| diana_coman: | talking of which hanbot there's a table of yours with resources around here, lemme know if you want coords | [02:15] |
| diana_coman: | clearly the bot could do with some more reliable movement strategies, probably some bounded/within area approach thing; shouldn't be too difficult to implement in there either if anyone takes the time to do it | [02:29] |
| diana_coman: | hm, 270 explore attempts -> 230 flotsam + 1 ord | [06:07] |
| diana_coman: | time-wise at this rate it would take about 10 days to get 20k flotsam; might be worth it just to check exactly *how* unreliable such estimates are; if I don't run out of tools by then ofc | [06:09] |
| mircea_popescu: | o nice. | [08:30] |
| mircea_popescu: | lemme see here | [08:30] |
| mircea_popescu: | 2371 attemts, 60+28 flotsam. and various other stuff | [08:31] |
| mircea_popescu: | of course im bh. | [08:31] |
| diana_coman: | lol; but yeah, bh can't compare as speed with tools for sure | [09:13] |
| diana_coman: | the most interesting part is that the data quite clearly indicates an area where one should pretty much search for flotsam if that is the main thing one wants | [09:14] |
| diana_coman: | you can find flotsam on a large area for sure, but if I stick to a smallish circle, I clearly find *mainly* flotsam | [09:15] |
| diana_coman: | might run some numbers later if I get the time for this as it looks quite interesting really | [09:15] |
| mircea_popescu: | hm. | [09:53] |
| mircea_popescu: | so while untouched, the table got filled with 1 units of cr | [09:54] |
| mircea_popescu: | is there some case where the bot confuses itself ? | [09:54] |
| mircea_popescu: | diana_coman yeah but with cs ? | [09:54] |
| diana_coman: | what with cs? unlikely you'll get flotsam with cs there, not related | [09:55] |
| mircea_popescu: | yeah. | [09:57] |
| mircea_popescu: | i'd like a cs f spot. | [09:57] |
| diana_coman: | re bot: you mean only cr? otherwise it sounds like a confusion regarding slot number basically - something threw it off but it's hard to say what exactly like this without details | [09:57] |
| diana_coman: | so go find yourself one? | [09:57] |
| mircea_popescu: | tryin' | [09:57] |
| diana_coman: | out of curiosity: how are you trying exactly? | [09:57] |
| mircea_popescu: | by mining :) | [09:57] |
| diana_coman: | bh? | [09:57] |
| mircea_popescu: | anyway re bot : what could i say to make it more detailed | [09:57] |
| diana_coman: | lol, how would that work? | [09:58] |
| mircea_popescu: | well atm im mining bh, but hey, i burned ~300 sticks last week | [09:58] |
| diana_coman: | any error messages it gave : maybe it did not find the table at some point for instance and then everything got missplaced | [09:58] |
| mircea_popescu: | it also has a pile of tiny keys in inv, for some reason, and a small cr + key that it didn't mine | [09:58] |
| mircea_popescu: | the tinies i guess it mined, unless it somehow threw out the enums | [09:58] |
| diana_coman: | but even with sticks how would that work other than sheer luck, I just don't see the approach you are taking there | [09:58] |
| mircea_popescu: | well yes, exactly. see where there's flotsam, see if it's reliably flotsam | [09:59] |
| mircea_popescu: | so far haven't found any. | [09:59] |
| diana_coman: | aha, so for some reason it failed when trying to move the key to the claim and then as a result the empty slot got messed up so it probably failed to move the table for some time too and then the slots got mixed up etc; basically lag | [09:59] |
| diana_coman: | but yes, it can go downhill like that at this stage if it catches some lag at the wrong moments basically | [10:00] |
| diana_coman: | well, after my previous combing of the island I'd say relying on luck like that might work if you are really really lucky only | [10:00] |
| diana_coman: | otherwise, fat chance | [10:00] |
| diana_coman: | the way I had found a grass spot was by plotting previous data and looking at the cracks essentially - quite a pain | [10:01] |
| diana_coman: | not even sure there is one for flotsam but I have to say I did not go through all the trouble | [10:01] |
| diana_coman: | what I'm trying to say is that you have there a field of overlapping haystacks in which you are naively trying to find a needle by ...lifting straws or something | [10:02] |
| diana_coman: | but anyway, sounds funny :D | [10:02] |
