Hide Idle (>14 d.) Chans


← 2019-07-27 | 2019-07-29 →
shrysr: diana_coman: I've posted a summary on younghands. It's not as comprehensive as I;'d like... but i am getting better at capturing tasks and stuff from the discussions here.
shrysr: I guess it should also include plans for next week.
shrysr: I did.
diana_coman: shrysr: I rated you yesterday so now you have access to more of deedbot's functionality, use it wisely; esp note that you can !!up yourself in #trilema but note that speaking there before you really know your way comes with the danger of negrating and being turfed out
diana_coman: shrysr: re log eater, you know, it's not the worse one can be but the...digestion is the more important part there, as it were
diana_coman: re picture and updates of a blog post: the natural "update" is either another post linking to the old one or a comment; only where there really is a good reason, I'd go actually changing the post content as such and even then, in most cases it's an addition rather than a change (even if something is obsolete/to be deleted, it can be marked as such rather than removed)
diana_coman: the main reason being that deleting stuff you once wrote doesn't actually have much benefit for you (if it even has any) but it has some significant potential downsides (e.g. people linked to it and/or come back looking for that particular part of it and...it's gone, not even marked or anything)
diana_coman: from a purely technical perspective though, once I publish it, from my point of view it lives there and in the backups of the whole blog/site
diana_coman: I still keep the text files but more as drafts/different sort of backups rather than anything else
diana_coman: shrysr: do get into the habit of checking your moderation queue on your blog, wtf is this
diana_coman: and then read in #trilema Mircea's feedback and answer his questions.
shrysr: diana_coman: actually, i did not go through mp's article before i wrote it... but i just did and I can see why he says that particular point is utter fucking non sense
shrysr: in fact, my first instinct or thoughts were actually aligned to that article. but somehow... it still came out as 'meaningless', maybe because I think - my requirements in a situation are best known only to me, even if I endeavour to define it clearly. Relying on A's opinion (rating) about whether to trust B - is useful if they have had dealings with them, but this still does not translate as an absolute
shrysr: as to whether I can trust B. Perhaps saying entirely 'meaningless' is incorrect. i.e i'm saying that i think individual evaluation is needed, irrespective of any WoT ratings.
shrysr: shrysr: lol... i will get into the habit of checking the moderation queue.
diana_coman: shrysr: I *had* given you the reference before, so not going through it was pretty ...dumb, you know?
diana_coman: the point there is that you use the WoT to get the information *you* need in order to calculate *your* trust in a person
diana_coman: A's trust in them doesn't feed in as such, no, how could it
diana_coman: shrysr: a clear structure can help you loads to stay on track, you know? so you don't end up like that, starting with the right idea but then entirely ignoring it because you got sidetracked or whatever
shrysr: yes
shrysr: i;m trying to explain what led me to think ratings are ultimately meaningless in the new post. Hey I checked out the canonical post right after you gave me the link... i did not finish reading it then. not that its a valid excuse.. but i think i got the whole thing, but did not capture my thought process well..
diana_coman: shrysr: honestly, it sounds like you have the right direction as it were but you lack the discipline & structure to stick with it and you end up on a random-walk instead of a clear thought process
diana_coman: at any rate, up yourself in #trilema and at the very least thank Mircea for this feedback + answer his questions.
diana_coman: you get to practice !!up too
diana_coman: I'll be around the next hour or so
diana_coman: shrysr: you can paste a link to the logs in #trilema and the bot in there will quote it as in : http://btcbase.org/log/2019-07-28#1925199 -> it's because of ...
shrysr: diana_coman: okay yes - i intend to respond to mp on trilema.
shrysr: i just did.
diana_coman: and well done on voicing yourself
shrysr: hehe thanks..
shrysr: diana_coman: The relationship with a person, and your opinion of said person, i.e everything constituting any connection between you and somebody in your WoT is reduced to a single number. This means that the definition of the number is obviously variable from person to person. So when I view a graph reaching to a person of interest X - the number need not indicate the increased probability of the
shrysr: presence of any thing I require from X.
diana_coman: I suspect what you mean re "meaningless" is that "ratings as numbers are meaningless *by themselves*"
diana_coman: which yes, doh.
diana_coman: the trouble with the second part seems to be that you still carry some model that doesn't quite fit the task
diana_coman: what you require is information; how does probability come into it?
shrysr: diana_coman: ^
diana_coman: shrysr: your "universal" seems to mean "irrespective of persons involved" which ....no, of course not, how could it?
diana_coman: basically any instrument will "help" only to the extent that you can extract usefulness out of it
diana_coman: there is no "universal" instrument, think of it, what would that be?
