[17:31:37] NASSP Logging has been started by n7275 [17:31:39] hey [17:35:11] morning! [17:37:03] hey guys [17:38:22] what's up? [17:39:50] got my ass kicked by Faraday's 2nd law last night [17:40:30] hahahaha how so? [17:44:06] datebook says each fuel cell uses about 0.000324 g/Amp/second of hydrogen [17:46:01] Faraday says 2.02g/mol/(2*(96500Coulombs/mol)) [17:46:16] which are very different [17:47:16] like, by a factor of 31... [17:47:37] I'm probably just having a moment [17:48:06] hehehe [17:56:38] I'm sure I'll kick myself when I figure it out, lol [20:36:43] oh god... [20:38:45] it's exactly 31 [20:39:41] ...which is the number of cells [20:39:55] lol [20:54:20] I feel very smart right about non [20:54:27] *now [21:06:27] hahahahaha [21:06:29] nice! [15:27:41] hey [15:29:20] good afternoon [15:31:02] I think I finally found the right solution for this LVDC orbital guidance [15:31:04] Hey guys [15:31:53] remember how the Space Shuttle was basically ready in 1979, but then one thing, the thermal protection tiles, caused a delay of more than a year? That's how I basically felt :D [15:39:12] now that I probably solved it, the person I contacted at NASA who had access to a Equation Defining Document can send it to me, so that I have to do it all over again [16:14:18] wouldn't that figure, haha [16:35:22] hey [16:35:33] hey Alex [16:38:16] finally making progress with LVDC orbital guidance, so I'm back to the stage where I can test fly missions [16:38:24] hi alex [16:46:57] morning! [16:50:11] indy91, good to hear [16:50:20] hey [17:22:13] hey Mike [17:22:37] I found a new piece of Comanche 45 info last night [17:22:49] I remembered a report that had word counts for most of the programs [17:22:56] took me a bit to figure out how they were counting [17:23:08] but from what I can tell, rev 45 has exactly 5 more words than rev 44 [17:23:10] what document is that from? [17:23:21] https://www.ibiblio.org/apollo/Documents/a042186.pdf [17:23:45] 5 words works perfectly [17:23:51] a jump somewhere [17:23:53] a jump back [17:24:03] VLOAD ABVAL [17:24:05] DELVSIN [17:24:08] STORE DELVSAB [17:24:14] yeah [17:25:35] so another piece of good information that doesn't really help us solve the puzzle :D [17:26:25] hahaha yeah :D [17:26:58] but it at least tells us that they didn't do a lot more changes somehow to banks 16 and 17 [17:27:11] also as soon as the air force opens more fully back up after the pandemic, I'm going to attempt to FOIA the records used to make this [17:27:56] you mean Rome Laboratory? [17:27:58] yep [17:28:29] "FOIA requests for records located at AFRL/RI can be accomplished via e-mail to rrs.foia@us.af.mil, or by sending a written request to AFRL/FOIA, 26 Electronic Parkway, Rome NY 13441-4514." [17:28:42] that's a can of worms that could turn out to be whales [17:28:57] hahaha [17:29:01] but how do you even know what to request? [17:29:42] this report says "This report provides a description of the data delivered to RADC for inclusion in a Software Data Repository.", and gives the number of records in that data [17:29:58] so I can describe what and where it is reasonably well [17:31:00] so we could just randomly request the Saturn V EDD from MSFC, as I am fairly sure they have one [17:31:13] sent an email 3 days ago, no reply yet [17:31:51] like requests to put things back on NTRS, it would probably take a few months [17:32:00] the argument that I can make in this case is that the first (Solarium) and final (L210/A72) lunar AGC programs are already publicly available, so they're not releasing anything that isn't publicly known in terms of capability or anything [17:32:06] it's just making the change history more clear [17:32:15] yeah [17:35:54] Solarium being a lunar AGC program is funny [17:36:03] the LVDC version for Apollo 4 and 6 is lunar [17:36:17] Solarium is like: get back home as quickly as possible WITHOUT going to the Moon [17:36:20] lol, I was trying to exclude Skylark/Digfly from that list but I guess "lunar" isn't the best way to do it :P [17:36:25] right [17:36:52] everything before that PCR that removed lunar capability :D [17:37:17] hehehe [17:37:19] PCN SL410 [17:37:25] a sad PCN [17:40:01] possibly the saddest of all [17:40:44] although some black friday-associated ones may have been similarly sad for the developers :P [17:41:54] what was done on black friday again? [17:44:20] I don't know any specifics, but that was when Tindall first came in, and held a bunch of meetings to remove a ton of features from the AGC code to make sure it would fit in the available memory [17:44:38] Don has a printout from that day that might have some details [17:48:22] man if we can actually get these records though [17:48:28] we'd be able to reconstruct every revision [17:56:32] they had planned to give the AGC the same capabilities as the IGM [17:56:37] that was one thing they cut then [17:58:16] and what I found interesting in the notebook from Mike Collins was the planned feature for Colossus IIA [17:58:26] "Boost polynomial 2nd and 3rd stage" [17:58:44] oh that is interesting [17:58:54] wouldn't be better than an astronaut and a chart though, probably why they didn't implement it [18:09:54] hmm [18:10:03] that five words is tricky [18:10:29] that might actually be a bigger hint than it seems on the surface [18:10:41] because you're right, it's exactly enough for the three new words and a jump back [18:11:05] but that doesn't account for the fact that something else needs to be deleted in S40.1 in order to make room for the first jump [18:11:19] and whatever goes away presumably must also be accounted for in the new 5 words [18:13:20] of course this also assumes that the report is correct, and that I have the right number of words in rev 44 lol [18:15:14] haha oh god [18:17:29] can't move the entrance to S40.1 either, it's called from bank 24 [18:18:03] and they wouldn't put the calculation before the check on XDELVFLG [18:19:07] yeah [18:19:08] hmm [18:19:39] and DELVSAB gets used not that many lines later [18:19:46] fairly contained range of words [18:26:11] yeah how is this five words [18:28:00] well how are they counting [18:28:06] they probably have to move some words [18:28:10] in addition to the 5 [18:28:31] this is only giving us total number of words in the assembly [18:28:38] so the total increased by 5 [18:28:56] seems correct [18:31:15] not sure if you tried that version, but maybe they didn't store anything in DELVSAB [18:31:50] uhhh what do you mean? [18:32:33] just had the ABVAL of DELVSIN in MPAC [18:33:02] I feel like the Norton document wouldn't be showing DELVSAB = |DELVSIN| in that case [18:33:31] without a black bar on it [18:33:41] yeah [18:35:45] hmm, even with a limited number of places where it could be, still so many variations... [18:37:06] so if I wanted to try and help with the problem, what is the best way to do it? I don't even have a setup where I can run yaYUL on anything [18:40:56] I'm not sure yaYUL even runs on anything besides Linux? I think Mike wrote pyYUL, that might work. [18:41:42] I mean, I have a yaYUL.exe :D [18:46:25] lol, no way in hell can we use pyul for this yet [18:46:41] the only listings that are in punch card form are Solarium, Retread, and 80% of Aurora [18:46:57] Then running yaYUL MAIN.agc from the console in any rope directory should produce a bin file for you [18:47:00] I feel like we just need to go through all variations of the jump location to bank 17 [18:47:30] I did manage to create .bin files, but it said 0000/2000 words used for every bank [18:47:49] weird [18:47:59] if you can get a virtualagc checkout to the point where you can run "make" in