| mircea_popescu: | well... i might try combing my ancient finds library... not so sure how useful it'd be | [10:03] |
| mircea_popescu: | diana_coman so in your estimation, if i stack the cr in the first right slot, will this fiox it ? | [10:04] |
| diana_coman: | so what do you have in your table now? only cr 1 in each slot? | [10:04] |
| mircea_popescu: | normal products, then a bunch of cr | [10:05] |
| diana_coman: | uhm, so then it thinks that one of those slots is cr in fact when it is something else; trouble is: which one | [10:05] |
| diana_coman: | basically you can't really guess which one - you can have a look when it finds resources whether it gets them out correctly or not | [10:06] |
| mircea_popescu: | ima try this, got all the spurious cr (tiny product) in inv ; next it unloads that q cr it'll mix and hopefully it's fixt. | [10:06] |
| diana_coman: | it won't be | [10:06] |
| diana_coman: | because it will try to move it to a slot that is occupied | [10:06] |
| diana_coman: | and so the server will put it instead in the next free slot available in table | [10:06] |
| diana_coman: | and there you go again | [10:06] |
| diana_coman: | basically the only way to fix it is to figure out which slot it thinks cr is in | [10:06] |
| diana_coman: | the only way I can think of doing that is by looking at it for a while to see what resources it moves from where and go through elimination | [10:07] |
| mircea_popescu: | how's it occupied ? it's not occupied, it'll take the tinyq cr out of table, put it in inventory, put it back into table | [10:08] |
| diana_coman: | ugh | [10:09] |
| diana_coman: | mk, have fun | [10:09] |
| mircea_popescu: | lemme explain : currently table in order is sm b f cr wpl wpl cr b f sm | [10:09] |
| mircea_popescu: | there's two of each because 2 quals. | [10:09] |
| mircea_popescu: | before there was also cr cr cr cr after this ; it was all tiny quality, and i took it in inventory. | [10:09] |
| diana_coman: | aha | [10:09] |
| diana_coman: | so? | [10:09] |
| mircea_popescu: | now in the model in my head, next it finds a tiny cr, it will mine it, put the thing in inventory, | [10:09] |
| mircea_popescu: | which will mix with what is alread ythere (same quality) | [10:10] |
| mircea_popescu: | and then put it back into table | [10:10] |
| mircea_popescu: | so all good. | [10:10] |
| diana_coman: | the model in your head is stubbornly ignoring the bot's model which is amply described though | [10:10] |
| mircea_popescu: | how ? | [10:10] |
| diana_coman: | how do you think that cr ended up in a lot of slots instead of one single slot? | [10:10] |
| diana_coman: | in the first place? | [10:10] |
| mircea_popescu: | lag. | [10:10] |
| diana_coman: | nope | [10:10] |
| diana_coman: | not directly | [10:10] |
| diana_coman: | indirectly yes, but not as a direct result, how would that even be? | [10:11] |
| diana_coman: | what do you mean: it was laggy so slot 10 became slot 20 | [10:11] |
| diana_coman: | ? | [10:11] |
| mircea_popescu: | because "lag" is a wild card that can do anything ? :D | [10:11] |
| diana_coman: | or what? | [10:11] |
| diana_coman: | ahahahah | [10:11] |
| mircea_popescu: | i mean lag so the bot got a different result, undefinedly. | [10:11] |
| diana_coman: | could be, could be, oh mighty gods of lag | [10:11] |
| mircea_popescu: | wwell just found a cr tiny, lessee | [10:11] |
| diana_coman: | mnope, that's not how it works, but not sure what doesn't get through | [10:11] |
| mircea_popescu: | yeah it didn't mix the 33 in inv with the 1 in table | [10:12] |
| mircea_popescu: | but put it in a new spot. | [10:12] |
| diana_coman: | obviously | [10:12] |
| mircea_popescu: | let's see what it does with sm | [10:12] |
| mircea_popescu: | sm it mixed correctly. | [10:12] |
| mircea_popescu: | so the cr is misplaced ? with 1 other thing ? | [10:12] |
| diana_coman: | by the sounds of it, yes | [10:12] |
| mircea_popescu: | aha. | [10:12] |
| diana_coman: | basically and last time: bot thinks cr is in slot 2 let's say (or 5 or 8) but that is wrong: there is something ELSE there | [10:13] |
| diana_coman: | so bot tries to move cr to slot 2 but because that one is occupied | [10:13] |
| mircea_popescu: | but then, why doesn't bot take that thing out of inventory when trying to stack the cr ? | [10:13] |
| diana_coman: | server is uppity and puts it instead into next free slot | [10:13] |
| diana_coman: | and next and next etc | [10:13] |
