2023-05-11 02:27:03 <veltas      > Nope, there were more bugs but I'm getting closer
2023-05-11 08:06:32 <Zarutian_iPad> hmm… I recently discovered something regarding teaching techniques that I find quite irritating: the assumed-to-be obvious implication
2023-05-11 08:09:27 <Zarutian_iPad> this leads to non explaining “explanations” that certain kind of teacher/profesor/educator favour for some reason
2023-05-11 08:10:01 <unjust      > is it a way to gloss over gaps in their own understanding?
2023-05-11 08:10:50 <Zarutian_iPad> I suspect it is
2023-05-11 08:12:55 <Zarutian_iPad> and I have seen this aforesaid propensity corralates with those that think they are some sort of “authority”.
2023-05-11 08:14:01 <unjust      > what field/domain did you notice this pattern in?
2023-05-11 08:15:14 <Zarutian_iPad> a bit in what USAians call STEM but mostly in what often is refered to as humanities
2023-05-11 08:16:02 <Zarutian_iPad> though this is a misleading translations of similiar concepts in Icelandic
2023-05-11 08:16:47 <Zarutian_iPad> vísindi and samfélagsfræði are the two I was trying to translate
2023-05-11 08:17:14 <KipIngram   > I agree with you totall about that sort of thing, Zarutian_iPad
2023-05-11 08:17:32 <KipIngram   > I think unless you can explain something clearly you don't really understand it.
2023-05-11 08:18:14 <Zarutian_iPad> the thing I suspect is that many never learn how to explain something
2023-05-11 08:18:21 <KipIngram   > I think this effect it partially caused by people who don't bother to acquire "foundational knowledge."  They'll learn level, say, four - but ignore mastering 1-3.
2023-05-11 08:18:22 <unjust      > those would make sense to me as "sciences" and "social studies" after translating them
2023-05-11 08:19:01 <KipIngram   > STEM also includes math.  I've always found it odd how "buzz terms" catch on and then just get tossed around willy-nilly.
2023-05-11 08:20:21 <unjust      > maybe the emphasis on pursuit of theory and lack of practical application of theory for a lot of academics, is what allows them to get away with this behaviour unchecked
2023-05-11 08:20:26 <Zarutian_iPad> maths seem to atract certain kind of eccentric that often refuse to elaborate and use notation that is confusing as hell
2023-05-11 08:21:01 <Zarutian_iPad> unjust: those translations might be close enough what I mean
2023-05-11 08:21:13 <KipIngram   > Yes, they develop a "priesthood" of terminology that turns it into a "clossed domain."  They tend to forget how to explain things to "outsiders."
2023-05-11 08:21:28 <KipIngram   > It really is particularly bad in math.
2023-05-11 08:22:08 <Zarutian_iPad> math and if I read what 0xabad1dea says, musicology
2023-05-11 08:22:47 <KipIngram   > Many times I run into that in math; I'll read a paper or a treatment or whatever, and just can't fathom it.  Over and over.  Then at some point later, when I finally find an explanation made in ters that are more commonly understood, I see that the thing really isn't THAT hard.
2023-05-11 08:23:03 <Zarutian_iPad> s/says,/says correctly,/
2023-05-11 08:25:31 <Zarutian_iPad> indeed
2023-05-11 08:26:23 <KipIngram   > I get the impression that most technical papers in math are not written to educate - they're written to communicate with other pepole that are already experts in the field.
2023-05-11 08:26:47 <Zarutian_iPad> onething I realized around my twenties or so is education is not the same as knowledge or wisdom.
2023-05-11 08:27:13 <KipIngram   > I've been tryinig for YEARS to get my arms around Roger Penrose's "twistors," and I still havn't even begun to.
2023-05-11 08:27:38 <KipIngram   > And I do think it's primarily a terminology barrier.
2023-05-11 08:28:12 <Zarutian_iPad> sure, I get that when I watch iBology lectures/seminars on molecular biology but at least there I can look up the bloody terms they use and gain some glimmering of understanding of what they are talking about
2023-05-11 08:28:58 <KipIngram   > It's not quite as bad in science, I think.  It's "there," but it seems easier to dig into and get some insight.
2023-05-11 08:29:35 <KipIngram   > But in math these terms form some sort of "circular network" an dit's hard to figure out where to start untying it.
2023-05-11 08:29:44 <Zarutian_iPad> I like to think of this progression: data -> information -> knowledge -> wisdom
2023-05-11 08:30:19 <KipIngram   > Maybe it's because science ultimately connects with "the world," but math is more of a stand-alone domain.
2023-05-11 08:30:51 <Zarutian_iPad> and turning knowledge into wisdom requires quite a bit of reflection
2023-05-11 08:31:02 <KipIngram   > It does.
2023-05-11 08:31:34 <KipIngram   > There's been a general trend downward in people's attention span the past few decades though - so many people just can't be bothered to really invest significant time in somehting.
2023-05-11 08:31:47 <KipIngram   > And that "reflection" takes time.
2023-05-11 08:32:44 <Zarutian_iPad> I do think that attention span per say is not descreasing that much but amount of interruptions have
2023-05-11 08:33:43 <Zarutian_iPad> both internal “I am so bored” and external “here is an obnoxious ad in this deep dive youtube video you are watching”
2023-05-11 08:34:00 <unjust      > the decision fatigue from those interruptions can definitely knock your attention span
2023-05-11 08:36:56 <Zarutian_iPad> I noticed recently with telly shows on Netflix and such how short some shows are or feel like they just get ended prematurly. Both the episodes and the series.
