Announcement

Collapse
No announcement yet.

13 Patches Published That Effectively Bring RadeonSI To OpenGL 4.5

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • kalrish
    replied
    Originally posted by atomsymbol View Post
    In my understanding, usefulness does not necessitate an objective. A measurement of usefulness can happen after many years, even when the authors are long after their death.
    Usefulness depends on an objective in the sense that something is useful to some ends (and not others). Whether something is useful thus depends on what you want to get. If making our lives more comfortable is part of the objective, science is useful.

    Leave a comment:


  • starshipeleven
    replied
    unapproved post for kalrish above, it will be shown eventually.

    Leave a comment:


  • starshipeleven
    replied
    Originally posted by kalrish View Post
    No. That an experiment set in accordance to some assumptions works does not imply that those assumptions are true.
    Yeah, we all knew you have no idea of proper experimental procedures because you are an idiot with moderate philosophical knowledge playing sophist vs Socrates but getting schooled hard by the latter.
    You ever heard of Scientific control? https://en.wikipedia.org/wiki/Scientific_control
    An experiment without control is invalid, and this since a LONG time ago.

    With such an assumption, you could go ahead and test that, and you would find it to be repeatable, yet that wouldn't prove there's such a god.
    And with scientific control you would get that also that by repeating the same test without the same assumption you get the same results as with it, so the answer is that the assumption is bullshit.

    Really, you cannot argue science if you don't know the basics of it. Consider your whole point invalid, I'm answering it only for fun.

    Newton's gravity laws are not valid because things do not behave as they state. What happens is that, at low speeds, the error is so small that our technology barely detects it.
    And this is enough to state that newton's laws are true at low speeds.
    Due to other obvious reasons you cannot claim there is an error if it is undetectable by the sensors used, so according to all evidence the above is still true.

    In any case, it doesn't matter if it's only in corner cases that the error pops out. There's an error, and that's enough. Of course you can build many things in accordance to them and they will work. And this, by the way, is another argument for what I have said above: two centuries ago they built things in accordance to Newton's laws, and they worked fine, yet that they work fine do not and can not imply that Newton's laws are true, for that would be contradictory.
    You seem to default the meaning of "truth" to "absolute all-encompassing truth that is true in any and all conditions".
    Newton's laws are true and correct within specific conditions, as they give the same answers as real-life measurements within these specific conditions, and still do so and will keep doing so.

    Really, isn't something true if it is true within specific conditions now? Like the statement "the letter o is in the following word: word" This statement is true or false depending on conditions (is letter o in it?) and on the observed phenomenon (the word).

    True ‘enough’!? That's quite mediocre logic for someone so fond of sciences.
    Yes, because by its own admission you cannot use science to get to the "absolute all-encompassing truth that is true in any and all conditions" because one of its requirements is that to state something "true" you need proof and to follow a rigid procedure. You can't have true definitive proof to reach this "absolute all-encompassing truth that is true in any and all conditions" as sensors will always have a non-infinite sensitivity.

    If you ever even tried to think of science like something that wanted to reach "absolute all-encompassing truth that is true in any and all conditions", you are a complete moron and you should study what science is before wasting everyone's time.

    This is not against science. What this is against are the beliefs that science is absolutely true and that there's an essential difference between science and so-called “superstition” or religions.
    Science never claimed to reach the absolute philosophical truth, but to reach things that are "true given this list of CONDITIONS", while religion or superstition always claim they are absolutely truthfully always true and then fail to provide any kind of proof.

    But both have assumptions, so they're not radically different.
    Well, every human idea is based on assumptions anyway so I don't think you can use this to say if something is different or not from something else.

    It's like saying that cars and skyscrapers are the not "radically different" because both are physical objects.
    You are using "radically different" in a wrong way, relearn english.

    An example of this situation is your beloved constitution, before which any two persons, being of course two different persons, are yet both people, i.e., not radically different.
    Don't assume that all the world is 'Murrca and everyone speaking english is 'Murrcan. You were lucky as my nation's constitution also states that, but many places lack a constitution or such statement in theirs.

    Leave a comment:


  • funfunctor
    replied
    I forgot to add one final point about truth in generality when one speaks about the scientific method vs. dogmatic paradigms.

    Critically - within the scope of the scientific method the resulting inferred model need not have to be true at all, it need only be useful as that is both sufficient and complete ! Dogmatic principles however typically _do_ require/consider themselves to be universally true however not necessarily directly useful. Truth and usefulness do not have equivalency.

    In fact, usually when you model you want to remove absolute truths out from your model so your results are not baked into your actual model aka "over fitting". The idea is to motivate a model from the scientific method such that it is the absolute minimum mechanics to reliably predict a set of results under a given set of constraints.

