, but I don't bother to turn them off.
2023-04-21 09:55:08 <KipIngram   > I mean, my fingers are right their over the ctrl-letter keys; it's just faster to use them, and doesn't pull my hands out of home typing position.
2023-04-21 09:55:26 <KipIngram   > I dislike having to use a mouse for the same reason.
2023-04-21 09:55:59 <KipIngram   > I keep telling myself I should invest the effort needed to get good at that little "mouse button" in the middle of the keyboard, but I never have, and it's not very friendly without practice.
2023-04-21 09:56:58 <thrig       > the trackpad I mostly use to raise xterms depending on where the gnat is
2023-04-21 10:11:29 <veltas      > Touchpad's faster IMO, trackpad just feels nicer to me
2023-04-21 10:12:24 <KipIngram   > Yeah, I use the pad.
2023-04-21 10:12:45 <KipIngram   > Most of the time I can reach down and get it with a thumb without too much hand motion.
2023-04-21 10:13:11 <KipIngram   > It's just easier to be precise with it than that track button.
2023-04-21 10:13:31 <KipIngram   > And usually mouse stuff needs to be fairly precise.
2023-04-21 10:16:09 <veltas      > When you get used to it you can be 'precise enough' and fairly quick
2023-04-21 10:16:17 <veltas      > And it's more comfortable because hand doesn't need to move at all
2023-04-21 10:16:53 <veltas      > But I do think (and I've demonstrated with small sample size) even people used to trackpad are faster with touchpad when they're used to that too
2023-04-21 10:17:05 <veltas      > They just use it very occasionally
2023-04-21 10:17:27 <veltas      > Also trackpads are highly variable in quality, most Dell ones feel awful
2023-04-21 14:16:18 <KipIngram   > Hey, where is root's PATH set?
2023-04-21 14:17:00 <KipIngram   > And where should I put commands I want to run on *boot* (not login), but guaranteed last, after the os has done everything else it's going to do setup-wise?
2023-04-21 14:17:13 <KipIngram   > I found /etc/rc.local suggested, but it's not guaranteed to run last.
2023-04-21 14:17:44 <KipIngram   > Worst case I could fire a script from there that waits for some delay period and then runs my commands, but that seems awfully ugly.
2023-04-21 14:18:41 <thrig       > with systemd one might run something after multi-user.target
2023-04-21 14:19:08 <KipIngram   > veltas: Yeah, I really should work on that track button.  I suspect that once I got used to it I'd far prefer it.
2023-04-21 14:19:35 <KipIngram   > What are you distinguishing with "trackpad" vs. touchpad?
2023-04-21 14:19:46 <KipIngram   > Are you calling that little button trackpad?
2023-04-21 14:20:07 <KipIngram   > To me "pad" seems like something large and flat.
2023-04-21 14:20:33 <KipIngram   > thrig: Ok, I may have to tackle systemd.
2023-04-21 18:05:14 <veltas      > Who even knows what I mean
2023-04-21 18:05:47 <veltas      > I think i mean TrackPoint instead of trackpad
2023-04-21 18:06:23 <veltas      > The thinkpad nub thing
2023-04-21 18:14:08 <veltas      > Yeah I barely use my touchpad but I can still get to exact positions faster with it than the trackpoint
2023-04-21 18:14:17 <veltas      > And then a mouse is another level above that
2023-04-21 18:15:07 <veltas      > But the trackpoint is just cozy
2023-04-21 18:30:53 <veltas      > crc: https://ilo.retroforth.org/snapshot.tar.gz is empty
2023-04-21 20:15:06 <crc         > veltas: fixed; thanks for pointing this out