2023-05-11 08:38:20 <Zarutian_iPad> my reaction is often when I felt like I have really gotten ‘stuck in’ and the thing ends “that’s it? where is the rest?”
2023-05-11 08:40:11 <Zarutian_iPad> also so recreational writing I have read recently has suffered in samish way but also in quality. Like Charles Stross’es Dead Lies Dreaming
2023-05-11 08:42:37 <Zarutian_iPad> too many shallow characters and too much pandering to Polly Anne types in the audience.
2023-05-11 08:43:49 <Zarutian_iPad> broke immersion quite a bit and the story felt too mechanical in its execution
2023-05-11 08:45:46 <Zarutian_iPad> something that Stross is not wont to do. I blame his Covid infection.
2023-05-11 08:48:04 <Zarutian_iPad> but back to the topic of the aforesaid non-explainers
2023-05-11 08:49:18 <Zarutian_iPad> as I said I suspect that these folks are glossing over gaps in their own understanding and have never learned how to explain something
2023-05-11 08:49:42 <thrig       > many bodices were ripped to bring you these killer moon plans
2023-05-11 08:51:14 <Zarutian_iPad> the ways I have learned to explain something is pretty much ‘stolen’ from Feynmann and others
2023-05-11 08:51:18 <KipIngram   > My problem with TV is that I seem to just want a different degree of depth than "the mainstream" wants, and shows are definitely targeted at the mainstream.
2023-05-11 08:51:51 <Zarutian_iPad> thrig: I think you nailed it with that as a description of that kind of genre
2023-05-11 08:52:22 <thrig       > (someone wanted chatgpt to churn out a stream of dopamine daze for them, so that's what I came up with at the time)
2023-05-11 08:52:24 <KipIngram   > Plus I just get really tired of seeing "political correctness" overtly injected into one show / movie / book after another.  Even when I have absolutely no "personal concern" about the specific issue at all, it's still an easily recognizable thing, and it detracts from entertainment value.
2023-05-11 08:52:46 <KipIngram   > I.e., it's not that the material "offends me" - it's just a "there they go again" thing.
2023-05-11 08:54:02 <drakonis    > you mean hamfisted as opposed to something natural?
2023-05-11 08:54:24 <KipIngram   > Well, I guess so.  Stuff that seems "artifically inserted" to check a box.
2023-05-11 08:54:55 <KipIngram   > If the show is actually ABOUT that in some fashion, then it's fine, but often it feels artificial and "sidebar" to me.
2023-05-11 08:59:46 <KipIngram   > What "concerns" me about it is that we'll develop a culture in which some particular "list" of such things becomes almost a requirement - i.e., that otherwise good material won't get launched without that list being tended to.
2023-05-11 08:59:58 <drakonis    > we already have
2023-05-11 09:00:01 <drakonis    > we always have
2023-05-11 09:00:32 <thrig       > blazing saddles, or remember that one saturday night live cartoon that they never aired again (it was not really pro Corporations)
2023-05-11 09:00:45 <KipIngram   > Well, and we also had a culture in the past where one COULDN'T present such things, and that was bad too.
2023-05-11 09:02:19 <drakonis    > that we still do
2023-05-11 09:04:25 <thrig       > where's that video... https://www.youtube.com/watch?v=jzMFoNZeZm0
2023-05-11 09:05:37 <KipIngram   > drakonis: I guess so, but it was considerably worse in the past.
2023-05-11 09:06:01 <drakonis    > thrig: its funny because there's people that think you couldnt do the office today
2023-05-11 09:06:39 <KipIngram   > Heck, at one point in time you couldn't even air a  TV show in which the "cops didn't win."
2023-05-11 09:07:08 <KipIngram   > I think that didn't really begin to loosen up much until the 1970's or so.
2023-05-11 09:08:08 <Zarutian_iPad> KipIngram: yeah, PC-checkboxed shows like StarTrek Discovry meant that it was single watch only and often with lot of fast fowarding over certain bits.
2023-05-11 09:08:23 <KipIngram   > Yes.
2023-05-11 09:08:56 <KipIngram   > And it's not that I don't think some of these issues are worth giving thought to.  I just don't want that thought mixed in with my entertainment.
2023-05-11 09:09:37 <Zarutian_iPad> contrast it with DeepSpace Nine which tackled a lot of this in rather intresting ways
2023-05-11 09:10:57 <thrig       > and DS9 came out after Paramount looked at a Straczynski proposal, somehow
2023-05-11 09:11:20 <KipIngram   > Yeah, it was their attempt to get in on the Babylon 5 game.
2023-05-11 09:11:28 <KipIngram   > Those shows came out right together.
2023-05-11 09:12:05 <KipIngram   > I assume JMS pitched B5 to them; they turned it down and then went and did their own.
2023-05-11 09:12:10 <thrig       > http://sequart.org/magazine/67733/deep-space-nine-and-babylon-5-remarkably-similar%E2%80%94or-similarly-remarkable/
2023-05-11 09:12:27 <KipIngram   > I guess if you OWN Star Trek, it's easy to see why you might prefer it as a vehicle over a "brand new universe."
2023-05-11 09:12:59 <Zarutian_iPad> the thing about DS9 is how well written and beleavable the stuff is
2023-05-11 09:13:01 <KipIngram   > But JMS had a different story to tell - you COULDN'T have told that story in the Trek verse.
2023-05-11 09:13:07 <KipIngram   > It is quite good.