| diana_coman: | hence all the 1 cr in each slot thing | [10:13] |
| mircea_popescu: | so by this model, wouldn't i see the wrong thing moved into invetory when it's trying to stack it and fails ? | [10:14] |
| diana_coman: | you might not see it for 2 reasons: it might be less than the whole stack and it goes too fast; it might be *more* than the whole stack and then server ignores | [10:14] |
| diana_coman: | so nothing happens as far as you can see | [10:14] |
| mircea_popescu: | oh wow, it doesn't take "all" stack but a finite count of items ? | [10:15] |
| diana_coman: | probably 2nd come to think about it, as otherwise you should see something remaining in inv | [10:15] |
| diana_coman: | well, it can't take all stack iirc so it has to ask exactly how many if it is a slot (because idiot client doesn't keep a count that is easily accessible) | [10:15] |
| mircea_popescu: | im spoiled because client has this ctrl-select which means "All" | [10:16] |
| diana_coman: | the slot itself would have that bloody count, but the bot can't access the slot | [10:16] |
| mircea_popescu: | but i suppose bot can't use that, has to compute huh. | [10:16] |
| diana_coman: | it's a bit convolute: each time you put the table down, the client creates a new object for that table | [10:16] |
| mircea_popescu: | so basically, this is unfixable, even if i stacked it in the right slot, the poor bot still wouldn't know what to take from there | [10:16] |
| diana_coman: | with a new set of slots etc | [10:16] |
| diana_coman: | well, it wouldn't matter if it takes less than what is in there or more, would it? | [10:17] |
| diana_coman: | ah, it would actually, gosh | [10:17] |
| mircea_popescu: | aha. | [10:17] |
| diana_coman: | the idiot server won't mix them in the table | [10:17] |
| mircea_popescu: | this may actually be the true avenue of the problem. | [10:17] |
| diana_coman: | ugh, so yes in fact unfixable, sigh | [10:17] |
| mircea_popescu: | lag made it take LESS from stack than there was | [10:17] |
| mircea_popescu: | and now has 2 stacks in table. | [10:17] |
| mircea_popescu: | i say this because of the fact that table slot in which it originally put the tinyq cr has 1 item ; whereas the smallq cr has 30. tinyq is much more common, so it looks like an off by 1 error. | [10:20] |
| diana_coman: | hey, it's open source - you get to fix it :D | [10:20] |
| mircea_popescu: | if my principal strategy in fixing problems i encountered weren't to whine at people, this would even have been good news for you :D | [10:20] |
| diana_coman: | lol, whine away; gotta build more immunity to that anyway, so I guess it IS good news :D | [10:21] |
| mircea_popescu: | lol | [10:21] |
| hanbot: | diana_coman> talking of which hanbot there's a table of yours with resources around here, lemme know if you want coords < wow srsly? coords yes please | [12:04] |
| hanbot: | diana_coman ^ | [12:06] |
| hanbot: | pesky arrows. | [12:06] |
| diana_coman: | 350 289 hanbot - basically at the far end of the village hill | [12:06] |
| diana_coman: | prolly grenadine fell off the cliff that way and lost the table | [12:07] |
| hanbot: | tyty. i went looking for it for ~20 mins the other day, gave up | [12:07] |
| hanbot: | kinda incredible sweeper didn't eat it? | [12:07] |
| diana_coman: | seems to be a rather lazy sweeper these days at least on containers from what I saw | [12:08] |
| diana_coman: | I was actually looking for mine when I found yours, lol; foxy had that kind of accident too after exploring in a line the whole night | [12:09] |
| hanbot: | ahahaha right on the tip. | [12:10] |
| hanbot: | well thanks for spotting it | [12:10] |
| diana_coman: | yw | [12:10] |
| mircea_popescu: | i'm currently exploring in a 60 long line by voyna's tent | [12:11] |
| mircea_popescu: | i love long lines | [12:11] |
| diana_coman: | they are perfect in some situations | [12:11] |
| diana_coman: | Foxy had a great time exploring like that on flat land under water, lol | [12:11] |
| mircea_popescu: | lol | [12:12] |
| diana_coman: | what, that's where stones get polished nicely! | [12:12] |
| mircea_popescu: | hehe | [12:12] |
| mircea_popescu: | diana_coman same thing happeend, with boulder this time | [16:20] |
| mircea_popescu: | i think it may have to do with the terrain here or such | [16:20] |
| mircea_popescu: | o look at that, 100 flotsam, 600 cr. | [21:24] |
| mircea_popescu: | i musta hit a big one | [21:24] |
Comments feed: RSS 2.0