e.g. the LUM69R2 directory, that will be the easiest way for you to try [18:48:51] https://www.ibiblio.org/apollo/download.html#Windows [18:49:16] I have a folder with the most recent git version [18:49:28] and another where I installed that Windows version [18:50:24] I can figure all that stuff out [18:50:36] but where can I check how many words in each bank, banksums etc? [18:51:07] `make` in the Comanche045 folder will automatically run the banksum checking script and tell you by how much your banks are wrong [18:51:16] oh ok [18:51:22] let me try to get to that point then :D [18:51:46] I'm going to have to formally commit all of the Comanche 44 stuff too lol [18:51:48] I'll do that tonight [18:52:59] wait which branch do I need [18:53:08] there isn't even 44 or 45 in master [18:53:25] just comanche045 I guess [18:53:39] if you want to mess with it today, get yourself to the point of being able to build LUM69R2, and then I'll give you my work-in-progress folder to play with [18:53:56] I'll try [18:57:46] this is fun on Windows [18:57:48] not [18:57:49] hehehe [18:59:10] Yeah, might be easier to just set it up on Linux TBH, [18:59:53] I probably already have done so on my laptop [19:01:18] I'm pretty sure Ron also provides a linux virtual machine image that's already set up [19:01:24] because of the annoyingness of this lol [19:02:00] from 3 years ago though [19:03:23] yeah, if anything I'll start my laptop [19:17:36] STORE UT [19:17:36] VLOAD ABVAL [19:17:37] DELVSIN [19:17:37] STCALL DELVSAB [19:17:38] S40.1 [19:17:47] putting that at the end of bank 17 makes the bank 17 checksum very close [19:18:11] what about 16? [19:19:04] I'm just trying various things in bank 17 right now [19:19:12] ah ok [19:19:20] no idea how to get to this chunk from 16 yet, or if it's even possible :P [19:19:36] with "STORE UT", it wants to jump back to S40.1 +9D [19:19:56] which might be a bit too early in the function [19:23:02] ok, Virtual AGC on Windows plan has been abandoned :D [19:23:21] hahaha [19:36:11] how in the heck is this only +5 words [19:39:07] I'm not sure what you mean by that [19:39:16] I thought it is fairly clear we know the 5 words [19:39:45] two jumps, the three lines to do the calculation and the difficult part, moving words around [19:39:49] so that labels stay the same [19:40:02] just saying it like that makes it sound easy [19:40:06] but doing it is extremely difficult [19:40:13] unless I'm missing something obvious [19:40:37] jumps in interpretive normally take 2 words, not 1 :P [19:40:42] GOTO [19:40:43] WHEREVER [19:40:50] ah right [19:41:18] maybe something can be made a word shorter [19:41:31] you can sneak in a single word jump by turning a STORE into an STCALL [19:42:29] there are some STOREs, but you always need the MPAC to stay intact [19:45:19] did we look at the theory that something much later in S40.1 was moved, and the DELVSAB stuff is already as in later revisions? [19:45:47] not a lot [19:46:08] this is why it will be helpful to have multiple people trying things lol [19:46:14] there's a lot of possibilities [19:47:06] yeah [19:47:39] could be anything down to THETCON [19:47:54] the very end might make the most sense though [19:48:03] right [19:48:06] lets you jump to bank 17, and then straight to 24 [19:48:13] instead of going back to 16 [19:48:15] seems clean [19:48:50] maybe the last six lines [19:49:20] STORE VGTIG in S40.1B still there [19:49:23] then the goto [19:49:34] hmm no [19:49:43] that removes one word too much [19:51:04] can't you have an empty, nonsense word in there for convenience? :D [19:51:13] lol [19:51:22] possibly [19:51:47] but we can't have all 6 of those words in bank 17 anyways [19:51:50] that's one too many [19:53:37] oh there's a flag set there [19:53:40] there's another trick [19:53:42] SETGO [19:58:13] rarely