2023-05-11 09:13:53 <Zarutian_iPad> rather than the ‘psychic crybaby caused the lithium burn’ crap in StarTrek Discovery
2023-05-11 09:14:29 <Zarutian_iPad> dilithium*
2023-05-11 09:14:45 <Zarutian_iPad> anyway be back later
2023-05-11 09:17:35 <KipIngram   > I'm a big Babylon 5 fan.
2023-05-11 09:19:11 <KipIngram   > I thought pretty well of DS9 too, but I do think they "lifted" some ideas from JMS's pitch.
2023-05-11 09:19:33 <KipIngram   > Not the first time we've seen two very similar shows pop up on different networks at the same time, though.  :-)
2023-05-11 09:22:24 <KipIngram   > You see it in movies too.  I don't remember the other franchise name, but on one side you had Olympus Has Fallen.
2023-05-11 09:22:36 <KipIngram   > There was a very very similar string of movies right along with it.
2023-05-11 09:24:03 <thrig       > there's only so many hollywood writers, and they probably know or can "look over the cubicle wall"
2023-05-11 09:25:51 <drakonis    > the problem is hollywood actually
2023-05-11 09:26:02 <drakonis    > and there's a lot of nepotism in there
2023-05-11 09:26:10 <KipIngram   > Yeah.
2023-05-11 09:26:12 <drakonis    > plus it tends to follow trends a lot
2023-05-11 09:26:32 <drakonis    > so they go after whatever sells the most
2023-05-11 09:26:51 <KipIngram   > Yes, certainly if a movie comes out and does well, others want in on that.
2023-05-11 09:26:52 <drakonis    > its not a matter of sticking "political correctness" everywhere, its because it makes someone money
2023-05-11 09:27:07 <KipIngram   > In some cases, though, it's so close to the same time you have to figure there's been some spying going on.
2023-05-11 09:27:15 <drakonis    > bad media gets made all the time
2023-05-11 09:27:25 <KipIngram   > Unavoidable.
2023-05-11 09:27:31 <KipIngram   > Just giving the QUANTITY of media made.
2023-05-11 09:27:59 <KipIngram   > It's like staffing a company.  Small companies can at least TRY to acquire all excellent people, but a big corporation is stuck with the law of averages to some extent.
2023-05-11 09:28:48 <KipIngram   > They can try to shift their bell curve to the right, some, but the bigger they are the harder that is to do.
2023-05-11 09:28:49 <drakonis    > do consider that hollywood is a realm of knowing people to do things
2023-05-11 09:29:21 <drakonis    > there's some really bad things getting made because the people making them are connected
2023-05-11 09:29:28 <drakonis    > see also nepo babies
2023-05-11 09:30:44 <drakonis    > which is slang for people who're born to successful people and are propped up by their parents to success
2023-05-11 09:31:08 <thrig       > born on 3rd base and thinks he hit a triple
2023-05-11 09:31:10 <drakonis    > its pretty much people who shouldnt have any business being successful
2023-05-11 09:31:13 <drakonis    > yes
2023-05-11 09:31:37 <drakonis    > see also ya boy musk rat
2023-05-11 10:03:50 <KipIngram   > Yeah, nepotism just isn't likely to go away.  It's "wired into us" t some extent to look out for our kids.
2023-05-11 10:04:11 <KipIngram   > Family, neighbors, friends, etc.  Lots of levels of that.
2023-05-11 10:04:20 <thrig       > social primate gonna socialize
2023-05-11 10:04:27 <KipIngram   > Yup.
2023-05-11 10:04:35 <thrig       > well, most of them
2023-05-11 10:04:50 <KipIngram   > And "most" is what really drives the world.
2023-05-11 10:05:01 <KipIngram   > The exceptions make interesting stories, but they're edge cases.
2023-05-11 10:07:05 <drakonis    > and then they're held up as examples that the system works
2023-05-11 10:13:02 <KipIngram   > Good point - and that's a horribly skewed way to look at things.
2023-05-11 10:13:13 <KipIngram   > People just don't operate "rationally" a whole lot of the time.
2023-05-11 10:13:54 <thrig       > Spock was reportedly a hot ticket when he had his ears on
2023-05-11 10:18:33 <KipIngram   > My favorite "Spock" video:
2023-05-11 10:18:39 <KipIngram   > https://www.youtube.com/watch?v=BC35cQKHwzg
2023-05-11 10:19:44 <thrig       > I wonder if the Tolkien Estate would let that fly these days
2023-05-11 10:25:40 <KipIngram   > No telling.  I'm glad it's out there, though.  It's fun, and it doesn't hurt that the girls are pretty adorable.
2023-05-11 10:28:05 <KipIngram   > It's odd - if you ask me, that's "free advertising" for the Tolkien estate.
2023-05-11 10:28:23 <KipIngram   > And I doubt anyone made any kind of serious money on it, but I guess you can never tell.
2023-05-11 12:22:24 <KipIngram   > Well, it looks like I did a lot of work last weekend for nothing.
2023-05-11 12:22:51 <KipIngram   > This drive has some announce data.  Various teams then established their start dates based on that - you can work your way backwards.
2023-05-11 12:23:17 <KipIngram   > And the outcome of that process resulted in me having a May 12th due date to hand off a particular set of test results, that I get by testing the drive.
2023-05-11 12:23:29 <KipIngram   > So I knocked myself out over hte weekend and earlier this week to get that done.