    Mathematics is kind of half way between these two schools of thought in a sense because Mathematics does require things to be true in a absolute sense however only up to that of the axioms. You may ask, isn't then Mathematics dogmatic in that you can make up your own axioms and claim things to be true, well yes and you would be right ! You can indeed build your own Mathematics, however would its formalism be as useful as the common one, possibly not. For the interested reader, this is hot topic in Mathematics today, where in the school of maths "hot topic" means its been raging for only a few hundred years Any way, look up HoTT vs. Set Theory..

    Cheers,
    Edward.

    Leave a comment:


  • kalrish
    replied
    Originally posted by starshipeleven View Post
    Yes, and if the child theoretical claims are proven true by repeatable experiments, also the assumptions are proven true, within the conditions tested by the experiment.
    No. That an experiment set in accordance to some assumptions works does not imply that those assumptions are true. Imagine, for example, that you assumed that there's a god by whose will you generally feel pain when you are hit. With such an assumption, you could go ahead and test that, and you would find it to be repeatable, yet that wouldn't prove there's such a god. That's cooking ‘truths’. Note that falsificationism doesn't have anything to say here, as to negate the assumption that there's such a god is as impossible as to negate causality. Both are beliefs. It doesn't matter if causality seems true, as that's just a sensation. In ancient times people had such sensations with deities.

    Furthermore, scientific experiments add their own assumptions to the theory behind them: they assume that many factors don't affect the result. That's necessary and ok for practical purposes –this is not what I argue against–, even though in rare ocasions it makes them fail, but renders them unsuitable for yielding truths –what I insist on–.

    Originally posted by starshipeleven View Post
    Newton's gravity laws are still 100% valid even now, within the same constraints where his laws were deemed valid by experimentation. (i.e. these laws fail to give results consistent with real-life measurements when the objects go at relativistic speeds, that is with objects that run at like 20% or more of lightspeed). But as long as you deal with non-relativistic stuff, you can use Newton's laws without ill effects.
    Newton's gravity laws are not valid because things do not behave as they state. What happens is that, at low speeds, the error is so small that our technology barely detects it. In any case, it doesn't matter if it's only in corner cases that the error pops out. There's an error, and that's enough. Of course you can build many things in accordance to them and they will work. And this, by the way, is another argument for what I have said above: two centuries ago they built things in accordance to Newton's laws, and they worked fine, yet that they work fine do not and can not imply that Newton's laws are true, for that would be contradictory.

    Originally posted by starshipeleven View Post
    The assumptions are tested and if they aren't validated they are discarded, what is deemed true enough can be used again in the future. Welcome to scientific method.
    True ‘enough’!? That's quite mediocre logic for someone so fond of sciences.

    This is not against science. What this is against are the beliefs that science is absolutely true and that there's an essential difference between science and so-called “superstition” or religions. There are plenty of differences between sciences and religions. Their assumptions are different. Their methods are different. Science builds rockets, religions do not. But both have assumptions, so they're not radically different. An example of this situation is your beloved constitution, before which any two persons, being of course two different persons, are yet both people, i.e., not radically different.

    Leave a comment:


  • dungeon
    replied
    Great, next 13 January is friday - release it then

    Leave a comment:


  • dungeon
    replied
    Just release mesa 13 on friday

    Leave a comment:


  • funfunctor
    replied
    Hi,

    Very fitting for the "final bits" of mesa to invoke a discussion about the merits of the scientific method. So much so I registered to this forum!
    Alright, let me try to explain the scientific method in a fashion, that perhaps, has not been explain in this manner before..

    The scientific method isn't, in fact, about "proving" 'things' as it is about "disproving" 'things'. Rather the scientific method isn't really concerned with the notion of "proving" at all and is more analogous to a statistical paradigm over one with absolute determinism. In fact, statistical methods are exactly the mechanism employed with dealing with the analysis of experimental outcomes.

    When a scientific study is designed, it is constructed in a manner that allows for primarily two hypotheses, the null hypothesis and the alternative hypothesis. The point here is to find *evidence towards* (key words here) *rejection* of the null hypothesis. This basically means you find evidence to "disprove" one possible explanation out of a sea of possibly infinite explanations towards "why" something occurred in some way. In other words, we don't "prove" or "disprove" anything in the strictest sense of the words and this is critically important to make note of! Further, we are only interested in getting rid of the highly unlikely ( above 95% confidence ) explanation and leaving us with a suggested alternative *not* proven alternative, just a suggested one coupled with a *plausible interpretation* such that further predictive modeling can be deduced.