seen a SETGO, but it could work [19:59:18] the thought process could have been, move the last few lines to bank 17 and then make it somehow work [20:01:37] hmm [20:01:45] also could have moved thetacon and the last 3 lines with it [20:06:45] true, that would be a bit cleaner [20:08:35] doesn't give the right checksum though lol [20:26:53] thewonderidiot, all set up on Linux I think [20:27:02] fun fact, German word for Desktop is Schreibtisch [20:27:11] that's the name of the folder on Ubuntu for me [20:27:26] only thing that failed was creating the Desktop shortcut for Virtual AGC :D [20:27:26] hahaha [20:27:30] just had to fix that in the make file [20:27:32] makes sense :D [20:27:53] I think I can build now, although I only did make install [20:28:13] got a LUM69R2.buggers file [20:28:22] if you go into LUM69R2 and do "make", you should see it running the check_buggers script [20:28:49] indeed [20:28:56] perfect [20:29:10] now let me undo everything I've done here so I can give you a clean Comanche 44 [20:29:39] ouch [20:29:56] I am just on master so far, should I switch to comanche045? [20:29:57] it's okay, what I was trying was clearly not working lol [20:30:00] nah, no need [20:32:06] https://drive.google.com/file/d/1phCZKp9CSR7GMOli2d-VttpOoBGAdxRm/view?usp=sharing [20:32:22] hmm [20:32:25] put that Comanche045 folder into your virtualagc folder, and you should be able to just run "make" in it [20:32:48] right now its content matches Comanche 44 [20:32:50] I might duplicate [20:32:56] I did duplicate [20:32:58] lol [20:33:08] for my convenience, can you post that link again :D [20:33:31] it didnt help me on my desktop PC... [20:33:41] https://drive.google.com/file/d/1phCZKp9CSR7GMOli2d-VttpOoBGAdxRm/view?usp=sharing [20:33:44] thanks lol [20:35:21] ENDSLECT in JET_SELECTION_LOGIC marks the end of bank 17 in Comanche 44, by the way [20:35:35] so when messing with bank 17 I've been defining stuff at that label [20:35:48] bugger word mismatch, who would have though [20:35:49] t [20:36:26] Makefil:21: recipe for targrt 'diffComanche045sums' failed [20:36:31] Makefile:21: recipe for targrt 'diffComanche045sums' failed [20:36:36] uhh, more typos [20:36:43] diffs 10666 and 22500? [20:36:46] you get what I mean, it showed that [20:36:54] yes [20:36:57] cool [20:37:04] now make them be 0 :D [20:37:15] why does it have that failure? Something with the script? [20:37:44] or does it that show that when there is a mismatch [20:37:48] the exit code for the script is the number of banks wrong -- and `make` considers any nonzero number to be an error [20:37:54] ah ok [20:37:56] so the build fails if the bugger words are wrong [20:38:20] what editor are you using? [20:38:36] vim [20:39:15] I'm normally editing in vimdiff mode, with `vimdiff P40-P47.agc ../Comanche055/P40-P47.agc` [20:39:33] so I can pull stuff over quickly from Comanche 55 as I break things in 45 and have to undo them [20:41:45] ok [20:41:47] back in a bit [21:16:27] I know why you like vim, it looks like Dwarf Fortress to me [21:16:33] vimdiff at least [21:30:17] hehehe [21:34:09] trying to just build something right now [21:34:12] getting used to vim [21:34:31] where do you usually place the bank 17 code? [21:39:11] You can always run vimtutor if you want a nice walkthrough on how to use its features ;) [21:39:27] at ENDSLECT in JET_SELECTION_LOGIC.agc [21:40:14] right, end of bank 17 [21:40:18] you can try to place it earlier, but I think there are some inter-module references to some stuff pretty late in bank 17 [21:40:21] Thymo, sounds useful [21:48:03] yay, it builds [21:48:19] just did a simple jump to bank 17 and back [21:48:27] nice :) [21:48:27] of course I now have 10 bank errors [21:48:30] hahaha [21:48:57] yeah the whole thing is pretty sensitive to word count changes :D [22:25:57] night!