2023-05-11 12:23:40 <KipIngram   > But now the local guys are asking "Why are we doing that so early?"
2023-05-11 12:23:59 <KipIngram   > They expect to have further improvements to the firmware, and they naturally want those results to reflect their best work.
2023-05-11 12:24:09 <KipIngram   > This happens EVERY CYCLE.
2023-05-11 12:24:25 <KipIngram   > I think the root cause must be that the whole plan is made backwards.
2023-05-11 12:24:56 <KipIngram   > You'd THINK that we'd establish what the desired product needed to look like, then let all the teams make their own plans, and let that DETERMINE the announce data.
2023-05-11 12:25:22 <KipIngram   > But I've got a strong suspicion that some executive just makes an X and a calendar, based on business considerations, and declares that when announce MUST BE.
2023-05-11 12:25:41 <KipIngram   > And the rest of the whole business is everyone jockeying to push the pain off of themselves and onto other teams.
2023-05-11 12:26:37 <KipIngram   > In past cycles I've been explicitly told to not do that work yet, but that particular guy wasn't paying as much attention this time.
2023-05-11 12:26:54 <thrig       > thankfully we've left such practices as the I Ching or geomancy behind us
2023-05-11 12:27:31 <KipIngram   > Anyway, getting all the work done isn't something that can magically be accomplished in whatever length of time is convenient for an executive.
2023-05-11 12:28:11 <KipIngram   > I'll wind up doing all this work again in a month or two, with new firmware.
2023-05-11 12:28:56 <thrig       > 
ACTION goes back to pulling quotes from https://ctext.org/analects/yao-yue#n1600

2023-05-11 12:30:00 <Zarutian_iPad> sounds like someone got their ?Gantt charts? backwards.
2023-05-11 12:30:09 <KipIngram   > At least in the past when that guy told me outright not to do the work, I had cover.  This time I had to do it, because... it was on the calendar.
2023-05-11 12:30:25 <KipIngram   > Not doing it would have shown up as MY failure.
2023-05-11 12:31:06 <KipIngram   > 90% of the work I do is for guys local here in Houston, the guys that actually develop the drive.  This thing is really my ONE deliverable that the guy who's actually my manager (in England) cares about.
2023-05-11 12:31:16 <KipIngram   > So I had little option that to demonstrate I could meet the date.
2023-05-11 12:31:30 <KipIngram   > "other" than
2023-05-11 12:31:40 <Zarutian_iPad> calanders for planning should only be used when doing acriculture and observing astronomical events
2023-05-11 12:32:28 <Zarutian_iPad> wait your manager is in England yet you work in USA?
2023-05-11 12:32:39 <KipIngram   > My supervising prof in grad school had a good line - "plans are useless, but planning is essential."
2023-05-11 12:32:44 <thrig       > globalization, bitches!
2023-05-11 12:32:59 <KipIngram   > It's worth having your expectations written down, for people to calibrate around.
2023-05-11 12:33:03 <Zarutian_iPad> no, not what I meant
2023-05-11 12:33:07 <KipIngram   > But they can't be turned into religious artifacts.
2023-05-11 12:33:22 <KipIngram   > They need to be constantly updated based on unfolding events.
2023-05-11 12:33:30 <KipIngram   > And that includes the last date on the plan - the announce date.
2023-05-11 12:33:39 <Zarutian_iPad> dates with error bars is what I have seen
2023-05-11 12:35:43 <Zarutian_iPad> the error bars start out big as hell then the date gets refined as the stuff winds along
2023-05-11 12:37:42 <KipIngram   > Yeah, that's useful too.
2023-05-11 12:37:50 <KipIngram   > Information is useful, but only if it's "real."
2023-05-11 12:38:00 <KipIngram   > If you're not allowed to move a date, then that date stops being real.
2023-05-11 12:38:14 <KipIngram   > And everyone just tries to make sure they won't be the ones that get in trouble when it's missed.
2023-05-11 12:38:27 <KipIngram   > And that's energy wasted in the big picture.
2023-05-11 12:39:01 <Zarutian_iPad> have you thought about getting your boss kidnapped so that it is HE who misses the date?
2023-05-11 12:42:23 <KipIngram   > :-)
2023-05-11 12:42:35 <KipIngram   > I actually like my boss.  Nice enough guy.
2023-05-11 12:42:44 <KipIngram   > Plus I just got a raise, so...
2023-05-11 12:42:50 <KipIngram   > Guess I should be nice to him.
2023-05-11 12:43:11 <Zarutian_iPad> okay so he is not the problem but his boss is?
2023-05-11 12:43:30 <KipIngram   > It's the whole setup, I think.
2023-05-11 12:43:54 <KipIngram   > It wouldn't surprise me to learn that no one executive really has much ability to change how things are done.
2023-05-11 12:44:06 <KipIngram   > It's this whole interlocking mess of habits and practices.
2023-05-11 12:44:37 <KipIngram   > I suspect there's an element of "the executive does X because X is what successful executives do."
2023-05-11 12:44:51 <Zarutian_iPad> so, give your time estimates in made up units that you refuse to elaborate on
2023-05-11 12:45:05 <KipIngram   > If he let the announce date float to its natural point he'd get accused of not being aggressive enough, etc.
2023-05-11 12:45:14 <KipIngram   > :-)
2023-05-11 14:14:46 <KipIngram   > Ok, I think that while meta-compiling a new system of the sort I envision, there will be four separate RAM regions that will have content saved in the load image.