    This method of reducing out the highly unlikely and suggesting plausible alternative that construct workable models with predictive capacity has been the quintessential element of the success of the scientific method as it is known today.

    As a motivating example, quantum mechanics actually says nothing what-so-ever about _why_ things act in the way they do however explains precisely the mechanics of exactly _how_ they do indeed behave in a computational manner. The subtle distinction of the _why_ and _how_ is of the most highest order of importance to understand as this is the kernel of a real understand what the whole point of science even is. I simply cannot stress this detail enough.

    I think it is unfortunate that the words "prove" and "disprove" get mixed up and confuse a great number people. The confusion is also compounded by mixing in how pure mathematics works over how statistical studies are done in the scientific method. Folks sometimes say things to the effect of "but isn't physics just mathematics?" actually no they are nothing to do with one another. It would be like saying "isn't shakespeare just english?".. I'll endeavor to explain that last bit now:

    Mathematics isn't so much a science as it is a language in the truest scene. In the case of mathematics there are two domains of thought, mine is from one known as constructivism and so I shall define from this paradigm. In constructivism, we define/pick/make-up a axiom or a set of axioms, these axioms can not be proven nor dis-proven we simply assert them to be true for the purposes of obtaining useful results. If the axioms you assert are well thought out you may obtain very useful and fruitful results. You may actually "prove" your results this time in the true sense of the word "prove" unlike the scientific method because we can either directly derive the result from the axioms in a deterministic manner by a set of rules or conversely assume some result to be true and then show it conflicts with the axioms and thus must then be false.

    A motivating example for the reader would be the proof of the square root of any prime is not a rational number. We can do this for the special case of p=2 by assuming that the square root of 2 is indeed rational and so cab be written as n/d where n and d are integers and d cannon be zero and that n and d are co-prime. If you follow this though for yourself you will see this becomes a contradiction. Thus this is a "formal proof" which is a very different thing to the scientific method.

    In fact, Mathematics does not care in the slightest about reality, it is agnostic of reality while Physics/Biology/Chemistry care exclusively about reality.

    I hope this is a useful and clear explanation to folks and helps clear up some of the misconceptions and terminology that is so often thrown about.

    Kind Regards,
    Edward.

    Leave a comment:


  • starshipeleven
    replied
    Originally posted by kalrish View Post
    To assume something is to believe it (a synonym in this context is «suppose»), usually as a preparation step for further theoretical claims.
    Yes, and if the child theoretical claims are proven true by repeatable experiments, also the assumptions are proven true, within the conditions tested by the experiment.

    For very basic stuff like causality, they have been validated in pretty much every experimental condition in the last centuries, so it is very safe to say that it is "absolute truth" or very close to it.

    Because if you find something where causality is invalidated it must be in a condition that was not tested repeatedly every second in the last few centuries. And pretty much all experimental conditions for causality are tested by now.

    By the way, truths, in the sense of absolute truths, would obviously be atemporal (which is practically equivalent to, albeit not the same as, eternal), while scientific theories change. That, along with the fact that science is based on assumptions, should be enough for anyone to notice that science doesn't yield absolute truths, which is what I first said.
    If you actually knew something about science and weren't an imbecile with moderate philosophical knowledge trying to pose as some kind of prophet, you would know that new scientific theories don't invalidate older ones.
    New theories are better because they cover more conditions than the older ones, but what was proven true before is still true now, WITHIN THE EXPERIMENTAL CONDITIONS that proved it true.

    The usual example is Newton's gravity laws and Einstein's gravity laws.

    Newton's gravity laws are still 100% valid even now, within the same constraints where his laws were deemed valid by experimentation. (i.e. these laws fail to give results consistent with real-life measurements when the objects go at relativistic speeds, that is with objects that run at like 20% or more of lightspeed). But as long as you deal with non-relativistic stuff, you can use Newton's laws without ill effects.

    Einstein's gravity laws give the exact same results as Newton's laws while within the same experimental conditions, BUT they also give results consistent to real-life measurements in areas where Newton's laws failed (relativistic objects).

    The latter is an EXPANSION of the former, it does not mean the old is invalid.

    Now, the course of philosophy in the last hundred years (approximately) has lead to the general consensus that it is not possible to set out to know without assumptions or to get rid of them, and that many assumptions are not chosen or that one is not even aware of them (our biology and culture being the most studied).
    The assumptions are tested and if they aren't validated they are discarded, what is deemed true enough can be used again in the future. Welcome to scientific method.

    Leave a comment:


  • pal666
    replied
    Originally posted by kalrish View Post
    It seems you didn't understand anything.
    it seems that you are idiot

    Leave a comment:

Working...
X