diana_coman: shrysr: did you get to read the 2nd trilema article on the wot? re attacks?
diana_coman: in your concrete example, the trouble rests with ....Z, not at all with the WoT nor with the other 2; basically Z made *very poor verging on dumbass* use of the wot if all he asked (of people he did not know at all!) was "is p a good pimp?"
shrysr: no no thats all Z got as response
shrysr: diana_coman: 2nd article ? no ?
diana_coman: he either did not ask enough/correctly-formulated questions or he should have discarded then those ratings as meaningless basically; think of it: you ask 2 strangers "is X good?" wtf will the answer actually tell you?
diana_coman: basically you designed there a popularity contest, not a way to extract information
shrysr: lol >> I shd have mentioned Z had a verbose questionnaire which X and Y skipped and responded "P rocks"
diana_coman: let me fish that for you, it shows an actual attack as opposed to your basic misuse-case
diana_coman: it matters, you know? if I ask you specific questions and you answer with generic advertisment, the conclusion is I will negrate you for being an ad-machine
diana_coman: and depending on whether you are in turn rated by the advertised party, I might negrate *them* too for being most likely scammers
diana_coman: you see? plenty of information in there but yes, it depends on the skill of the one looking for it
diana_coman: basically the WoT won't help dumbasses *at all*
diana_coman: shrysr: I hope you're following the discussion in trilema too :D
shrysr: yes i did... i get it. I do, but somehow - I can't 'believe' the WoT ratings. Yes, despite being a couple of weeks, one cannot say it is impossible that i wd rate you a 10... however, why shd it not be interpreted that I'm just an impulsive dumbass whos ratings dont mean much? Somebody who rated me -1 would think so... somebody who rated me +5 would probably not. THe numbers mean different things to
shrysr: different people - and therefore cannot be translated as a 'trust' for any specific interaction ?
diana_coman: how does that "therefore" follow there?
diana_coman: there is no requirement to "believe" anything, that's the point
diana_coman: so one X thinks shrysr is an impulsive dumbass; so ..what?
diana_coman: are you fighting there some deep-drilled "fairness" thing ?
diana_coman: i.e. it's "not fair"
diana_coman: it's *because* the numbers mean different things to different people (i.e. depending on *context*) that *a* number can serve as indicator of *one* specific interaction
diana_coman: not for any, not universal, quite deliberately *not* that
shrysr: diana_coman: nope. not worried about "not fair"
diana_coman: good for you then; but you keep circling this "universal" thing and I can't (yet) see why/where it's got its hooks in you
diana_coman: so... what's bothering you at the fact that it's not universal?
diana_coman: think also from the other side: if it were universal, it follows that *all* possible interactions between people *have to* fit ...20 numbers?
shrysr: yes >>> and thats not possible right?
shrysr: no >>it is Said to be universal in the canonical article....or atleast universal in the sense that you will know it wont work fo ryou
diana_coman: not without significant loss of meaning, yes; therefore (basic proof here!) the assumption is wrong aka "universal"
diana_coman: lol, don't mix universal, yo!
diana_coman: i.e. *what* is or isn't universal
shrysr: see "The important parts here are the easiest to overlook : " in the canonical article ?
diana_coman: shrysr: do the right thing when you reference something: select the part you mean and then provide the link
diana_coman: your referencing so far is a *pain*
diana_coman: yes; if you select a part of it, the link will change to provide a specific hook to *that* part
shrysr: how do i link to a particular part of the blog article ??
shrysr: really ??
diana_coman: (moreover: even if there IS only one thing, you should provide it rather than ask the other person to find it)
shrysr: holy fuck... i did not know this
diana_coman: shrysr: yep, I mentioned it before
shrysr: i thought it worked for the logs!
shrysr: damn
shrysr: ok that makes things SOOOO MUCH better
diana_coman: shrysr: tell me from your selection *what* is said to be "universal" ?
shrysr: The WoT process
diana_coman: aha, the *process*
diana_coman: NOT the meaning of each number ffs
diana_coman: now, what is this process that we are talking about?
diana_coman: it's not really "the wot process" but the process of finding out something about an unknown party using the WoT
diana_coman: there is no wot process; the wot is not a process but a ...graph
shrysr: ok yes, understood.. I mean the process of rating people and relying on WoT to put together information that helps you with dealing with an unknown entity.
shrysr: this entire process
diana_coman: perhaps a simple analogy: the WoT is a tool like any other, like a microscope if you wish; now what is "the microscope process"??