2023-05-11 14:14:58 <KipIngram   > In a perfect world I'd be able to build into four disk buffers.
2023-05-11 14:15:27 <KipIngram   > I don't know if that's enough space for all the four regions, but it would be nice not to have to spread any of them over multiple buffers.
2023-05-11 14:16:07 <thrig       > four... Death, Famine, War, and Pestilence?
2023-05-11 14:16:08 <KipIngram   > Then I'll have a word that will take a block # and load a system into RAM from the four continguous blocks at that point.
2023-05-11 14:16:16 <KipIngram   > Oh, ok, that is cool.
2023-05-11 14:16:24 <KipIngram   > I'll have to play with that.  Nice.
2023-05-11 14:16:48 <KipIngram   > There will actually be two other regions in a running system - a region for disk buffers and one for error recovery snapshots.
2023-05-11 14:17:00 <KipIngram   > But those won't have any pre-existing content in the load image.
2023-05-11 14:17:19 <KipIngram   > The four are 1) code, 2) CFA table, 3) PFA table, and 4) headers.
2023-05-11 14:17:55 <KipIngram   > I'm inclined to embed information at the start of each one specifyng how much RAM to allocate for that particular region.
2023-05-11 14:18:15 <KipIngram   > And each one will also contain a "region type" indicator, that guides me in how to relocate it while loading it.
2023-05-11 14:19:13 <KipIngram   > Code region will the cells corresponding to the other regions set to the right values.  Tables will need the base address of the code region added to the initial table entry offsets.
2023-05-11 14:19:15 <KipIngram   > Etc.
2023-05-11 14:20:11 <KipIngram   > So starting up will involve allocating the six regions, modifying things to link them all together (which includes setting some registers and setting a list of variables), and jumping to the right place.
2023-05-11 14:21:07 <KipIngram   > I think it may fit.  My xt's are going to be 16 bits, and usually a 16-bit system can come in around 8k.  Four 4k blocks will give me 16k to work with.
2023-05-11 14:21:52 <KipIngram   > There won't be more than a few hundred words at startup, and the tables will have an eight-byte entry for each one.  So long as I'm under 512 words those will fit.
2023-05-11 14:29:08 <KipIngram   > I think the region that will overflow, if any of them to, will be the code region.  I'm pretty sure the other three look almost certain to fit.
2023-05-11 14:29:33 <KipIngram   > And, if it does, then oh well - that just means I have to use two blocks for that.  Annoying, but not a show stopper.
2023-05-11 14:30:45 <KipIngram   > I'll also need a way to start a system up from the Linux command line - same logic, but packaged differently.
2023-05-11 14:31:28 <KipIngram   > Maybe I can find a clever way to quickly give control the Forth loader - haven't contemplated it  yet.
2023-05-11 14:33:13 <KipIngram   > Maybe if it became necessary to go past one block for that I'd just set it up to process a list of items, and if the same region type got called out again it would just append to the first one.
2023-05-11 14:33:54 <KipIngram   > Because I might like to be able to take a running system that I'd loaded a bunch of new code into and zip it out to a batch of disk blocks so it could be reloaded later.
2023-05-11 14:40:33 <KipIngram   > Yeah, I should plan it that way; I'm bound to want to be able to save a larger system later.
2023-05-11 15:23:07 <veltas      > YAGNI though
2023-05-11 15:50:25 <KipIngram   > Maybe not, but I've decided it would be easy.  And even though I'm unlikely to write an initial system that needs it, I might want to save a system after building other things.  Maybe - who knows?
2023-05-11 15:51:03 <KipIngram   > But like I said, I think it just involves appending, and adding a flag to the block header that indicates when a block is the "last of that region."  Or an "n of m" indicator.
2023-05-11 15:52:25 <KipIngram   > I will be pleased if/when the initial system comes in with only one block of each type.
2023-05-11 16:44:18 <vms14       > KipIngram: If I have a way to merge lists/hashes into a bigger hash, if one of the two lists is missing a value how would you expect it to work?
2023-05-11 16:44:45 <vms14       > for example: [ 1 2 3 ] [ 4 5 ] merge
2023-05-11 16:44:58 <vms14       > {'5' => undef,'1' => '2','3' => '4'}
2023-05-11 16:45:24 <vms14       > see the 5 => undef  it's because the both lists together don't have a % 2 value
2023-05-11 16:45:40 <vms14       > so I just add an undefined value to compensate
2023-05-11 16:45:53 <vms14       > but, wouldn't you expect to be the 3 who gets the undefined?
2023-05-11 16:46:12 <vms14       > or should it be an error?
2023-05-11 16:47:29 <thrig       > the feds might block the merger
2023-05-11 16:48:51 <vms14       > if the sum of both lists have a len divisible by 2, then I don't bother
2023-05-11 16:49:18 <vms14       > but if not, I don't know how I should behave
2023-05-11 16:49:45 <vms14       > for now I just add an undef at the end, so the last element will get that as a value
2023-05-11 16:50:11 <vms14       > but maybe you wanted the 3 with undefined value and the 5 be the value of the key 4
2023-05-11 16:50:51 <vms14       > looks like the most correct option is to be an error
2023-05-11 17:01:37 <vms14       > I'm mostly thinking the best choice is to make it an error
2023-05-11 17:04:57 <vms14       > nah, I'll make the list that has the non multiple by 2 len receive the undefined
2023-05-11 17:05:32 <vms14       > [ 1 ] [ 3 4 ] ....
