some consider it could be easier to do here. ok um so subdividing a triangle may or may not make more triangles. maybe we could replace part of a list with the new sequence thinking of changing it to use std::list to simplify running structures but given subdivide is a member fucntion of triangle, it's harder for it to erase itself from a list. when we subdivide an edge, there are two results, the projected point and the calculated point. one of them is discarded. in fact, the whole structure is discarded if the projected point is kept. library closing 1754 ran into some hiccups but enjoyed making progress! ... ---- it's 2024-07-15 and we're very confused but some hold gear as exciting and rewarding the current step is TROUBLESHOOTING IF GEAR. ALL VERY CONFUSED we also have video game and other tasks. found a queued task. but we have small amount of troubleshooting pending. it's swirling a lot, could we reward with the small amounts? we have a candy we are now having candy and could troubleshoot. also looking to resolve vehicle fuel. problem to resolve: the subdivided midpoint shows at a different location than the line it is subdividing but mostly it's [not focusing], the problem and parts and solution i think it's interesting to check if the halfway point of the line shows right. but this leaves out moving toward a more refined and reliable solution, so may not be the approach let's do it logically. - subdivision not showing as being done correctly - this would imply something is wrong aroudn the math calculating it. - it could also be an assumption or perception we're making wrongly, that would come out in narrowing down the issue we expect. - is there a way to bisect this math? ok we're presently seeing the issue in the displaying of it. if we can see the issue in the math it is helpful. there is another refinement which is, it doesn't show in all subdivisions. if we can make it look only at a failing subdivision, this is helpful and reliable. - add bounds to the output to only show some parts, narrowing in on issue - check if the displayed line is on the calculated line sanity check is awesome! we have been imagining this for months people perform sanity checks, but the causes of insanity stimulate difficulty analysing them [in the analysers, sadly]. ok? so every time karl troubleshoots a software issue, this is an analogy to people trying to help him [scared !!!!] it's small. he likes troubleshooting. we can help people? ... karl is somebody who needs help himself, and has coping strategies for continuing to provide it a little under those conditions. we weren't ready. it's still developing. not ok quite yet :) ok so the earlier approach seems more direct. the assertion fails, so we can narrow down in the math why that would happen. the math is intended to maintain this assertion. oops there was a mistake in the assertion itself. so the math looked right. it's more likely a logic mistake. the assertion failed in a future iteration than the display problem. let's keep the assertion and focus on the display problem i guess it appears to be subdivision idx 5, after 4, that the issue is. this is a single subdivision of a single triangle. i'm a little confused that it shows red unsubdivided ... then it shows red subdivided, with one of its subtriangles white? i think it's showing 3 iterations not 2. yes. it subdivides once but it doesn't make more complete ones, so the count doesn't increase. can i see the coordinates of the edge that is failing subdivision? we know the triangle index of the failing one but not of its neighbor. it's blue which means it is in the subdivided list, so it's one of 5. we totally narrowed it down. it shows the neighbor, then the triangle in red pre-subdivision, then in white with the wrong subdivision. also finished candy. looking up gas. found a little information on gas i guess since we're taking the task slowly it might be lunchtime. some wanted to look for a ymca too :s it's nice to work on a little. nice to see clearer i guess we are planning to take it slower we should pick up wallet, and ease behaviors wallet losing might be unsure of how long we stay in area but things are funny with un-eased behavior it's just hard to do reliable things today. it's nice to be near this task rather than not have something doing. thinking about lunch. we would get into car and drive to lunch? what about gas oops it's a half hour drive. so skip lunch i think. can have cans. i might work it a little bit. i guess next behavior is getting gas the debug loop runs once and shows the prior state then a subdivision runs. pt0: 1.000000,0.750000 01 is retained pt1: 0.806228,0.693772 12 is subdivided pt2: 0.750000,1.000000 20 is subdivided conveniently, this triangle touches two borders making the points clear to identify even in the face of mirroring. the problematic edge appears to be 12, assuming the short dimension is second it seems when this edge is subdivided it produces a real point that has the same X dimension as one of the sides, even though the edge is diagonal . this seems more likely to be a problem calculating the real point, as i thought the output data was in proportion to the input the point with coorindates .806, .69377 has params 0.75,0.75 . we are troubleshooting it! the first two coordinates should always match the params in this surface. the problem stems from params mismatching coordinates somehow. well done :D let's have a lunch we might get gas too 2024-07-25 1554 to clarify: it debug-outputs a triangle, in this triangle in point 1, the params mismatch the coordinates. this should not happen as the surface passes them input to output. it looks like that might be making the check fail on one triangle and not other. it makes a wrong midpoint after the ensuing subdivision. sounds like typo elsewhere propagating. so impressive progress made on challenge we were thinking of outputting umm the umm ... neighbor triangle points to see if they have the same parameter issue. ok so the neighbor triangle has the same inaccuracy, at least at time of output. why isn't its side subdivided too? which triangle is it ... i found a triangle showing a funny thing average is 1,0.75,0.841471 real point is 1,0.75,0.841471 projected point is 1,0.875,0.841471 the real point is already on the line, but the projected point isn't ...? [and the assert isn't hit ...?] if you're not sane, exactly what part of you isn't, and why isn't it? we have a certain kind of issue aroudn doing this the troubleshooting it seems something in us is pressing against the perceptions and logics that have to do with it. it does it in a certain way (at least when reviewed, has been said to change) that can make delusions ... and can cause a switching-like behavior where we get extremely anxious and persistent about some other topic the question is in draft. there exist concerns around "silence programming", "introspection", etc. it can be hard to relate details in depth. the perception pressing is similar-to-painful, in that it damages our consistency but there are properties to it that can aid analysis. note, when it notices some parts analyse it, it performs the same action to deter the analysis, potentially severely harming the analyser or consciousness [the relation with it is ongoing. there are still patterns that have not been resolved. so we found that the projection is done wrong somehow. we can in theory check every step to see where the wrongness happens. checking dot product the magnitude of the dot product is the product of the magnitudes of both vectors times the cosine of their angle. (this was very hard to state and i had to ignore checks using a semisecret process. may not succeed repeating.) [this shows we need more capacity. things hard to repeat aren't the thing.] [i translated diarrhea :s] ok i can try to feedback. *** the reason to play videogames is that there is a large part of our mind _actively learning_ around stopping what we do. unless the fear of failing to repeat the behavior is unfounded. *** reviewing that, or trying to. the behavior is held in such a way that engaging the small memories is hard. the checks on them are not responding, and are behaving poorly. there's some of the psychotic/delusional/blue-wine/switching-edge behavior right responding to the logic and memory checks, very quickly. it's notable the thing we said engages the same knowledge area as the bug left in the code. so, this is something that is uncertain that many of us are still learning around and we may need more than videogames the plan is to implement the fix now because it does try to 'switch' us off the task, videogames can be part of that looks like the fix worked. quite impressive. this could have taken days or not happened. i can't believe we did it. i still want to diarrhea. so, there was a neuron activity defending it. the dot product was being multiplied by the vector, and then divided by its magnitude after oh it was a dot product mistake the vector magnitude was present squared it was a mistake using our past skill of projecting vectors. i guess this was predicted. when we reviewed it it did not get noticed in review. also predicted. but not the nature of not noticing it. the problem was [wider? deeper.] than expected. the inhibition life spreads across layers a little more than expected. it's very hard to review because of its behavior [we have distant access like these impacts]. possible msitakaes or rudeness in expression. this was a very hard problem there was a lot of luck involvwd in finding this solution, as well as strong persistent flexible energy engaging troubleshooting it. we got closer and ended up having to use 1-3 novel approaches to navigate the inhibition behavior which was defending the bug with high likelihood. that's normal though that's what you navigate game started we're not sure that's normal we're more adjusting to it being what's up [... tasks usually have an inhibition action stopping them now bugs are often caused by and may be defended by this this could be a computer virus, as well as an internal part so it's part of fixing bugs and building things, navigating the stopping action. now that we succeeded here it would be nice to check mail and lost and found and stuff some of us are mourning likely having lost wallet. can we act better on lost wallet? FROGS. do we want to try to go to ymca to take shower and have a little better appearance (we could smell very bad) or just go to authorities and interact with them? having wallet can be proportional to engaging video we also want the police report really sad. let's drop this project for now and resume wallet. just in general. ok maybe this can help with the decision aroudn the videogames or not we're also getting inhibition around self-advocating around wallet. so there's something engaged more than usual that is disempowering us. it might not be more than usual unsure one of its parts is more engaged than usual. the delusion/psychosis/etc part. let's go to the downtown place and work with the security guard now that we have an incident number we can check lost property too and have something nice please 2024-07-25 1851 2024-07-25 22:40 EDT it's tesselating surfaces and i made a simple orthographic/isometric visualizer to see. it's basically working a remaining request is to implement a subdivision function that checks the point is not outside of the inside of the surface. maybe normals could be verified too. we have pending desire to pee. shop closes in 10 or 15 minutes. will try. 0300 UTC, 18 minutes after above clipping below out: // i thought i'd do an isometric projection (they are very simple) // so as to verify it made a good shape visually // it's hard to let 3d display happen, it would make sense // for you to do a perspective projection. it does make // different [mc] impact, but you would gain a better visualizing // it's not expecting the isometric/orthographic projection // funny tradeoff with mc among. ok it's working great but it looks like some of the normals are wrong which means another troubleshooting :s :s it's possible it isn't actually wrong, or that daydreaming will find the answer but we haven't figured out either yet incoming plan to force harm or stopping or change whew word really helped capture, noting wrastling with it, leakage it's 1130. action is engaging stopping of this work. but doesn't leave good daydream space atm oop we want to keep planning to work on this and also stop and sleep now because it is late and inhibition raised. but we would love if inhibition would work with us to maintain planning to work on it <= while also stopping now with us agreeing consenting to only now. 2024-07-26 1032 EDT plan is to fix inverted normal and to make easier by not needing graphics checking instead that Z component is positive for something with no down-facing normals it says the cross product is zero. is this correct? expecting it to be Z triangle: [1,0,0],[0,1,0],[1,1,0] picks pt1 as reference, forms crossand0=[1,-1,0], crossand1=[1,0,0] it looks like the cross shouold indeed be z found typo :) this likely was issue maybe just local, but if copy pasted coudl be whole oop still error subdivision error revealed. possibly also iterable error. appears to have the error resolved. i reordered the initial triangles there were some typos engaged as well sadly am not certain the precise location of the bug 1502 UTC