diana_coman: sure, you will follow a process to find something out with a microscope but it's *your* process
diana_coman: I suppose that's part of the thing: you see the process as predefined and external to you, somehow secreted/imposed by the wot
diana_coman: it's not; and the quality of the results will depend more on you (i.e. whose process are we talking about) than on the wot
diana_coman: basically the WoT is a block of marble + a chisel; if it's Michelangelo's process with those tools, one can get David; if it's some 5 year old with the same tools, well...
shrysr: ok
shrysr: this feels acceptable to me.
diana_coman: lol, feels.
diana_coman: shrysr: as a side for now here , be careful with feel vs think ; here's a bit on that http://ossasepia.com/2017/01/25/feelings-are-helpful-but-not-for-idiots/
shrysr: diana_coman: ok > how do I make the selection to URL thingy work for my blog ??
diana_coman: shrysr: note that there is some trouble in that the original thing (which is working on younghands.club too) relies on the structure that apparently is sometimes different on different browsers; so there is the original thing described at http://trilema.com/2015/that-spiffy-selection-thing/ AND the latest find of the idiotic browser behaviour + a solution for it http://trilema.com/2019/proper-html-linking-the-crisis-the-solution-the-resolution-conclusion/
shrysr: okies.. I will go through that.
shrysr: ok i revise my statement to I 'think' that is acceptable. But I don't see the process as external -- I understand that each person has his own process and comes up with a number. But the processes could be totally irrelevant to the other, and I could be potentially blind to this fact despite thorough investigation. If we consider my earlier example, then can the whole process still be considered universal
shrysr: i.e why the whole process failed
diana_coman: shrysr: you are still confusing 2 processes there: 1 is "how do I give whatever ratings I give"; 2 is "how do I find out useful information about an unknown party from the wot"
diana_coman: both 1 and 2 are universal in the sense that they *apply* to all possible cases (i.e. they cover them) but not in the sense that they consist of predefined steps/recipe
diana_coman: the easy analogy is sex - it's universal aka can be done between any 2 people but not in the sense that it's predefined or with the same meaning or whatever other fixed quantity you try to attach to it
diana_coman: shrysr: makes sense?
shrysr: Ok - yes. you are saying that there are no pre-defined steps for both 1 and 2, and it will vary from person to person, depending on how they use their WoT. Ok. I'm good with that. But - I'm not yet good with the definition of universal in the canonical article selection.
shrysr: because it also goes forward to say "...If it worked you know it worked, and of what quality its results are. If it failed you know it failed, and why and how come. "
diana_coman: what's your trouble there?
diana_coman: "if the sex was satisfactory, you know it was and just *how* satisfactory; if it wasn't, you know it wasn't and why and how come"
diana_coman: the point there is: you have the means to *also* evaluate an outcome (yet another process!)
diana_coman: or in other words, the *absence* of information is some information in itself in the WoT context
diana_coman: at one extreme - if P is not in the WoT, that tells you something about them already
shrysr: So in my example, Z would know AFTER his inteaction with P that P does not rock. and does not house exotic whores. But can it be entirely Z's fault that he could not know this before hand via some custom complex procedure in WoT? Can any procedure no matter how complex (connected to data from the WoT) insure him from failure? Can the system Ensure that he will find out WHY it happened? how does it help
shrysr: him know that X preferred blondes and Y did not give a shit who/what he fucks -
diana_coman: shrysr: ah, so that's your "not fair", the failure-insurance; the world never hands a failure-insurance, there is no such thing
diana_coman: the trouble with your Z is that he doesn't have a clear understanding of what "a pimp that rocks" means basically; that's his failure and nothing to do with the WoT; I suppose at one extreme you can think of the WoT as a sort of Oracle - if you ask stupid questions, you'll get useless answers
shrysr: oh lol! I did say that Z prepared a questionnaire
shrysr: lol
diana_coman: but from all your "counterexamples" , he had no idea how to interpret the data basically
diana_coman: it's not even that he needs a questionnaire; he just needs to A. clearly define what HE means by "pimp that rocks" B. look for the data that either supports or rejects the hypothesis "P rocks"
diana_coman: absence of some required bit of data for "pimp that rocks" is automatically supporting "P does not rock" (some parts may directly imply it even, not just supporting)
diana_coman: possibly the usual "exceptionalism" trips you over as in "oh, he IS actually a great pimp, just it so happens he's new and so nobody yet knows him in the WoT"
shrysr: yes that too
diana_coman: there is no such thing; by the definition of terms, if nobody yet knows him then and THEREFORE he is NOT actually a great pimp
diana_coman: he might become one *in the future*
diana_coman: but he's not, NOW
shrysr: right..
shrysr: ok
diana_coman: shrysr: you all right there? lol
shrysr: :) yes.
diana_coman: good then :)
← 2019-07-27 | 2019-07-29 →