2023-05-11 17:05:39 <vms14       > {'3' => '4','1' => undef}
2023-05-11 17:07:38 <unjust      > for comparison: python has a function called zip that combines elements from iterables into a sequence of tuples, which you can use to build a list of pairs or to create a dictionary/hash(map/table)
2023-05-11 17:07:46 <unjust      > ie.
2023-05-11 17:07:47 <unjust      > >>> dict(zip([1, 2, 3], [4, 5]))
2023-05-11 17:07:47 <unjust      > {1: 4, 2: 5}
2023-05-11 17:07:55 <unjust      > >>> list(zip([1, 2, 3], [4, 5]))
2023-05-11 17:07:56 <unjust      > [(1, 4), (2, 5)]
2023-05-11 17:12:28 <vms14       > i'll also have zip, but it will work only with lists
2023-05-11 17:12:37 <vms14       > althought you can convert a list into a hash later
2023-05-11 18:48:25 <KipIngram   > vms14: I'm not sure exactly what you mean by merge.  Just concatenate?  Interleave using some order?  What exactly?
2023-05-11 18:49:21 <KipIngram   > So, how big is a cache line in x86?
2023-05-11 18:50:14 <KipIngram   > I want to say 64 bytes, but that may be "dated" knowledge.
2023-05-11 18:51:12 <KipIngram   > https://igoro.com/archive/gallery-of-processor-cache-effects/
2023-05-11 18:52:44 <vms14       > KipIngram: nvm I made a decision, althought I think maybe it's better to have an error instead
2023-05-11 18:53:03 <vms14       > but it mainly takes two lists or two hashes and creates a new hash from those
2023-05-11 18:53:27 <vms14       > the thing is sometimes the both lists are missing one value to have a multiple of 2
2023-05-11 18:53:36 <vms14       > as I need to have key/value pairs
2023-05-11 18:54:14 <vms14       > [ oh my ] [ cat is ] ....    this would create a hash with { oh => my,  cat => is }
2023-05-11 18:54:31 <vms14       > the thing is I had to decide what happens when one of those is missing
2023-05-11 18:54:39 <vms14       > like [ oh ] [ my cat ] .....
2023-05-11 18:54:58 <vms14       > will the oh take my as a value and cat have undefined as value?
2023-05-11 18:55:33 <vms14       > what I decided is the list which has the non multiple length is the one that will take undef
2023-05-11 18:55:46 <vms14       > so in this case oh => undef, my => cat
2023-05-11 18:56:15 <vms14       > but if the sum of both list elements is multiple of 2, then there's no problem
2023-05-11 18:56:27 <vms14       > [ oh ] [ my cat is ] ....
2023-05-11 18:56:35 <vms14       > oh => my, cat => is
2023-05-11 18:56:51 <vms14       > it's a bit weird, it's cleaner to mark it as an error
2023-05-11 19:00:22 <KipIngram   > Oh, you want to take keys from one list and values from the other?
2023-05-11 19:16:31 <KipIngram   > I asked about cache lines because, while I do want to separate my data (and definitions) from code, just about the very first thing I need to pin down in a system image is the key variables that track the dictionary state.
2023-05-11 19:16:59 <KipIngram   > So I'm picturing one cache line - eight variable cells, right at the start of the image, and THEN code will follow that.
2023-05-11 19:17:20 <KipIngram   > If they're not sharing a cache line, it seems like that ought to be ok.
2023-05-11 19:18:01 <KipIngram   > Four of them will be "next available byte" pointers for the four regions.
2023-05-11 19:23:23 <KipIngram   > I'll figure out what the other four should do later.
2023-05-11 19:36:10 <vms14       > KipIngram: no, both things from both lists xd
2023-05-11 19:37:00 <vms14       > if in one list there's not enough stuff to get key/value pairs it will continue with the other list
2023-05-11 19:37:34 <vms14       > unless the sum of both lists len is not multiple of 2, then the list that is missing some value will get an undef
2023-05-11 19:49:45 <vms14       > I have something that I like named 'with.buffer'
2023-05-11 19:50:10 <vms14       > it redirects . to a temporal buffer and will return that buffer at the end
2023-05-11 19:50:19 <vms14       > [ 'oh... . ] with.buffer
2023-05-11 19:50:47 <vms14       > instead of printing on stdout, it will write it on a buffer and return a string with the contents
2023-05-11 19:51:36 <vms14       > it also accepts the name of a word, so you can redirect whatever a word prints into a string
2023-05-11 19:52:13 <vms14       > crc: not sure if it would be useful for you, the implementation is quite easy
2023-05-11 19:52:31 <vms14       > althought it may slow a bit the '.' word, unless you use another
2023-05-11 19:53:01 <vms14       > I can attach a filehandle to '.' and once it's attached '.' will print on that filehandle instead of stdout
2023-05-11 19:53:32 <vms14       > that 'with.buffer' creates one filehandle that when written to it it will fill a string
2023-05-11 19:53:43 <vms14       > you can also attach into a socket and you'll write there
2023-05-11 20:04:31 <crc         > my `.` equivalent (`n:put`) is defined as `n:to-s s:put`, with the `n:to-s` being in assembly and `s:put` being `&c:put s:for-each`
2023-05-11 20:05:08 <crc         > I have some facilities in RetroForth for capturing output to a string (all output runs through `c:put`, which is basically equivalent to a DEFER'd word, letting me redefine it to write to a string buffer). These would also be usable under Konilo with minor tweaks.
2023-05-11 20:07:14 <crc         > update on the javascript stuff: all but saving blocks & rom to IndexedDB is working. I should have those bits done in another day or two
2023-05-11 20:07:45 <vms14       > that's nice
2023-05-11 20:07:47 <crc         > My son is working on a version using <canvas>, with support for graphics; this is still quite buggy on the input handling, but he thinks it'll be ready to share sometime next week.
2023-05-11 20:07:55 <vms14       > I was wondering if you had something similar
2023-05-11 20:08:43 <vms14       > crc:  how it is that you have "rom" in indexeddb, but not able to save?
2023-05-11 20:09:09 <crc         > rom & blocks aren't in indexeddb yet; just cached in memory at present
2023-05-11 20:09:21 <vms14       > I saw a video that gave a nice intro about indexeddb, but I assume you've passed that knowledge already
2023-05-11 20:10:23 <crc         > I have an (uncomitted to repo) branch that I'm testing with this
2023-05-11 20:10:49 <vms14       > https://www.youtube.com/watch?v=yZ26CXny3iI&pp=ygUJaW5kZXhlZGRi
2023-05-11 20:10:53 <crc         > I tried using localStorage, but the limits are too small for the full block set.
2023-05-11 20:11:00 <vms14       > it served me to get a general idea
2023-05-11 20:11:11 <crc         > Thanks, queuing to watch tonight
2023-05-11 20:11:37 <vms14       > it explains the basics, and all the minimum stuff you need to set to get it working, which is a bit
2023-05-11 20:11:52 <vms14       > but it does explain everything very well, it was nice for me
2023-05-11 20:12:08 <vms14       > althought he mentioned something that I expected him to explain, but he didn't xD
2023-05-11 20:12:30 <vms14       > anyways, with that video you can comfortably go to mdn documentation and understand everything
2023-05-11 20:16:29 <vms14       > crc: did you use microdata on your html?
2023-05-11 20:16:55 <vms14       > I'm going to play with it
2023-05-11 20:17:07 <vms14       > althought I have to make an html generator first :D
2023-05-11 20:17:12 <crc         > no
2023-05-11 20:17:13 <vms14       > html generators are fun
2023-05-11 20:19:50 <crc         > My HTML generation is similar to: http://retroforth.org/examples/HTML.retro.html (the one in my employer's application is larger, with many more elements, but not open source)
2023-05-11 20:20:36 <vms14       > I had this way at first
2023-05-11 20:20:49 <vms14       > I like the lisp style much more, which is kind of the reverse
2023-05-11 20:21:42 <vms14       > [ html [ head ... ] [ body [ h1 #title " Hi..." :color red .title-class ] ] ]
2023-05-11 20:22:18 <vms14       > i had this in the js version, instead of writing html it created dom elements
2023-05-11 20:23:10 <vms14       > the reason I don't need a lisp interpreter is because I already have it
2023-05-11 20:23:17 <vms14       > "I have lists" xD
2023-05-11 20:23:44 <vms14       > just add a word expecting a list and it's done
2023-05-11 20:24:05 <vms14       > I could also have some kind of execution when building lists, but meh
2023-05-11 20:24:42 <crc         > My [ ] construct anonymous functions, not lists :)
2023-05-11 20:24:50 <vms14       > oh
2023-05-11 20:25:06 <vms14       > I use lists as code, I can even "compile" that code
2023-05-11 20:25:23 <vms14       > the fake compile seems to add a lot of performance
2023-05-11 20:25:48 <vms14       > 0 1 1000000 range [ 1 2 3 drop drop drop + ] do.list .
2023-05-11 20:25:59 <vms14       > this takes 34 seconds to complete on my machine
2023-05-11 20:26:11 <vms14       > 0 1 1000000 range [ 1 2 3 drop drop drop + ] compile do.list .
2023-05-11 20:26:17 <vms14       > this takes 4 seconds instead XD
2023-05-11 20:27:04 <vms14       > 4.08 real
2023-05-11 20:27:20 <vms14       > https://termbin.com/rn7jj
2023-05-11 20:27:32 <vms14       > you can try it on your machine with "time perl this.file"
2023-05-11 20:27:54 <thrig       > time to die
2023-05-11 20:27:57 <vms14       > if you go to the end of the line and remove the 'compile' word, then it takes 34 seconds
2023-05-11 20:28:45 <vms14       > my machine is slow, a guy told me it took 1.17 on his machine
2023-05-11 20:30:22 <vms14       > the whole thing of "lists as code" makes for a lisp
2023-05-11 20:32:03 <vms14       > I assume your [] is equivalent to my [] compile
2023-05-11 20:32:40 <vms14       > compile returns a subroutine, unless it compiled an immediate word that does not return any result, like :
2023-05-11 20:33:38 <crc         > 10.696s vs 1m0.104s on my phone, 5.348s vs 38.530s on my openbsd box
2023-05-11 20:34:12 <vms14       > I had a compiling flag, thought could be useful to have words being able to know if it's adventure time and to act differently, also they could push stuff on a word being defined
2023-05-11 20:34:17 <vms14       > crc: lol
2023-05-11 20:34:24 <vms14       > I didn't try on my phone :0
2023-05-11 20:34:47 <vms14       > but your openbsd box gives kind of similar results
2023-05-11 20:35:14 <vms14       > I was a bit amazed by the time difference
2023-05-11 20:35:28 <vms14       > also scared of how much overhead every word adds
2023-05-11 20:35:38 <thrig       > 
ACTION ponders adding pledge and unveil to retroforth

2023-05-11 20:35:43 <crc         > I do a lot of my programming on the phone
2023-05-11 20:35:44 <vms14       > the main point of colon words is that they're cheap :/
2023-05-11 20:36:16 <vms14       > crc: trough ssh?
2023-05-11 20:36:25 <vms14       > through*
2023-05-11 20:36:38 <crc         > thrig: exposing those is on my todo list
2023-05-11 20:36:51 <crc         > vms14: no, offline, under termux
2023-05-11 20:37:02 <thrig       > it's not too hard, I've done it for TCL and Common LISP
2023-05-11 20:37:19 <vms14       > why not just your computer?
2023-05-11 20:37:29 <vms14       > programming on the phone sucks so much
2023-05-11 20:37:46 <vms14       > with a usb keyboard it kind of improves, but a pc is just better
2023-05-11 20:38:01 <vms14       > I could understand the use of a phone like a rpi
2023-05-11 20:38:10 <crc         > not really that bad. I work in a terminal environment, and use actual keyboards (I have several)
2023-05-11 20:38:16 <vms14       > for example to have a server 24/7
2023-05-11 20:38:33 <vms14       > the phone does not heat, and I'd bet the uptime is even better
2023-05-11 20:38:54 <crc         > The computer is less practical. I like being able to setup and work in a variety of places.
2023-05-11 20:39:11 <vms14       > crc won't you like a gpc micropc?
2023-05-11 20:39:19 <vms14       > it's a computer that fits on your pocket
2023-05-11 20:39:23 <vms14       > I love it xD
2023-05-11 20:39:52 <vms14       > a bit expensive, but if it breaks and I were able to buy another one, I'd do
2023-05-11 20:39:59 <crc         > I like the idea of that, but with my RSI issues, will end up needing to lug a keyboard anyway.
2023-05-11 20:40:09 <vms14       > hmm
2023-05-11 20:40:48 <crc         > My normal setup is a phone (s22 ultra) and ipad mini, with either an atreus or ergodox connected.
2023-05-11 20:41:46 <vms14       > did you try andronix + termux?
2023-05-11 20:41:56 <vms14       > it fakes a debian distro into termux
2023-05-11 20:41:57 <crc         > Just termux
2023-05-11 20:42:12 <vms14       > then you can use debian repo for arm packages
2023-05-11 20:42:42 <vms14       > you can also choose other fake distros, I just choose debian always
2023-05-11 20:43:21 <vms14       > you won't bind port 80 or do that kind of stuff, but you have a fake (and broken) root account
2023-05-11 20:43:25 <vms14       > and can create users, etc
2023-05-11 20:44:00 <vms14       > andronix is actually just a bunch of scripts that get copied on your clipboard and you paste on termux xD
2023-05-11 20:44:31 <vms14       > then it gives you a ./start-debian.sh
2023-05-11 20:45:17 <vms14       > if you rooted your phone then I assume you might be able to bind port 80
2023-05-11 20:45:28 <vms14       > althought it does not matter too much
2023-05-11 20:45:57 <crc         > I have everything I need under termux :)
2023-05-11 20:46:34 <vms14       > yeah, it comes with most common packages
2023-05-11 20:46:54 <vms14       > on that fake debian I've installed apache + modperl xD
2023-05-11 20:47:45 <crc         > I mostly just run RetroForth & Konilo, and Termux has enough other languages to let me build my VMs and do any non-iOS stuff I need.
2023-05-11 20:48:02 <vms14       > crc I think you're the first person that tells me he prefers to program in the phone than on a computer
2023-05-11 20:48:46 <vms14       > the cool stuff is the phone has also key bindings
2023-05-11 20:49:05 <vms14       > like the windows key
2023-05-11 20:49:11 <vms14       > the alt+tab also works
2023-05-11 20:49:20 <crc         > it's not really that different. I'm in a terminal, running the same stuff I do on my desktop. It's just a smaller display.
2023-05-11 20:49:43 <vms14       > yeah with a keyboard it changes a lot
2023-05-11 20:49:50 <vms14       > but I still prefer a computer
2023-05-11 20:55:33 <crc         > Nothing wrong with that
2023-05-11 20:56:26 <crc         > I've still got several (3) desktops setup; they just mostly don't get used by me now.
2023-05-11 21:25:13 <KipIngram   > I did that for a while.  Used termux (I couldn't think of that name last time it came up) and a bluetooth keyboard for most of my stuff.
2023-05-11 22:03:26 <KipIngram   > You know, I thin if I hardened my error recover so that it was capable of restoring all the registers that hold critical stuff in my system, I could arrange to be able to launch a metacompiled system from my Forth, and if it committed some sin (most likely would probably be invalid RAM access) the resulting exception would send me through the error handler and control would come back to my main system.
2023-05-11 22:03:56 <KipIngram   > The register restore would have to be done because the target system would have to have the regs set in a different way in order to be able to run at all.
2023-05-11 22:04:32 <KipIngram   > As it stands now the error recovery system relies on certain registers retaining their system critical value.
2023-05-11 22:07:51 <KipIngram   > Might be hard, though - just finding the restore buffer requires that the register that points to the system base address be set.
2023-05-11 22:15:02 <KipIngram   > I guess I could have a buffer to save those critical things that I could access with instruction-pointer relative addressing.  Then so long as the OS got me to my handler, I could get it all back.
2023-05-11 22:30:34 <tabemann    > I'd probably opt for some QEMU-type setup for that